Upload
khangminh22
View
0
Download
0
Embed Size (px)
Citation preview
Development of a Measure of Perceived
Entropy
I^iíthin The Famí1ies of Respondents
by
Deborah J. PhiliPs
A Thesis
Presented To The University of lvlanitoba
In Partial Fulfillment.of The
Requirements For The Degree of
Master of Arts
in
Psychology
Winnipeg, Manitoba
I 985
Deborah J. PhiliPs
DEVELOPMENT OF A MEASURE OF PERCEIVED ENTROPY
I{ITHIN THE FAMILIES OF RESPONDENTS
DEBORAH J. PHILIPS
A thesis submitted to the Faculty of Graduate Studies of
the University of Manitoba in partial fulfillment of the requirentents
of the degree of
MASTER OF ARTS
"/@ ,1985
Permissio¡t has beert grattted to the LIBRARY OF THE UNIVER'
SITY OF MANTTOBA to lend or sell copies of this thesis. to
the NATIONAL LIBRARY OF CANADA to microfilnr this
thesis and to lend or sell copies of the film, and UNMRSITY
MICROFILMS to publish an abstract of this thesis.
The author reserves other publicatiolr rights, artd neither tlte
thesis nor extensive extracts from it may be printed or other-
wise reproduced without the autltor's writte¡r perntissiott'
BY
Abstract
TheFamilySystemsliteraturedealingwithentropy
contains h1çotheses that certain " entropic" f amily
interaction patterns are dysfunctional' and may lead to
problems in the children (or other members) of these
families. case study reports have described entropic
families as behaviorally and attitudinally rigid, and as
lacking calmness and cohesion, knowledge of individual
members' thoughts and feelings and/or clear norms" These
and other possibte characteristics of these " entropic'o
families fatl on a continuum, and may lead to increasing
problemsastheybecomemorepronounced.However,êñ
instrument to assess the degree of entropy present in a
family envíronment has until this time been lacking" such a
scale has been recently developed by this author, and as the
subject of this thesis its psyci:ometric properties have been
explored" The instrument, or Entropy Scale' measures
respondents' perceptions of interaction patterns in their
families. It is eomprised Of 33 entropy items, and ís
complemented by various items which will aid in making
future normative comparisons. The subjects were 28l- male
and female unversity students, between the ages of l8 and
2Q, selected from Introductory Psychology classes at the
university of Manit,oba. summary statistics were obtained
frorn their responses, and analyses of the Entropy scale's
ternporal stability, internal consistency and factoral
structure were Performed "
Temporal stability was assessed. for a one-month
interval, and the instrument was found to be reasonably
stable for this period. Itên-total correlations were
calculated to determine the internal consistency of the
scale, and the Cronbach's. Alpha obtained revealed that the
overall level of internal consistency of the scale is good'
Item distributions htere examined for normalcy, and were
found to be normal or positively skewed for the most part"
This contra-indicated a risk of correlation attenuation,
which may have led to deceivingly low factor loadings in the
factor analyses. The scale was correlated with a measure of
social desirability, and it was determined that social
desirability bias was not sufficient to invalidate the
subjects' responses. Finally, a factor analysis was
performed to explore the factoral structure of the Entropy
Scale. The factors, which were extracted with a Principal
Axis Factoring procedure, correspond quite well to several
aspects of entropy which have been described in the
literature "
Ideas for future research 'wittr the scale have been
proposed, such as administering ít to "cIinical" groups of
delinquent and schizophrenic adolescents. It has been
hypothesized that entropy may Promote schizophrenia and
delinquency in adolescents, especially in combination with
ottrer dysfunctional interaction patterns such as
centripetality and centrifugality" Thus it will be possible
to investigate the relationship between various aspects of
entropy in the fanity and specific problems experienced by
family members" It may also be possible to eventually make
diagnoses and plan treatment according to scores obtained on
the Entropy Scale, and on the items of particular factors"
The determination of the scale's psychometric properties,
which was the purpose of the present study, will form the
groundwork t.or these and other f uture uses of the scale "
1
List of
Ëist of
Chapter
I
Table of Contents
Tables
Figures
General IntroductionPresent Evaluation Methods
General Systems TheorY
Centrípetality and CentrifugalityRelation to Schizophrenia and DelinquencyThe Notion of EntroPY
Entropy in the LiteratureModels of EntroPYNeed for an EntroPY Sca1e
Composition of the EntroPY ScaleAim of the StudY
SubjectsInstrumentsProcedureAnalyses
Summary StatisticsContent ValiditYTemporal StabilitYInternal ConsistencYFactor AnalysesRelationship of the Entropy Scale to
the Social DesirabilitY Scale of
Page
iiiiv
I6
B
9
1lL4
l6l92L
24
26
2A
2B
29
30
3l32
34
34
35
36
2 Method
the Personality(Jackson, L967)
Research Form AA39
3 Results
ii
42
42
4B
49
52
54
5B
65
6B
6B
77
BIB3
83
92
94
Y4
r09
L22
L24
r41
Summary StatisticsContent ValiditYTemporal StabilitYInternal ConsistencYFactor Analyses
Item Composition of FactorsRelationship of the Entropy Scale to the
PersonalitY Research Form AA(Jackson, 1967)
4 DiscussionSummary StatisticsTemporal StabilitYInternal ConsistencYFactor Analyses
Interpretation of FactorsFactor Summarization
Relationship of the Entropy Scale to the SocialDesirabilíty Scale of the PersonalityResearch Form AA (Jackson, L967)
Conclusions and Future Research
Appendices
A Instrument Packet, Including Directions toSub j ects
B Jackson's Personality Research Form AASocial DesirabilitY Subscale
Histograms of Entropy Itern Frequencies
Rotated Factor Matrix, Varimax Rotation
References r04
D
E Entropy Scale Factor Loadings,Rotation, Pattern Matrix
ObliminL42
F Summary Statistics,Jacksonrs PRF
Responses of Subjects toSocial DesirabilitY
111
L44
Scale ------ 143
c Correlation matrix for Entropy Scale
I
Table
t0
I1
L2
t3
L4
15
Tables
Responses Based on Items Using.."Mother"'"FatTIer " or "Wltole Fami ly" asthe Referrant
Perception of lrlho in Family Has Most Power
Subjects' Evaluations of Some of fheir Actionsby Self and Others
iv
Page
44
45
47
5l
53
57
5B
59
2
3
4 Entropy Scale Test*Retest Product-Momentðôrrelations, Means and StandardDeviations
Entropy Scale Item-Total Statistics
Summary of EntroPY Scale Factors
Factor I: Togetherness and Tranquility
Factor II: "War & Peace" in the FamilY
Factor III: "Transparency" of FamilyMembers
Factor IV: RigiditY and ConformitY
Factor V: ClaritY of Power and Rules
Factor VI: Social APPearance andInterpersonal- En joYment
Factor Correlation Matrix
Relative Contríbution of Each Factor
a Wish
5
6
7
I9 60
6L
61
Entropy Items Most Influenced bYto Seem SociallY Desirable
62
63
64
66
Chapter I
Development of an EntroPY Measure
General Introduction
schizophrenia and delinquency are two widespread
phenomena which pose serious problems to society, due to
their effects and their resistance to treatment " For
example, schizophrenics represent approximately 50 per cent
of resident populations of mental institutions, and 20 per
cent of first admissions. High readmission rates
(approximately 50 per cent within two years) contribute to
the former percentage (Mosher & Menn, L978) " As well,
disappointingly low levels of psychosocial functioning are
reached by most discharged schizophrenics. T?lis last fact,
as well as the expense Of caring for so many of these
individuals, Ttas contributed to a heavy financial drain on
socíety" In addition to the costly services rendered'
potentially valuable contributions of these individuals are
lost" Si¡nilarly, the FBI Crime Report states that in L97B
crimes such as murder, rape, assault, and robbery occurred
on the average every 30 seconds, with a disproportionate
percentage of the crimes being committed by teenagers (i'e"
juveníle detinquents) " This delinquency too is costly, in
terms of human suffering as well as the financial burden on
society. Despite the great need to ameliorate the societal
costs of juvenile delinquency and schizophrenia through
2
understanding and prevention this has not occurred, possibly
because ttrere appear to be no single causative agents.
Evidence of a genetic contribution to schizophrenia has
been found by many (Rosenthal, L97L; Gottesman and shields,
1971i Rosenthal, Wender, Kety, Schlusingêt, Wilner & Rieder,
f975). Biological correlates such as abnormal brainwave
patterns (Iti1, L977); left hemispheric dysfunction
(girchsbaum, L977) i and autonomic overarousal (ltednick &
Sclrlusinger), 1968 , Lg73) have also been found- Brain
chemistry has been inplicated, especially the excessive
activation of dopaminergic receptors (Post, Fink, Carpenter
& Goodwin, L975) "
Biological correlates of juvenile delinquency have also
been examined. Both adoption and twin studies of older
criminals .ld psychopaths point toward a genetic
vulnerability for antisocial development in certain
subgroups of adolescents. Abnormal brain wave patterns
(Arthurs & Calhoun, L964) are among the main factors
invest igated .
on the side of psychosocial determinants of
schizophrenia, the intrusive, overprotective, schizophrenic
mother ( ¡,tark, 1953 ) has f igured prorninently in much
research. Bowen ( 1978) described enmeshed, ê9o-fused
families where the transfer of anxiety proceeds from one
member to the patient. Lidz, Fleck and Cornelison (fOeS¡
3
described marital schism and marital skew, which appeared to
be linked to schi zophrenia: via rnarital conf Iict and
alliances between parents and children involved in the
former pattern, and ttre passive acceptance of the
psychopathology of a dominant parent by a submissive one
involved in the latter " Bateson, Jackson, Haley and
lrleakland (fgSe ) found the double bind, in which a chitd can
neither comment or, withdraw from nor ignore two
contradictory messages, to be important in the background of
schizophrenics "
Much of the above-mentioned research on ctinical
populations is observational in nature, lacking control
groups and relíable, blind ratings. Better controlled
research has led only to partial support for many findings.
For example, Jacob (fSZS¡ performed a comprehensive survey
of direct observat.ions of farnii-y interaction, and found
little evidence that farnilies of schizophrenics were
different from control families in the degree of expression
of positive or negative emotion, relative dominance of one
parent by another, or conflict'"
similar difficulties plague researchers of the
psychosocial determinants of delinquency. sociologists have
stressed poverty, unemployment, minimal family controls over
children and reduced social controls as contributing
factors" An abundance of deviant role models, due to high
4
crime rates, has also been implicated. cloward and ohlin
(r960) stress the discrepancy between cultural aspirations
and chances for lower class children to achieve those
goals. McCord, McCord and ZoLa (f 950) emphasize t'he
importance of peers. The famíly is said to promote
delinquency, through such mechanisms as aggression-inducing
frustrations, reinforcement for antisocial behavior ?ttd a
lack of teaching of cognitive control. Johnson and Lobitz
(tgl+) found marital maladjustment significantly correlated
with children's deviancy, and Rutter (fgZf) researched the
importance of the nodelling of aggression in the marriage
relationship. The home environment of delinquents has also
been found to be chaotic and lacking in discipline (Burgess
& Conber, Lg76). As wittr schizophr.enia, possible causes and
effects are confounded-
Thus ¡nany difficulties arise from attempts to isolate
the effects of the above factors on schizophrenia and
delinquency. However, this thesis has been an attempt to
establish a measure for "entropy" which may contríbute
substantially to research involving family interaction
patt,erns which may lead to these disorders" These patterns
can be observed on an ongoing basis, and their effects
gauged.
Haley (tgø+) examined verbal dialogue between family
members and determined that repetitive interaction patterns
5
differentiated "disturbed" fami1ies from "normal" ones'
Faunce and Riskin (fgZO) developed Farnily Interaction Scales
whícþ measured various aspects of commuRication, sucþ as
ctarity and topic continuity. They found that
differentiated patterns of interaction emerged for the
various family groups studied, although a clear distinction
between families wittr delinquent members and families with
schizophrenic members r¡tas not made, thus limiting the useful
cont.ribution of the results-
Delinquents have been used as controls in studies of
schizophrenia (Stabeneau, Tupin, Vterner & Pollin, L964), as
they form part of what is thought of as a similar family
constellation" Both .constellations contain a child with a
behavior disorder that is frequently disruptive, costly and
public " Thus the shame and guilt produced by socially
stigmatízed illnesses is controlled. However, the
symptomatology of delinquency is outward-directed, whereas
ttrat of schizophrenia is usually inward-directed. The
extreme and sometimes opposite nature of the two disorders
has motivated much curiosity regarding the etiology of
each. As a result the two disorders are now being studied
together, each in its own right, in order to glean
ínformation on ttre way each is differentially promoted by
family int.eraction"
6
Present Evaluation Methods
In areas other than patterns of dialogue, evidence
bearing on the various effects of family int'eraction is
sketchy. only in the last three decades has a focus upon
the whole family as a system emerged (Bateson et al", L956;
Haley, Lg64r Stierlin, L973i Beavers, L977). However, Haley
(LgøA) feels that a compilation of individual members'
perceptions may or may not be representative of the
interaction patterns which most affect those members '
Perhaps this wariness of anything but direct observation of
family interaction is partially responsibte for a near-fact
evident in the literature on delinquency and schizophrenia:
In few areas of investigation of famity interaction relying
on techniques other than direct observation, is clinical
impression refined to the point of empirical validation' In
particular, a valid and relíable scale vtrhÍch could be given
to family members to measure aspects of interaction
pertinent to the continuation or fostering of delinquency
and schizophrenia has not been developed. Instruments
devised to assess family interaction must be used by trained
raters and often accomplish nothing more than an assessment
of inter-rater reliabilitlr (Lewis, Beavers, Gossett &
Phillips, Lg76i Faunce and Riskin, f970). Other problems
inherent in this approach include the biases introduced by
non-blind ratings; the prohibitive amount of time necessary
7
for the observation procedure; and the requirement of having
the entire family available for observation" Lastly, there
is little evidence of either reliability or validity for
almost alI of the instruments presently used"
The development of a valid and reliable scale to be
given to family members would elininate some of the above
difficulties, but there is a need for a superirnposition of
order on some of the above findings" In order to separate
fact and fiction, cause and effect, the psychometric
properties of such a scale must be firmly established. This
is the primary purpose of the present study.
The theory that has most clearly differentiated between
family interaction patterns which contribute to the
development of schizophrenia or delinquency has been that of
Stierlin (L973, L974) " This theory tras been further
elaborated by Beavers (tgll ' L9B2). They postulate two
distinguishable patterns of interaction to be present in the
families of delínquents and schizophrenics, these being the
centrifugal and centripetal interaction patterns
respectively, which will be discussed below in greater
detail. The concepts, centripetality and centrifugality, in
turn derive from General Systems Theory.
B
General SYStems Theory
GeneralsystemsTheoryradvancedtoagreatextent.by
von Bertalanffy (1969), involves several notions" To begin
wíth, all reality is regarded as a collection of systems, a
system being defined as a set of elements standing in
interaction. The idea of holism, that neither the system
nor its elements can be understood by examining parts in
isolation, is a second premise. von Bertalanffy ( rg0g)
advanced the notion of a steady state of balanced tension,
whích reflects the interrelationships and dynamism inherent
in life. A certain amount of tension was said to be
necessary with respect to system boundaries, which separate
the system from everything external to it' including
potentially useful resources. The classification of' systems
as open or closed, based on tension or permeability of these
r- -r ^- .c^'r 1^,-,^,:t €rnm > lrndrr rrf fhcorw which pOStulatedfxJUfl(ìdL l- cÞ, !e¿lvwçu
the effects of various degrees of this tension" If the
boundary hras too permeable, the system hras believed to lose
identity and integrity, due to infiltration of the external
environment" On the other hand, .if the boundary was
impermeable integration wìttr the outside world was believed
to be Lost. This would lead to a state of life-threatening
entropy, described as disorganization or chaos. The systen
would grow only with interchange of materials between itself
and
and
9
its surroundings, âllowing for increased differentiation
organization, ox negentropy"
Centrip etality and CentrifuqalitY
The concepts centripetality and centrifugality fít into
General Systens Theory in the following way: a system's
interactions are centrifugal to the extent that tension
holdíng its elements together is lacking" It is centripetal
to the extent that tension precludes a permeable boundary
and thus the interchange of material between the system and
other systems surrounding it. The terms "centrifugal" and
centripetal" have been applied to the realm of family
interactíon" They have been expanded, to include more of
what is known about families to which these two labets
possibly apply.
On the basis of clinical studies, Stierlin, Levi and
savard (rgzr) write that centripetal forces tend to delay
the adolescene's natural attempts at separation from his or
her family. They state that members of "centripetal"
families view their families as holding nore promise for
them and of meeting their needs for relationship than does
the outside world. Anything which extends past the confines
of the family has ninimal perceived impact. Thus
interaction with the environment decreases. These families
are often in social isolation, regardless of apparent
r0
organized activities" Reiss (L975 ) spoke of consensus
sensitivity in these families, which involves a constant
need to achieve overt emotional consensusr êfld often
precludes separation of family members from one another.
The presence of pseudomutual i ty, a f ami l iar condi't ion
st.udied by lrlynne, Ryckof f , Day and Hirsch ( 1955), is
paralleled by this state of affairs; in which true feelings
are thwarted and various problems such as a low level of
marítal happiness, poor sexual relations between parents and
a lack of contentment wíth the family on the part of
individual members provide proof of covert frustration and
hostilit.y"
In many of these families a range of perceptions and
expectations hetd by the parents seem expressly for the
purpose of binding the adolescent to the family" He comes
to view himself as too inept or sick to deveJ'op a separate
identity, and as self ish and cruel should he at'ternpt to do
so" Parents provide rewards for remainíng tied to the
family, and punishment in the form of guilt and lowered self
esteem when this is not done.
For families whose interaction follows the centrifugal
pattern, primary sources of gratification are found outside
the family. The idea that the outer environment holds more
promise than the family environment is held by all, as well
as the idea that most if not all problems can be solved by
1t
moving from the family orbit. Open, often hostile
argumentativeness is usually evident"
Characteristic of parents from centrífugal families is
the abitity to emotionally wíthdraw from an adolescent after
emotionally charged arguments. Although this and other
qualities of these family members seem extreme, Stierlin et
al" (fgZf) point out that centripetal and centrifugal family
patterns fall along a continuum, with most families
experiencing the "symptoms" to a degree" Simple observation
is often insufficient to determine a famíly's pattern of
interaction. Stiertin et aI" (fgZf) describe a family whose
centripetality upon closer inspection seemed transient and
weak, and one whose seemingly centrifugal tendencies
concealed centripetal elements"
Relation t.o Schizophrenia and Delinquency
These two family climates have frequently been linked to
schízophrenia and de1ínquency, for example by Stabeneau,
Tupin, werner and PoIlin ( fO6S ¡ . In the families of
schizophrenícs these authors found overcontrolled affect
discrepant with speech content and behavior. Controlled,
unnatural sweetness was often evident, âs well as a rigid
family organization and distortion of role expectancy and
differentiation" As a result of withdrawal, isolation, and
prevention of assertiveness, some members were left with
L2
littte power. Guilt seemed inst.illed in children, whom
parents thought of aS extensions of themselves and
controlled by expressíng hurt and confusion. Conformity to
outside standards was noticeable.
In tlre families of delinquents, êffect was
'undercontrolled, intense and artificial. conflict occurred
frequently, often between parents, and role confusion bJas
prominent, with role 'comp'etition and wÍthdral¡ral pervading.
Members seemed self-centered and manipulative.
In both family "types", overt and covert affect
differed. In schizophrenics' families affect was seemingly
pathetic and shallow, concealing anxiety and hostirity. In
the families of delinquents it appeared counterfeit, but
depressiveness and self punitiveness 'rr¡ere evident to the
skilled observer. In delinquents' families each member
spoke more and pauses þrere fewer than in t'he families of
schizophrenics. However, the schizophrenic patient
ínterrupted least and was the least effective in
transmitting thoughts when compared to others in the
family. The greater effect of the family's interaction on
the patient was evident" All behavior patterns seemed most
intense in the triad comprised of the parents and the
patient, whereas interaction between parents and control
siblings seemed much more like that between "normal" family
members.
13
It should be noted that the use of the label
n del inquentn creates complexities in research involving
"delinquents'n families. Specifically, delinquency is not
an easily defined disorder or syndrone. There is no one
constetlaLion of behavior which once identified justifies
the use of the term odelinquent.n Labeling depends on many
variables, including whether one is r caughtn for
transgressing and, if caught, whether or not one can procure
a nfavorablen judgement, from others. These variables in
turn are influenced by other factors. Thus any research
involving the accepLance of individuals as ndelinquentsn is
only as'credible as is the labeling process which precedes
ir.To understand how families of delinquents and
schizophrenics may depart from normality, a composite
picture of the nhealthyn family is necessary. Lewis et a1.
(1976) state thatr oh the basis of case studies, a maximally
viable social system perhaps possesses flexible
organization, complex structural relationships observable as
it interacts, intrasystem determinism hlith respect Lo
componentsr behavior and a continual flow of experience and
information into and out of the system.
The flow of information and experience important for
psychological health involves individuation, i.e. becoming
responsible for self expression and being receptive to that
L4
of others. centripetal and centrifugal family members may
fail to profit, from this, for different reasons. Those from
centripetal fanilies .are often too bound to others in their
family to develop autonomy of thought and feeling" They are
prevented from receiving others' experiential input, âs it
would impinge on their dependence on the family. Ttrose from
cent.rifugal families are free to develop their own sense of
self, altlrough tainted by parents' negative attributions"
However, be'cauSe they need to be accepted and valued, they
may often fail to profit from others' experience in their
blind attemPts at self assertion.
The Notion of Entropy
Thus the interchange of "material" necessary to a
healthy system is perhaps lacking in centripetal and
centrifugal families. Much of this material can be thought
of as energy in the form of information and affect'
available to,enhance organization. General Systerns Theory
contains the notion thae a healthy life is made possible
through negative entropy or negentropy (as opposed to
entropy), the use of energy to develop structure and fight
the inexorable downward pulI found in closed systems" This
puII seems t'o best represent the common malady of
dysfunctional centripetal and centrifugal fanilies; other
than the common boundary problems, which are related to a
15
lack of and an overabundance of permeabirity in these two
family types respectively. It may also represent an
important difference between families which appear to have
centripetal or centrifugal characteristics but experience no
difficulties, and "dysfunctional" families, which do
experience diff iculties.
Lewis et al. (L976) note that the more a system
possesses energy to maintain structure and flexibility, the
more negentropy and the less chaos or rigidity is
experienced" Thus both centripetal and centrifugal families
may experience entropy, which perhaps manifests itself more
as chaos in the former and as rigidity in the latter. For
example, entropicr cêrltripetal family systems may ' be
dysfunctional in the sense of being feedback-governed and
error-activated. Rigidit.y may be prornoted by adherence to a
negative feedback loop and disregard of growth-producing
i nformat ion "
Put more simply, the existing energy available to the
family system may be used to prevent change. In the extreme
form, this requires much vigilance of family members with
respect to signs of change in the family system" Evidence
of variation in family proceeditrgs, for example a slight
decrease in the amount of time spent together on weekends,
may be acted upon such that. time and effort (energy) will be
spent to reverse the change; rather than to determine the
t6
reasons behind the change, ot the possible benefits of it,
sucl.t as an increase in an adolescent' s competence in
interacting with peers. Morphostasis or stagnation of
family structure thus occurs, because family members do not
respond sufficiently to signs of the structure's need to
adapt. to members' changing needs.
Incontrast,entropic,centrifugalfamilysystems
experience morphostasis for different reasons. Due to
struggles for self assert,ion, coercion and manipulation, a
positive feedback loop may not occur. It is difficult if
not impossible to attend to cues signalling which actions
are conducive to the emotional health of famity members and
thus the positive growth of the family system itself, if too
much attention is placed on avoiding dominance by others'
The probability of these actions is thus not increased, and
morphogenesis, growth of family structure, rnay not occur'
For example, trow can a son realize that his father would
"loosen" the rules given some signs of maturity from him'
when the son believes that showing these signs would lessen
his independence?
Entr in the Literature
One
a lack
system.
way that entropy Ïras been described in studies is as
of input of information into the disturbed family
tennard and Bernstein (1969) and Riskin and Faunce
L7
( rgzo) found mind-reading statements like "You're not
hungry" were prevalent ín most disturbed families- Trhe
families studied by Riskin and Faunce ( r970) included a
multi-problem group containing both schizophrenic and acting
out children, âs well as a constricted group containing
children tabeled as neurotic and a group which experienced
"officiat" child-labeted problems, . êither delinquency or
underachievement. Entropy was not a sought-after,
operationalized variable ín this study. However, indirect
identifícation of at, least one possible form of entropy in
disturbed families was Provided.
In such families, individuals may not make use of
information as to how other family members are thinking and
feeling, even with respect to important matters " Bruch
(Lg62) found mothers of infants who later became
schizophrenic were insensitive to cues coming from these
children, tor example t,hose involving hunger" Lewis et aI.
(tglø) found many centripetal fanilies had external
scapegoats, such that people outside the family l¡rere scorned
for expressing emotions that "good" families keep hidden,
including joy, anger and sorrow. Thus not only did members
fail to learn from eacþ other, but outside information was
ignored as well. This entropic lack of informational input
into the disturbed family system seems so prevalent that the
necessity for assessing it along with centripetality and
IB
centrif ugality is evident. The studies below make t'he
importance of this assessment even clearer"
tewis et a1. (L976) found that the centrifugal families
in their case study used blame to avoid personal
responsibility" For example, statements in which someone
else r¡ras blamed for the speaker's actions were common. Thus
centrifugal families demonstrated a form of entropy, failing
to extract information and learn from their immediate
environment. Even midrange centrifugal families, which
possessed many of the same characteristics as centrifugal
families but to a lesser degree, nê9lected to make use of
much information available to the family system. For
example, since no-one in these families was seen as
virtuous, attempts to be good or competent urere seen as
fraudulent and were derided or ignored. Thus, in some areas
little information hJas used by the system" This, in turn,
seemed to perpetuate rigid beliefs and attitudes harmful to
family members
In contrast, I{ishler and Waxler (1968) found t'he highest
rate of acknowledgment of members' communications in
"hea1thy" families" Information was introduced into the
existing system and utilized in a manner which promoted
negentropy by adding to the existing resource base. This
then contributed to inforrned decision*making"
19
Beavers (Lg77 ) states that there is a degree of
centripet,ality or centrifugality in alI families, but that
it is within the confines of an entropic family system that
centrifugality or centripetality can lead to dysfunction of
its members. He states that the dramatic expression of need
for self definition found in schizophrenic adolescents is
not surprising, given this entropy" He writes that the most
entropic "style" may occur in centripetal families, when
communication is so obscure that power is neither clai¡ned by
one nor shared by alI. The "y"a"* is So intolerant of
directness that members are punished for uniqueness.
Differentiation of members is nrinimized, because
separateness is terrifying" However, it is evident from
some of the above literat.ure that there are many entropic
"styles", present to a greater or lesser degree in many
forms of dysfunctional families"
Models of Entropv
Unfortunately, the research surrounding entropy in the
famíly systems literature has been sparse. In fact, until
relatively recently ( Speer , L9'7O t Wertheim, L973, L975)
theory did 'not stress the importance of the negentropic
potential for change but rather focused on the tendency of
the family to maintain its status quo. fhe abo.ve theorists
emphasized that it is both stability and change that
20
distinguish "healthy" families and couples from
dysfunctional ones" Trhe Beavers Systerns Model (Beavers &
Voeller, 1983) focuses on the concept of'systemic expansion
rather than systemic change, oñ a continuum from entropy
(system death) to negentropy (system growth)" The recently
constructed circomplex Model (olson, 1983 ) , also using
principles derived from General Systems Theory, stresses the
concept of systen change; oï1 the continuum from morphostasis
(no change) to morphogenesis (constant change)' Chânge is
viewed by this model as a curvilinear dimension with respect
to irealth of the familY sYstem"
From this perspective, negentropy is a balance between
too little change ( teading to rigidity) and too much change
(leading to chaos). Entropy is systen stagnation or death'
resulting from either too litt1e or too much change over
- E .L i -^ ^
1 sl^^rr^rh ÊrÂ5rrârê :n/l \/nal 'l ofvarylng perlocl5 QI trllte. Ar'L¡¡vqYr¡ Desvv&s
(1983) view adaptability as a linear dimension, with more
adaptability associat.ed with greater family system health,
this conception could naively lead to the view that the
centrifugal family (witfr its lack of or very permeable
boundaries) is the least entropic of all famities. Given
this, the need for an empirically validated, reliable scale
to measure entropy becomes apparent. such a scale could be
used to help define one dimension of family health and could
also be of benefit in helping define dysfunctional fainilies,
2T
regardless of their place on the centripetal-centrifugal
cont inuum "
The reader may have noted a parallel between anomie and
entropy, in that both involve an inability to make use of
information in a way beneficial to a system, and both can
lead to chaos. Anomie has been defined in various ways, one
consistent theme being what has been termed a collapse of
values and norms. It has been generally agreed (Pope, L976)
that EmiIe Durkheim's concept of anomie describes conditions
of social deregulation and disintegration. Whether these
conditions are acute or chronic, according to the theory
surrounding anomie the entire social system is affected when
components or sub-systems of a social system fail to
function "appropriately" " In fact , inability of tl.e
subsystems to follow the system's norms reveals pathology of
the system as a whoie. Simiiariy, if entropl" did not
prevaÍ1 subsystems could perhaps adapt to the changing ínner
and outer environment, furthering elaboration and
differentiation. Thus it is the interaction of the whole
social and/or family"system which is problematic and should
be examined, including its adaptive use of information. It
is insufficient to focus only on an individual in that
systen, fot example the schizophrenic or delinquent
adolescent "
22
Need For an Entropy Scale
With the above considerations in mind, âs well as the
compelling evidence from case studies that farnilies of both
delinquents and schizophrenics may be plagued by an entropic
existence, it Seems unfortunate that an empirically
validated, reliable scale to identify forms of entropy in
these and other "types" of families has yet to be
developed. A scale of this sort would allow clinicians to
aSSesS interaction patterns without the time-consuming
training of observers and checking of inter-rater
reliability, or the possibility of experimenter bias through
non-blind ratings. It might eliminate the need to observe a
family for hours or the necessity of having the entire
famíly present in ordet to make an appropriate diagnosis.
Many of these difficulties are inherent in instruments
presently used to assess famify interaction patierns, and
little is known about eittrer their validity or temporal
stabilíty" The precise scoring criteria used also
frequently remains a mystery, so studies cannot be
systematically replicated by others, especially those in
other settings. A questionnaire, however, would allow one
family member to complete the items. Alternatively, alI
family members could give their perceptions of family
interaction, uninfluenced by the reactions of other members,
and both similarities and differences in perception could
23
represent valuable information to the clinician" On the
basis of results more objective and replicable diagnost'ic
Systems might be developed, as well as treatment strategies
to help families overcome problems in interaction" For
example, entropíc, centripetal families could be þelped to
identify and .discard old norms of rigid adherence to rules
of conduct which no one benefits from" fherapy could focus
on enabling them to develop flexibte norms, based on
êverchanging needs and feelings, perhaps promoting
negentropy.
An entropy scale thus could make an important
contribution. Such a scale would allow clinicians to place
families on a continuum not only with respecè to the degree
of openness or closedness of boundaries between them and the
environment but also with respect to the degree to which
-LÎ ^ !^ t: CC ¿'i -T^ ^-,1 a] =laar:{-a {-a môâ+Eney d'Ig dIJ¿e L(J (lIl-!ersrlL¿qus ql¡q s¿qvv¿gss
changÍng needs of farnily members. More specifically, the
ability of family members to develop and maintain individual
identities (dífferentiation) while naintaining some form of
organization which prevents chaos or breakdown of the family
system might be identified"
Such a scale could also be used to check effectiveness
of clinical intervention, by providing pre and
post-treatment comparisons. After treatment the members of
a family could again complete the questionnaire, to assess
24
progress. Significant differences in the amount of entropy,
centripetality, or centrifugality may signal an improvement
in functioning, which can be examined periodically over time
and compared against other objective and subjective measures
of increased family "happiness'r. For example, a decrease in
centrifugality and entropy scores on the questionnaire may
be associated blith the gradual reduction of delinquent
behaviôr.
Longitudinal predictive studies could also 'be made, which
might provide a means of assessing the diagnostic potential
of the scale. SpecificaIIy, adolescents from farnilies whose
members have completed the scale could be followed over
time" It could be determined, for example, whether
signifícantly more adolescents from families operationally
defined as entropic and centripetal (on the basis of the two
scales invotved) oeveiop schizophrenia iira¡: do adol-escents
from families diagnosed only as centripetal, as entropic but
not centripetal, or as "normal". These are simply some of
the more obvious uses to which such a scale might be put"
Ie þJas the aim of this thesís to develop such a scale.
Cornposition of the EntroPY Scale
The Entropy Scale ( See
basis of Stierlin's theorY
with entropy as it relates
A) was developed on the
study research, dealing
Appendix
and case
to family system dYsfunction.
25
The entropy items are underlined in Appendix A" Items from
thescale,whichiscombinedwithandembeddedinan
instrument which also measures centripetality and
centrifugality, deal with various manifestations of entropy
as it has been conieptualized by him"
The respondent'S perceptions of clarity and expression
of ¡is/her family members' t¡oughts, feetings, beliefs and
attitudes are examined, ãS well as his/trer perceptions of
rules, "power" stfucture and rOIeS. PerceptiOns of the
general degree of calmness, unrest or chaos represent
additional aspects of the respondent's farnily tife which t1"
explored. As well, the degree of rigidity in terms of
refusal to change plans, the amount of family togetherness,
the general amount of disagreement and orderliness, family
members' concern for each other, and the overall variety of
family routines are examined" Together, the items represeni
an attempt to assess the respondent's perceptions of his/her
family's ability to make use of information available to the
family system, both from within and from hrithout.
Examples of particular items may help to illustrate
this. Thus itçm number 35 reads: "We always stick to our
plans nO matter what", while it.em number 2L reads: "You can
tell what people in the famity are thinking", and item
number 57 reads: "People in the family consult others they
are making decisions about". Overall the scale attempts to
26
assess the degree, causes and effects of entropy within a
family, as perceived by the respondent"
Trhe form of the Entropy scale is that of a Likert-type
7-point scale, in which a I represents "alh/ays", a 4
fepfeSents "S6metimeS" and. a 7 COffeSpOnds tO "never". FOr
determining response set, some of the items in the scale are
reversed, such that for some itens the more positively an
índividual responds the higher his entropy score will be
whereas fox others the reverse is true" The rationale
behind the use of a 7-point scale concerns the generally
monotonic relationship between the number of steps in a
scale and its repeatability (Nunnally, f970). Reliability
of this typ'e increases very rapÍdty as the number of steps
is increased from two, but. it levels off at approximately
seven steps" Thus, including more than seven steps would
simply have served to complicate the scale and Possibly
confuse subjects. As wellr ârI odd number of steps permits
the inclusíon of a middle, "neutral" step, which may put
subjects at ease by permitting some indecision as they
respond to the scale's items. The scale is composed of 33
items. The scores derived from these entropy items may be
totalled to form an Entropy Scale score, based on perceived
entropy. Subjects' individual perceptions of the presence
of this variable in the family environment are reflected by
this Entropy Scale score, with scores for each item ranging
27
from I to 7. The Entropy scale score is obtained by a
linear combination, í.e. an addition of it'em scores in the
sca.le. The scores of the few items which have reversed
scoring due to their wording hrere reversed and then combined'
Aim of the StudY
The aim of the study was primarily to test the Entropy
ScaIe's psychometric properties. That is, trow reliably does
it measure entropy? How veridicat is the scale, "veridical"
here pertaining to the degree of correspondence between the
factor structure and the aspects of entropy detailed in the
Literature? Tfhe Entropy Scale items were combined with
those of two scales measuring centripetatity and
centrifugality, developed by Loff (1982). The three scales
were administered in one questionnaire. However, they were
analyzed separately as in reality they are separate scales'
specifically, the study represents an attempt to assess
the psychometric properties of the Entropy Scale by:
1" Examining summary statistics, including the means
and standard deviations of the Entropy scale and individual
entropy items.
2. Assessing the content/face validit.y of the scale"
3. Assessing temporal stabírity over a period of time,
i "e" test-retest reliabititY"
4" Assessing internal consistency"
2B
5.Performíngafactoranalysis,todeterminew}rethera
coherent factor structure exist,s which corresponds to some
of the components of entropy documented in the literature"
6" Assessing the relationship of the Entropy Scale to
the socíaI DesirabÍlity subscale of the Personality Research
Form AA (Jackson, L967) "
29
Chapter 2
Method
Subj ects
The study utilized 28L subjects. Forty seven of these
subjects formed an initial group, which completed the
questionnaire twice in order to supply test-retest
relíability data" Trhe remaining 234 subjects were added to
the group of 47 subjects, in order that the total number of
subjects would be sufficient for the factor analyses' The
entire subject group provided data for att but the
test*retest. analyses" AII subjects were selected through
the university of Manitoba Department of Psychotogy subject
pool, and ranged in age from l8 to 20 years' It should be
mentioned that the 18 to 20 year age range was chosen
because it represents an important transition period in the
lifespan of many indíviduals, i.e. it ís the time when many
separate from their families. consequently it is a good age
at wtrich to examine Stierlin's theories of centripetality
lnd centrifugalitY"
subjects were given a packet which consisted of a short
explanation of the purpose of the study, the Entropy scale
( combined wittr the centrifugality and centripetality
Scales), the Social Desirability Scale of Jackson's
personality Research Form AA (L967), and instructions
relati ng
Appendix
feedback
to the comPletion of the
A) " Subjects received either
íf they so desired.
30
questionnaire ( see
group or individual
Tnstruments
The administered questionnaire, consisting of the
Entropy, centripetality and centrifugality scales, is
composed of 82 items, most of which are rated on Likert-Èype
7-point scales" Randomly interspersed in this instrument
are:
l" TTre Entropy scale, consisting of thirty three items
simply worded in order to promote ease of readabirity
f or sub jects. This scale was designed t'o assess f ami ly
members' perceptions of the amount of entropy existing
in the home environment. Examples of entropy items
ínclude such items as: "People in my family express
their anger openly" (item number 74), and "I can tel1
what makes people in my family angry" ( item number 72) "
2" The Centripetal scaler cofisisting of 20 items
designed to measure the degree of perceived
centripetality in the respondent's home environment.
3. Ttre Centrifugal Scale, consisting of 20 items
designed to measure the degree of perceived
centrifugality in the respondent's home environment-
4. A drug control assessment item (item number 69).
31
5" Three items which attempt to specify who in the
family is being predominantly referred to as the subject
responds to the Ítems (item numbers '75, 76 and 77) "
6" One item which asks who has the most power in the
household ( item number 78).
7 " Four items which attempt to ascertain the degree to
whieh each subject perceives that one or more of his/her
behavíors could be considered as " schizophrenic" or
',delinquent',, or could be perceived by others as being
'o schizophrenic" or "delinquent" ( item numbers 79, 80,
81, and 82) "
B" The Social Desirability Scale from Jackson's
Personality Research Form (PRF) AA (Lg67), a scale with
20 true-false items, which was also administered to all
subjects as part of the total packet
Procedure
Each
above "
and/or
address
subject was provided wittt the packet
Those subjects who wished to receive
described
indi vidual
group results þJere encouraged to leave their name and
with the experimenters. As wett, a code name was
written on each individual's questionnaire rather than the
indi vidual ' s name, in order to make it irnpossible to
directly pair subjects with completed questionnaires. In
this way an attempt was made to assure subjects about
32
conf identiality by making them aware t,hat this procedure was
goíng to be followed and that there could be no stigma
attached to any responses they made" This hopefully
encouraged them to respond openly and honestly" Completed
questionnaires were returned to the experimenters"
One month following the administration of the
questionnaires to the group of 47 individuals an identical
procedure was carried out with these individuals again, in
order to assess the temporal stability of the Entropy
Sca1e. The remaining 234 individuals completed the
questionnaire only once. These data were pooled with data
from the first administration of identical materials to the
47 members in the "reliability group"" The pooled data on
the total group of 2BL subjects were then analyzed.
Anal-yses
. The purpose of the analyses was to investigate the
psychometric parameters of the Entropy Scale. fhis r¡ras
accomplished by:
I " Examining summary statistics, including the means
and standard deviations of the Entropy Scale and the
individual EntroPY items.
2" Assessing the content validity of the sca1e"
3" Assessing temporal stability over a period of time,
i"e" test-retest reliabilitY"
33
4" Assessing internal consistency.
5" Performing a factor analysis, to determine whether a
coherent faetor structure existed which corresponded to
some of the components of Entropy documented in the
I i terature "
6" Assessing the relationship of the Entrop)¡ scale to
the Social Desirability Scale of the Personality
Research Form AÀ (Jackson, L967) "
Summary Statistics
Means and standard deviations of the Entropy Scale and
indÍvidual entropy items were calculated for each of the tr¡to
administrations of the scale. The mean sf the Entropy Scale
was calculated as an appropriate measure of central tendency
for the scores, in order to summarize data. Examination of
the histograms, to be descríbed below, determined that the
distributions of most items were approximately normal and
all distributions were unimodal. Hence the mean, median and
mode tended to coincide in these distributions and all brere
good representatives of the central tendency for the data"
It can be shown that if a large number of dat.a sets from an
identical source are taken and all three measures of central
tendency are Calculated for each set, the means will show
less variation than will the medians and modes. Tttus the
mean is the most stable of these measures (Huntsberger and
34
Billingsley, L977) " The standard deviation of the scale was
calculated in order to summarize the spread of the Entropy
Scale scores
Means and standard deviations of individual entropy
items were calculated to províde information on item
consistency. The standard deviations of each item were
examined to determine whettrer the ctrange in an item's score
between the first and second administratiori was greater than
one standard error of estimate, this amount being the
approximate degree of error expected when first
admínistration scores were used to predict second
administration scores. If a change of this magnitude was
found the item was considered to be too subject to change
over time to be reliable (Glass ana Stanley, 1970)" Items
in this category htere not eliminated from the scale but were
E-- -: Ll ^ ^1 .i -.i Fãl-j ^É i - €rrl-rrra
slngleq 9ut f(Jt !,rJÞsrurc Etr!Ir¡r¡qq¿vr¡ ¿¡¡
administratíons "
Frequencies in the form of histograms were also
calculated for each entropy ítem" Histograms were used as
they allow one to discern whether item distributions are
normal or skewed. If distributions were normal and/or were
skewed in the same direction approximately to the same
degree, correlations large enough to allow for a
continuation of the analyses would be possible. Those itens
whose distributions were skewed in the opposite direction
35
from that of most other item distributions would perhaps not
correlate highly with other items, and this would perhaps
lead to deceivingly low factor loadings for the former items.
Content Validity
Content validity was assessed by obtaining judgments
from various members of the Psychology Department who were
familiar with Stíerlin and Beavers's theory Pnd concepts of
Entropy. They were asked to determine whether the scale
ieems could be considered to accurately matctr descriptions
of the concept as given by Stierlin and Beavers. These
individuals included a University of l4anitoba Psychology
professor wlro had studied the literature in the Fanily
Systems area for the last several years ( including that of
Beavers and Stierlin); a graduate student in Psychology who
had spent a number of years stu<ìying this iiteraiure as LÌre
focus of her graduate thesis; and another Psychology
graduate student who had completed an undergraduate thesis
involved in studying the relationship between the Ent,ropy,
Centripetal and Centrifugal Scales and the Family
Environment Scale (Moos and Moos, 1981).
Temporal Stability
Temporal stabilíty of the Entropy Scale was determined
using the Pearson Product-moment correlation coefficient, to
36
determine the degree of relationship between the scores
obtained upon the first completion of the questionnaires and
those obtained from the same subjects one month later when
the same questionnaires were again completed.
Internal Consistency
Homogeneity or internal consistency of the Entropy Scale
performing an item
The rationale forwas examined in the present studY bY
analysis using item-whole correlations.
the analysis, as stated by Nunnall1z (L967) ' hras 3
When items correlate predominantly positively withone another, those with the highest averagecorrelations are the best items. Since the averagecorrelations of items with one another are highlyrelated to the correlations of items with totalscores, the it,ems that correlate most highly withtotal scores are the best items. Compared withítems with relatively low correlations with totalscores, those which have higher correlations withtotal scores have more variance relating to thecommon factor among the items, and they add nore tothe test reliability (p. 215).
To test for homogeneity, the correlation of each item with
the total scale score less the itern score itself was
obtained in order to rank these correlations from highest to
lowest. An r of "7O or above was considered adequate, the
rationale being that approximately half of the variance of
these items would thus relate to a common factor (NunnaLly,
1e67 ) "
37
Cronbach's A}pha was used as the estimate of internal
consistency. This measure is generally used more than any
other internal consistency estimate (Nunnally, 1970). The
SPSS-X, a stat.istical package for t,he Social sciences, was
utilized for this analysis. fhe package provided means and
standard deviations, ôS well as the Cronbach's Alpha values,
wtrich would obt,ain f or the Entropy Scale gi ven the
elimination of each item" Al1 of this information h¡as
cumulatively useful in assessiñg the effects of eliminating
a particular iten upon the scale's internal consistency"
Factor Anal ses
Factor analyses were performed using both an orthogonal
and an oblique rotation, in order to examine the underlying
facltor structure of the Entropy Scale and to determine which
method was most appropriate. The factor analyses were
performed using principle factor solutions, as these
extraction procedures usually account for the most variance
with the fewest factors and yield a correlation matrix with
the greatest accuracy for a given number of factors
(Gorsuch, L974) . For communality estimates squared multiple
correlations were utilized. A PrincÍple Axis Analysis using
squared rnultiple correlations was appropriate because the
factor analyses were performed in order to explore the
factor structure of the concept (entropy). No information
3B
as to the breakdown of the factors, i.e. the combination of
elements in these factors¡ wâs given in the literature.
Gorsuch ( f983 ) writes that factor analyses of this type
justify the use of the Principle Axis Analysis.
. The criterion for the number of factors to be retained
requi red t,he calculat ion of eigenvalues . Eigenvalues are
measures of the relative importance of any function"
Specifically, they are the characteristic roots of the
correlation matrices upon which factor analyses are based,
and represent the amount of variance accounted for by each
factor" A "scree test" (cattell, L966; Cattell and Jaspers,
Lg67) was performed, and this test provided the basis for
determination of the number of factors to retain by the
following procedure. Scree plots were examined, and the
pÕint at which the curves of these plots leveled of f 'rras
l^! ^ø-.: -^l ml^-i ^ {.L^ ^ai n*- r.r'laara +laa erlmd nf cõllîrôc aÍLlgLgl¡llJ"tlgu. tl¡¿Þ wqÞ s¡¡s À/v!¡¡L w¡¡ç¿s
the loadings approached zero, and hence where more
variability due to error or specific factors was represented
by the loadings than was common variance" Factors whose
eigenvalues were found to be at or above this point on the
curve were retained for rotation.
A Varimax rotation was f irst 'attemPted, in order to
determine whether an orthogonal solution was appropriate for
the data" The Varimax procedure distríbutes the common
variance among the factors more evenly than do other
39
procedures, and for this reason it. is often the preferred
procedure. In the present study this procedure yielded
facLors which did not possess simple structure, i "e.
numerous large Ioadings were found on many factors with
respect bo several variables" Thus, it was assumed that
relationships among fact,ors existed" An oblique solution
was therefore considered more appropriate because oblique
solutions account for Some degree of relationship between
factor s "
Various Delta ( 6 I values between +1 and -I were
experimented with for the Oblimin rotat.ions, in order to
obtain simple st.ructure. These different Delta values
varied the angle between the reference axes with respect to
obliqueness in order t,o determine the best angle for these
axes. The best angle would be the one that' would result in
simple structure" Simple strucLure was considered to be
found when the pattern matr íX, and in particular a
simplified algebraic expression containing the Delta value
being experimented with, was found (Nie, Hul1, Jenkins,
Steinbrenner and Bent, 1975) . Thus, if the angle of the
reference axes !,ras relatively more or Less oblique, Lhe
fact.ors being investigated !ì¡ere relatively more or less
related"
40
The following criteria, found in Thurstone (L947, P"
335 ) , were used to determine when a simple solution had been
found (Note that a zero loading was considered here to
include values which approached zero):
',1. Each factor should have at least one zetoloading "2 " gaãh factor should have a set of linearlyindependent variables whose factor loadings arezeto.3.Foreverypairoffactors,thereshouldbeseveral variãUtãs whose loadings are zero for onefactor but not for the other.4" For every pair of factors, a large proportionof the vari.Lfãs should have zeÍo loadings on bothfactors whenever more than about four factors areextracted.5. For every pair of factors, there should only bea small number of varíables with nonzero loadingson both. "
Relationship of the Entropy Scale to the Social Des i rabi 1i t.y
Scale of the Personality Research Form AA (Jackson, 1967)
The social Desirability scale from Jackson's PersonalÍty
Research Form (PRF.) AA (L961 ) was administered to all
subjects as part of the total packet" For all subjects
scores obtained f rom the PRF lvere correlated wi th thei r
total Entropy ScaIe scores using a Pearson Product-Moment
correlation coefficient. As well, after items loading
highly on a particular factor were placed together under the
f actor, the f actor groupings of ttrese items I¡Jere correlated
with each Entropy Scale item. Again, a Pearson r was used.
comparisons between these correlations determined whether
Entropy Scale items correlated more highly with their
4L
respective factors or with socíal desirability. If an item
correlated more highly with its factor grouping, its
variability was considered to be due more to the factor's
content than tO "social desirability". Thus it' was assumed
that the responses to that item generally were more
influenced by respondents' actual perceptions of their home
environments than by "social desirabifity bias"" If an item
correlated more higtrly with the PRF scores its variability
was considered to be more attributable to " social
desirability" than to the content of the factor "
consequently the reverse of the above assumption was to be
made. Lastlyr means, standard deviations and frequencies
urere calculated for the qRF. scores. Jackson's normative PRF
data (tgøl ) were compared to these figures in order to
assess whether the present study's sample was comparable to
the average population with respect to the attempt to appear
socially desirable. If the sample proved to be comparable
in thi s respect " soc ial desi rabi lity bias " would .to! be
expected to unduly influence the responses to the Entropy
Scaler ëIS tþere is no reason to assume that subjects
responded to entropy iterns differently with respect to
social desirability bias than they responded to the items
from Jackson's scale"
42
Chapter 3
Results
Summary Statistics
All of the analyses, with the exception of the
test-retest reliability analyses, made use of the entire
group of 28L subjects. O.rfy th9 first administration scores
of the subgroup of 47 subjects who lvere tested twice with a
one month int,erval intervening were utilized for these
analyses. The means and standard deviations of the Entiopy
Sca1e, utilizíng this sub-sample of 47 individuals, were
calculated for each of the two testing periods" fhese
values for the first testing were 106 " 66 and f6 " 86
respectively. For the second testing the corresponding
values were f08.30 and 15 " 36 " As well, the mean and
standard deviation þtere calculated for the total sample of
2BL subjects who completed the questionnaire" These values
were 103.f6 and L7.52 respectively. A t-test perforrned on
the data determined that no significant difference exists
between the largest and smallest means above, t= "0175,
p > "25"Frequencies of each Entropy' item are depicted in
histograms in Appendix C. The histograms portray either a
normal, approximately normal or positively skewed
distribution for all but three items, these being item 39,
43
"People in my family seem to disagree in their opinions on
most things"î item 65, "My family is afraid to be seen as
different by others"; and item 68, "l^Iith respect to variety,
my family would be considered boring" " If item
distributions are aIl approximately normal or are at least
skewed in the .same direction, correlations based on these
items qre not generally attenuated. Therefore, attenuation
of correlations is not likely to be a problem in the present
study "
one item included in the packet administered to
subjects, a drug control question ( item number 69) , asked
subjects how frequently they take non-prescription drugs to
alter the way they think or feel. Responses on the EntroPy
scale of those subjects who take drugs "always" or
"usually'!, for instance, mêY be confused by drug effects.
ïn the total sample , L76 individuals replied "never" on this
item" only six individuals replied "always" and only six
replíed "usually" "
Results were obtained for the three items ( item numbers
75,76 and 77) which determined the degree to which, when
responding to the qúestionnaire, the subject referred to
his/her mother, father or whole family" Item number 75
reads: "In ans!,reríng the above questions I was thinking of
my mother"; item number 76 reads: "In answering the above
questions I was thinking of my father"; and item number 77
44
reads: "In answering the above questions I was thinking of
my whole f arni ly" . For the total sample ( N = 281 ) the
response breakdown in terms of how many subjects responded
to each of the possible answers. for these three items is
portrayed in Table I.
A very minute proportion of subjects (2 out of 2BI)
"never" referred to the entire family while responding to
the Ítems, and a very small proportion (fO of 2Bf) "rarely"
referred to the entire family. Sirnilarly, very few subjects
referred "always" to their mother or father when responding.
Table I
Re ses Based on ltems Usi "Mother t'
" Fat or T¡Ihole Fam l_ Y ast e Referrant
#7sNof
( ¡¿other )
Subjects#76
Nof( rather )
Subjects#77 (!'Ihole Fanily)
N of Sub'iectsResponse
alwaysusuallyf reqtientlysometimesoccasionallyrarelyneverdon't know
t9364393352925I
N 2AL
L43B4392432L"27_3
\J = 281
44B44774t9l0
2I
N 2AL
45
The responses to each of the three items were well
distributed for most subjects, with the bulk of them
occurring in the middle of the Likert type scale" It cannot
be said that a preponderance of individuals' questionnaires
give information only about one family member.
Item 78, which determined who in a subject's fanily has
the most pov,Jer, was included in order to gain inf ormation
about the perceived Power structure in the families of our
subjects. The results, based on a sample of 273 subjects,
are þortrayed Ín Table 2. These results indicate that most
subjects belong to "prototlpícal" families in which the
father is perceived as being the "most powerful". Mothers
were perceived as having the "most power" in approximately
PerceBtion of
Table 2
lrfho in Family Has Most Power
Item #78 The one with .the most(Describe your
power ln nyrelationship
family iswi th thi smy
person).
Response
fathermotherparentsbrothersísterfami lysel fno onedon't know
N of Subiects
160B7L4
52II30
N 273
46
half this number of families" Virtually no one responded to
this item with "I don't know", although a resPonse of "no
one" may reflect some uncertainty"
The last four items of the questionnaire ( ítem numbers
7g,80, 8r and 82) were also analyzed, ütilizing the entire
sample. Item number 79 reads: "sOme of my actions have
been considered by some to be schizophrenic", item number B0
reads "Some of my actions have been considered by some to be
delinquent", item numbe'r 81 reads: "Some of the actions I
trave engaged in I consider to be schizophreníc", and item
number 82 reads: "Some of the actions I have engaged in I
consider to be delinquent"" As can be seen from Table 3
very few subjects in this sample felt that some of their
actions could be considered "schizophrenic" actions by
others or themselves. Similarly, very few subjects felt
that some of their actions eould be considered as
"delinquent" actions"
Sub-iects' Evai-uations of, Some
Tabl-e 3
of fheir Actions bv Sel-f and Others
BY OTHERS BY SELF
Response
atr-ways
usual-J-y
f requentJ-ysomet imes
occasionai-i-yrarelyneverdon't know
Question #79" Schi zopTrrenic"N of Subiects
I0
7
l_9
2L
69
r595
N 2AL
Question #80t'DeIinquent "N of Subiects
II
T2
28
3B
88
106
7
ril 2AL
Questíon #8Inoschi zophrenic"N of Sub'iects
Io
I16
26
7l_
r47t2
Àï 2Br-
Quest.íon #82" De I i nquent "N of Sub-iects
I0
TO
2A
49
89
97
7
N 28L
+
48
Content ValiditY
The procedure used to assess content validit,y of the
Entropy Scale indicated that ttre items all accurately
describe the concept of entropy as described by Stierlin and
Beavers (Stierlin, Lg7L, L973, L974i Beavers, L977, 1983).
One hundred percent agreement was obtained between the
'oexpert" judges who took part in the content valídity
assessment with respect to which items to include ín the
Entropy Scale. These items were considered to be derived
from particular variables described as correlates of
,,entropy', in the "Family" literature " The variables
included degree of cohesion/togetherness and tranquility of
family members, upset and/or peace in the family
environment, openness of farnily members in terms of
expression of feelings and thoughts, rigidity in planningE--ia-- -^-r---.:!-- ' ^ñ: yrr'Ia ^l-*ì+rr in
lnvorvlng E,ng IilIltIIlr u(Jt¡I(JlrltrçJ, }/vwEI q¡¡u 4q4ç v!qr¡sJ
the family and evaluatíon of family activities in terms of
both social appearance and interpersonal enjoyment" The
factor analyses (fa¡les B to f5) described below indicate
that components of covariation. of the Entropy Scale are
consistent with the Family Systems literature dealing with
entropy" The factors will be described in detail below"
49
Temporal Stabilitv
The test-retest reliability analysis of the 33 entropy
items, performed on the questionnaires of the subgroup of 47
subjects, yielded the Pearson correlation coefficients
listed in Table 4. The table provides Pearson correlations
for each item, ês well as item means and standard deviations
for each administration of the scale" The results of the
procedure indicate that the Entropy Scale is reasonably
stable for the period examined. The Pearson r for the
seale, administered twice with one rnonth intervening, was
"4974, p ( .001. As well, 22 out of 33 items on the scale
have correlations which are significant at different leve1s
( í"e" the .05, "01 or "OO2 'levels), signifying that
subjects' responses to many items remained relati.vely
unchanged over the one-month period.
Examining the significance IeveL of test-retest
correlations for the items reveals that the preponderance of
items with relatively low test-retest correlations are found
near the end of the questionnaire. Reasons for this will be
discussed in the following chapter. However item number 26,
"Ttrings are peacefUl at home" and item number 39, "People in
my family seem to disagree in their opinions on most things"
appear before, or approximately ât, the halfway poínt in the
questionnaire. As well, item number 52, "Members of my
f ami ly don' t care about each other " , item number 54 , " In .my
50
f amily we care about, each other's f eelings n , it'em number 57 ,
nPeople in t.he famity consult others they are making
decisions aboutn and item number 6!, nIn my family we try to
keep Eime for each othern, are not near the end of the
questionnaire. These. six items were found to have
relatively Iow test-retest correlationsr âS did those near
the end of t,he scale, i. e. item number 65, nMy f amily is
afraid to be .seen as different by othersn, item number 68,
'with respect to var iety my family would be considered
boringn, iLem number 7:I, nMy family has a good time when
wer re t,oget,her o , item number 72, n I can tell What makes
people in my family angryn and item number 74, oPeople in my
family express their anger openlyn. In fact, item number 61
obtained a nonsignificant, ' negative correlation when
eorrelated with itself over a one-month period. Reasons for
the low and negative t.est-retest correlations found with the
early*occuring test items may be different from those
applying to the late-occuring items.
51
Table 4
-Retest Product-Moment Correlations,Entr Sca1e Test
Item # t
Means a Stan r Dev at ons
First Administration Second Administration
Mean S.D Mean S.D.
323433333333333434323223344444322
246I9
t0202L222326l9303233353B394547505254575861626365687L7274
" 61 52***.6113***"4443***" 6009***" 4400**5863***
" 5069***q'¿1 ?***
" 3930**.3389*"2339" 3539*"44LL**" 4984***" 6411***" 3463*"3L29*"o392" 3058*"4443**¿qrq***" 1569" 3r86"2L26" 6183***" OBI6"4657**" 5663***" 1957"o954" 2355.27L8"2L57
.0638"9787"1915"4255.0851" B5II.85r1"27 66.0000"L957"L702.9787"5652.239r"2826" 1087" 9565"5532"6170"9L49"6596"0213.9362"2979.4255"2340" 0000.1489"5957
1" 1r.13L.4669L.527LL"4707I " 47r9r " 84r3I " 0628r " r554I " 3I88L"3763L.4494L "32681.5585r " 6081L "6420r"5236L"6727I " 0593L "2947L.42691.3875r " 3i03r " 63391 " 6005L "5428r "402rr " 4891L "64l-6I " 9411r"63281. 398rL.4037I " 4303
3 "04263 "L2773 " 25534.L7023 " 40433 " 53193 " 65963 "29793 " 06383 "27663 "468L3 "7 6603 " 63833 " l9r53 "23404 "40434 " 08514 "04263"61703.085r3 " 6809z " f,Ivo2 "65963 "29792.93624.L4893 "72344 "59575 " 06384 "36L73.319r2.63833.0952
t. 1602L "3L24L.42L4I " 3BBrl. 5695L "5443L "273Lt.r7B0L "3254L.4552r.5583L "402LI " 3285L.296LL.4327I .43951.66591.3015L.2257L.48661.23551 a Â^4L.O+WZ1.3396I " 5311L .4656L "4443L.3465L.4395r"5238L "7L22L"5479r. t3l1r. 3400
"8298
" 8205
.0426" 8298
N.TE. ***p( "OO2, p3".or,*p4.05**
52
Internal ConsistencY
The internal consistency analyses were performed on lhe
responses of 273 rather than 281 subjects, because the dat'a
obtained from eight. subjects vÙas incompleLe, i. e.
insufficient for inclusion in the analyses. These analyses
resulted in. a Cronbach's Alptra measure of .8078. This value
may be considered to reflect a "goodn overall degree of
internal consistency (NunnalIy, I970 ) for the Entropy
Scale. Table 5 gives the it'em-totaI statistics for each
Ent,ropy Scale item. A few items had negative correlations
with t,he rest of t,he Ent,ropy scale, these being: it'em 10,
nïn my family it's clear who has the most powern; item 39,
npeople in my family seem to disagree in their opinions on
most t.hings'; item 63, nIn my home outsiders make social
rriei!-qtr. i!-êm 65, oMV family is afraid tO be Seen aSv ¿Ð¿us , 4ev'r' --, -'!
dif f erent by othersn i item 68, nwith respect t'o var.iety my
f amily would be cons ider ed bor ing n ; and it,em 7 4 , n PeopIe in
my family express their anger openly'. Item-total
correlations for the remaining items range from " 0706 to
"697 4.
53
I t,em
246
I9
10202L22232629303233353839454750525457586t626365687L7274
RELTABTTITY COEFFICIENTS
N OF CASES = 273.0
ALPHA = 0"8078
TABLE 5
Entropy Scale Item-Tot,al Stat,istics
Cor r ectedïtem-Total
Cor r eI at ion
.3036" 3036.4615.5681"4944
-.0895.4355"3322" 3830.5446" 6524" 0706"587 4
" 0969.5495" 2000.4845
* " 2983" 3866"2942.4102"JIOY"6r80.4446"3902.5839.4952
-"3141-.2533- "296r
"6974.3150
-.0323
AlphaIf ItemDeleted
.8031"7979.7 966"7 922.7946"8215"7 986.8023.7 999.7943" 7900
" 7900"7 922.8102.7 925.8068.7944"8251.7 997.8033.7 984.ÕUUY.7 899"797L"7 996"7 909.7935"827 2
.8290
.8259"7876.8026.8152
54
F ctor Anal ses
For the factor analyses Principle Axis Factor solutions
were util ized. The communality estimates consisted of
squared multiple correlations. This procedure produced the
eigenvalues shown in Figure L I plot,ted on a scree chart.
Examining the plot for the Ent,ropy scale reveals that. the
curve formed by the eigenvalues begins to level off at
f act.or 6 ¡ hence the retention of six f actors f or rotat,ion "
Símple structure was not obtained by the Varimax rotations,
as many items loaded on many fact,ors rather than only one.
Indications were that the factors were inter-reIat,ed. They
vüere not independent, and a Varimax rotation to obtain
simple Étructure vtas not a reasonable procedure. The
rotated Factor Matrix obtained with the Varimax rot,ation is
qhnwn ìn Annenriì¡¡ D^--rr :'- -- _-
Consequently Oblimin rotations viere attempted for
various degrees of obliqueness, the purpose of this being to
determine the angle between the two Èeference aXes which
would lead to ttre nbest solution" , i . e. t,hat' which most
satisfied the criteria of simple struct.ure. A solut.ion wit,h
a Delta value of 0.0 most satisfied these criteria. Of
course, beCause the solution v\tas not orthogonal a few items
vüere included in more than one factor due t,o their
relatively high loadings on t,hese factors. Specifically,
item number 26 was included in both Factor I and Factor II,
L 11 )0'o
a 1l'bç.t-'
se8'\
'Eeq' I
¡-s b'l
e
i.
,
ç1cl
1üAN19Í
Ql
z1q)9 hdailN3 ?n'L'd aJ s7n-7U^f\2s tA 'l glY't2tJE
L
s
b))
56
as was item number 30. Item number 7I \¡tas included in bot'h
Factor I and Factor VÏ.
A tot,aI of six factors resulted from Ehe Oblimin
rotation with DeIta value set at 0.0, this being t,he
accepted final solution. Appendix E protrays in a Pattern
Matrix the Entropy Scale factor loadings which were obEained
with this rot,ation. Table 6 relates the content and range
of loadings of each interpretable factor, and Tables 7 to 12
give the item composition of the factors. In t,he case of
most factors only items loading 2 .30 will be presented
here. It is primarily t,hese iEems which will be given
weight, in the int,erpretation of each f act.or r âs Gorsuch
notes that a lower boundary of .3 f or t,he def inition of a
nsalientn Variable may promote nmeaningfulnessn" The fairly
large sample size (N=281) of Lhe present sample may have
rendered even n uninterpretable" factors statisticaify
significant if computations to determine statistical
significance had been performed (Gofsuch' 1983 ' p. 210 ) .
Thus, Lhe simple criterion of meaningfulness ltas deemed to
be the best guideline in the inclusion,/exclusion of items
under each factor.
37
Table 6
Summary of Entropv Scale Factors
Range ofFactor Items Included f,oaainqs
,50,7 1,38,2,30 ,26
I 61, I58, 6
Content Area
"3225 to "7592 Togetherness andtranquílity offamily members
TT 33 ,23 ,9 ,26 ,32,39, 30
" 310f to "6544 Calmness of thefamity environment
rrr 74,6,72,2L
rv 29,35,65
v LA,22
"3053 to "7066 OPenness of familYmembers
"2647 to "754O Sponteneity offamily planning andconformi ty
"4854 to "5820 Power and ruleclar i ty
"2288 to " 6863 Social appearanceand interpersonalenjoyment
vr 68,7L,47
NOTE: Absolut.e values of factOr loadings are given"
5B
Item composition of factors.
Table 7
Factor I: Toqetherness and Tranquility
61 "7592
ï tem_t
B
50
7L
38
5B
62
30
26
.6268
"5844" 5r82
"4348" 3598
" 3585
"3327"3225
Loadinq Ïtem Content
In my family we try and keep time for eachother.We do things together as a familY.My horne is a shelter from the chaotic world"My family has a good time when we're together.People apologize after arguments in my family-Holidays are for the family to be together.I am much like other members of my family"Our family tries to keep things calm"Things are peaceful at home.
Factor I involves cohesion/togetherness
tranquility of family members. It is comprised of nine
with loadings ranging from .3225 to .7592.
and
i tems
Factor II involves uPset
environment" It is comPosed of
ranging from "3f01 to "6544"
and/or peace
seven items
59
in the family
with loadings
Table B
Factor II: "War & Peace" in the Family
Item# Loading
33
23
9
26
32
39
- "6544- " 5463
-.5428
-.4355- .4L36
"3877
Item Content
Argurnents are unending in'my family.My family is in an uproar.Going out is an escape from the disorder in myfami ly "
Things are peaceful at home.
Without rules my family is in chaos.People in ny famiJ-y seem to disagree in Èheiropinions on most things.Our farnily tries to keep things calm.30 -" 310r
Factor III relates to openness of family members in
terms of expression of feelings and thoughts, and includes
four items whose loadings range from .3053 to .7066.
60
Tab1e 9
Factor III: "Transparency" of Fa¡ni Iy Members
Item#
7 4 "7066
6 "4543" 331972
toadinq Item Content
People in my fanily express their angeropenly.People in my family are uP-frontI can tel1 what makes people in my familyangry.You can tell what people in the family arethinking.
2L " 3053
Factor IV involves rigidity in planníng involving the
family, as well as conformity. It has three items, with
loadings of "4812, .754O and "2647. Item number 65, "My
family is afraid to be seen as different by others", has
been included in the interpretation of this factor due t'o
the apparent theoretical importance of "conformity" in the
FamiIy Ent.ropy Iiterature, the lack of other items designed
to "tap" into this variable, and the fairly small numerical
difference between .2647, the item's factor loading, and the
"cut-off point" of "3"
61
Table I0
Factor IV: Ridiq i t.y and Conformity
ïtem# Loading
29
35"7 540
"48L2
It.em Content
My family hates toMy family alwaYsmatter what.My famity is afraid to be seen as differentby others.
change plans "
sticks to its P1ans no
65 "2647
Factor V, which deals
the family, is a doublet"
"4854 and "5820"
with por¡rer and rule
Its two factors have
clarity
loadi ngs
1n
of
Table 11
Factor V: ClaritY of Power and Rules
ïtem# Loading Item Content
10 - " 5820 In my family it's clear who has the mostpower.
22 -"4A54 fhe rules in the family are clear.
62
Factor VI concerns evaluation of family activitíes in
terms of social appearance and interpersonal enjoyment. It
has three items. These ítems range from "2288 to "6863 with
respect to factor loadings. Item number 7Lo "My family has
a gOod time when h¡erre together", and item number 47, "From
outside my family would look orderly", $rere included in the
interpretation of this factor due to the small numerical
difference between their respective factor loadings and the
"3 "cut-off poínt". As weÌI, their inclusion seems tO be
needed in order to interpret the f actor " Ttre f act'or i s
Revertheless difficult to interpret, and seems to overlap
with both Factor I and Factor IV" It has been included for
its possibte heulistic value, but may be deleted in the
future "
Table L2
Factor VI: Social,Appearance and Interpersonal Eniovment
ïtem# Loading
68 - " 6863
Item Content
I/ÀIíth respect to variety, my family would beconsidered boring.My family has a good time when we're together.From outside my famíly would look orderly"
7L
47"2683
* "2288
63
A factor correlation matrix was produced along with the
oblique solution factor extraction" Some of the six factors
extracted correlated quite substantially, as shown in Table
r3.
Table 13
FACTOR CORRELATION MATRTX
Factor I Factor 2 Factor 3 Factor 4 Factor 5 Factor 6
Factor I
Factor 2
Factor 3
Factor 4
Factor 5
Factor 6
r " 0000
- "4L29
"L544
" IèJ /IJ
" 0183
"2L67
1 " 0000
* " 0548
^^^ F
" UUðf,
- "L651
- "L526
1 " 0000
^^^^. v4+\J '*
- " 0302
" 0850
r r1^/lr^lI. \J\J\JV
- "L247
-.0239
I " 0000
.0896 1 " 0000
The finat statistics, ütilizing an oblique solution, are
found in Table L4" These statistics include the eigenvalue
of each of the six factors and the communalíty estimates of
each of the variables (Entropy Scale items), the latter of
these two values representing the total variance of a
variable accounted for by all the common factors.
64
In Table L4 the relative contribution of each fact,or can
be seen" For example, it can be seen that Factor I account's
f or 5 .7 times t,he amount of var iance among t.he i t,ems
accounted for by Factor rr. This can be verified by
examining the npercentage of variance" column.
Table 14
Item Communality and Relative ConLribution of Each Factor
Percentage Cumulative CommunalitTtem No. Factor Eiqenvalue of Vàr iance Percentage
246I9
102g2L222326293032333538394547505254575861626365687L7274
I23456
8.00551.40641 " 36401.0947
"7 866.6644
24 "34.34.13"32"42"0
24.328 .532 "736.038.440"4
.37 45
.4l-42" 4439"5447.5136" 3648.4053"2848.4385"57 24
" 6300
"547 5.4796"2357.6339.2544.4052.4005.2963"2698"32t4.5706" 6004"2692.2401.6189.42I2"6250.4770.5564.698s"5416.5330
65
Rel at ionsh ip of the Entropy Scale to the Social DesirabilitY
Scale of the PersonalitY Research Form AA (Jackson, 1967)
Jackson's Personality Research Form AA of L967 (PRF) was
administered to subjects to assess bias due to social
desirability. A Pearson å was calculat.ed to determine the
corretation between PRF scores and the mean Entropy scale
score for aIl subjects. This correlation was .4048, and u¡as
significant at t,he .001 1evel. Thus subjects in the study
received PRF scores which vÍere very comparable to those
obtained by Jackson I s st,anda.rdization sample, the overall
mean and standard deviat.ion of t'he present sample being
15.880 and 2.962 respectively. These results seem t'o imply
that the tendency of t,he present' subject sample Lo respond
wit,h respect to social desirability bias is similar to that
of the general population, âs indicated by Jackson's
normative sample. There is no reason to suspect that t'he
present samplefs social desirability bias was differentially
operative on the entropy items as opposed to the PRF items"
Thus confidence may be placed in Lhe reasonably unbiased
nature of the subjects' responses to the entropy items '
although results of any scale correlabe t'o some degree with
social desirabilitY bias.
In addit,ion, Pearsonrs I's calculated between individual
entropy items and the PRF scale indicate Lhat five entropy
items correlated more highty with social desirability than
66
threy did wit,h their respective factors. These items are
depicted in Table 15.
Table t5
Entro Items Most. Influenced a Wisltto seem socla v es ra e
ïtem#
45
54
2-
4
20
Correlation with PRF score(Pearson correlat,ion coefficients)
People in my family do t,hings to bug each other . "2479
In my family we care about each otherrs
Myof
family has values,/attiEudes similaroth e rs
Times my whole family spends wit'h outsiders areen joyable .
People in Ehe famity have the same opinion onth ings .
feelings. "2722
to t.hose"3966
.3299
.3416
67
Because the responses to these entropy items Ín
particular may reflect a wish to seem socially desirable,
they nay not as accurately reflect entropy as do those iterns
correlating more highly witfr the various factors established
than with the PRF" The PRF descriptive statistics are
portrayed in Appendix E"
6B
Chapter 4
Discussion
This study r¡ras performed to develop a measure of entropy
in the families of respondents obtained from their
perceptions of their families. In order to do this the
psychornetric properties of 33 items relating to entropic
family interact,ion patterns were examined. These items were
generated by a review of the Family Systems literature
dealing with Entropy, and this thesis is an attempt to
ensure that the measure (the Ent.ropy Scale) has content
validity, temporal stability, internal consistency and
factoral consistency. Pretiminary assessments of content
validity and statistical analyses of test-retest
reliability, . homogeneity' and factor structure ?rave revealed
â--r - i- lÀ- :-.:r.:^1 ¡^eÈ .i- r-'^r}.l¡rr nfEnat cne ¿nEroPy Þc¿lJ-e¡ III l" Ls L¡II LIq¿ !v!4r, !Þ wv! u¡¡¿ v!
being retained as a basis for further study of entropic
family interaction patterns. Revisions of the .scale will be
made ín the future, and will proceed in large part from the
results of the Present analYses.
Summary Statistics
As revealed by the t-test, the Entropy scale mean score
obtained with the subsample of 47 subjects was not
significantly different from that obtained with the total
69
sample of 28L subjects " The lack of a significant
difference between these meansr ðs well as the similar
values of the associated standard devíations, implies that
the data obtained from the two groups involved rnay be
pooled. As well, the test-retest relíability results, which
were based on the 47 subjects, ÍlâY be generalized to the
entire sample
Furthermore, as indicated by the histogram distributions
in Appendix C the values of correlations calculated in the
study may be assumed to be uninfluenced by problems wittr
"Skewness" of entropy item distributions. In particular, it
need not be feared that the factor loadings obtained for
some entropy items are deceivingly low due to attenuation of
the correlations involved. All but three items are normal
or positívely skewed, and only a substantial departure from
1ì+'¡ ^Ê ñãñr' a4 +ha Äì e{-rilrrrlinnc ñr .a l-onriênõ\t fr-¡rtÄv! rlrq¿ ¿ u-)' v ¿
many distributions to be skewe'd in opposite directions from
each other, would be suggestive of attenuation.
The drug control item, item 69, was included in the
questionnaire along with the Entropy Scale in order t.o
detect subjects whose behavior and/or emotions may
frequently be influenced by the use of non-prescription
drugs. The item reads: "I , have taken non-prescríption
drugs to change how I feel and think." Because only twelve
subjects replied either "always" or "usually" to this item,
70
the responses of our subject sample can be assumed to a
large degree to be unaffected by drug effects" If the
present non-clinical sample had been compared to samples of
delinquent or schizophrenic adolescents, this item would
perhaps have been even more important than it was in the
present study" That is, the item might have made it
possible to determine whether the use by some "normal"
subjects of LSD or another psychotomimetic drug at the time
of questionnaire administration was confounding the results
by precluding a "elean", accurate comparison of "normals"
and schizophrenics" The item will be administered, along
with the entropy items, to clinical populatÍons in the
future "
Items 75 to 77 read: "In answering the above questions
I was thínking of my mother", "In answering the above
questions I was thinking about my father" and "In answering
the above questions I was thinking of my whole family"
respectively" In the present clinical sample most subjects
(898) referred to their whole family at, least "sometimes"
while responding to these items, and a very minute
þroportion of the subjects (48) referred to their whole
family "never" or "rarely". This seems to suggest that much
of the information obtained from responses to the Entropy
Scale pertains to the whole family" At the same time, many
of the subjects ( 688 ) referred to their mother at least
7L
"sometimês", while 672 of the subjects referred to their
father at least "sometimes"" Thus a variety of responses
were obtained with these three items, and it cannot be Said
that the subjects were homogeneous in their "referral"
choices. This information will be useful in order to
ultimately make normative comparisons. It may be that some
subjects in "clinical" groups t"rill refer differentially to
varíous family members or to their whole family when
responding to the Entropy ScaIe than will those in other
diagnostic or "normal'n groups. Presentíng two of many
possibte examples, "clinical" subjects may predominantly
refer to a family member who they have a close alliance
with, while "normal" subjects might predominantty refer to
the entire family" In the individual case, if the subject
is a delinquent perhaps neither parent but rather a brother
or Sister would be referred to" In these examples it cannot
be assumed, for the purpose of either research or diagnosis,
that the Entropy score obtained relates to the entropic
Ínteraction patterns of all ímmediate family members" Ït
may be that the respondent, and one other family member have
a maladaptive, stifling alliance while all other family
members interact "normally" . In contrast, it may be t'hat
such an alliance is performing some function for other
famíly members, whose "ngrmal" interaction depends on it.
Many further examples of what, a high Entropy Scale score
72
might refer to could be given" In each case it would be
irnportant to plan further assessment and treatment according
to the responses to these three items" Thus the Entropy
Scale score, when interpreted in conjunction with the
responses obtained with these items, ñâY aid in the
detetrmination of wl.o in a particular family is in need of
treatmenti or perhaps what direction treatment could take,
for example loosening "bonds" with one member while
strengthening bonds with another.
Item number 7A reads: "The one with the most power in
my family is (Describe your relationshíp
with this person) " " In the present sample the father was
perceived as having the most power in a majority (598) of
families, with the mother being perceived as having the most
power in approximately half this number (322) of families.
Virtually no-one responded to this item with "I don't know",
although a response of "no-one", which may reflect some
uncertainty or a perceived division of power between
multiple farnily members, was given by a few subjects"
Thus it seems that in these "normal" families a definite
sense of who has the most poüJer, â11 important source of
information, exists. This item provides a means of
ultimately making normative conparisons between the present
subject group and "clinical" groups, and may be of
díagnostic significance if it permits a better descrÍption
73
of "normal" families and/or those of clinical groups. In
all cases it would be important to determine the ways in
which this pol¡rer is wielded in the family" . For example,
informãtion nay be obtained about how partícu1ar situations
are handled by the family, through interviews or direct
observation. This could be contrasted with information
obtained in parallel situations from families with different
power relationships as indicated by this item" If entropy
as measured by this scale conforms to its descriptions in
tire Family Systems literature, subjects from more entropic
families may be more uncertain as to family pobter dynamics;
oEt the most entropic families (on the basis of scores on
the Entropy Scale) may be the most skewed in terms of the
possessíon of power by one family nember" Comparisons of
Entropy Scale scores with responses to this item could help
to determine this"
Another question which could be asked is: "Does a
particular poürer configuration, or a lack of clarity of
po!.rer or rules, seem to be prevalent in the most entropic
families?" Interaction patterns involvíng power/rules are
important processes to be cognizant of in studying,
treating, or, indeed, being a member of an interacting
family" Thus this aspect of entropy, which emerged as
Factor V ( "Clarity of Power and RuIes" ) in the factor
analyses and consísted of items 10, "In my family it's clear
74
who has the most power" and 22, "The rules in the family are
clear", should be given a special focus in future studies"
Additional information should also be obtained through
interviews and direct observation, in order to clarify what
different responses to these items nean" This may involve
determining which situations the power is most and least
evident in, and which situations it is most beneficial
and/or counter*productive in. This information can then be
used in diagnosis and/or the planning of treatment, as well
as in the nosological assignment of entropic families for
research purposes" Of course, other aspects of entropy
might become evident once clinical populations have been
investigated and various "critical" or "diagnostic" iterns
and/or specific configurations of behavior are identified in
particular groups
fhe last four items on the questionnaire are it,ern 79,
'oSome of my actions have been considered by some to be
schizophrenic", item 80, "Some of my actions have been
considered by some to be delinquerlt", item BI, "Some of the
actions I have engaged in I consider to be schizophrenic",
and item 82, "Some of the actions I have engaged in I
consider to be delinquent". In a "normal" Sample such as
this, composed of first year university students, a greater
number of individuals might be expected to feel that they
"misbehave" (i.e. are or act delínquent) than those who feel
75
that they act in ways defined as "schizophrenic", and this
was indeed what bras found" This is perhaps because while
their behaviors often may not be approved of by the "older
generation" and hence might be deemed by the respondents
themselves to be "delinquent", this same behavior often
follows norms set by people their own age and thus would not
be deemed "scltizophrenic" " As well, their knowledge of what
defines "delinquent" would, probably be clearer than their
knowledge of what defines "schizophrenic", whÍch is far more
technical "
It will be important to include these items in the scale
when it is administered to "clinical" groups, âs a means to
increase confidence that thg "normal" subjects being
compared with the "clinical" samples perceive themselves as
'oï1ormal", and tO dif f erentiate between "cIinical" subjects
who pereeíve themselves to various degrees as normal,
delinquent and/or schizophrenic" If it is found that some
"normal" subjects differ from others by perceiving
themselves as detinquent, schizophrenic or both, these
subjects may be interviewed to examine their perceptions n
more dept,h. In this way the characteristics of their
families may be examined and compared with those of "normal"
subjects who feel they are. indeed normal, and the
possibility that treatment is needed may be explored.
Similarly, "clinical" subjects may be interviewed on the
76
basis of their responses to these items, and information
potentially valuable for diagnosis and/or treatment may be
obtained" fhe items will make it possible to compare the
diagnoses or labels applied by others to "clinical"
respondents with their own perceptions of their disorder,
and to examine some of the possible effects of various
"entropic'o famíIy interaction patterns on self perceptions.
The information obtained with these items can be
supplemented by that obtained in interviews. As well, these
items may help to assess the efficacy of treatment. Perhaps
therapy which proc.eeds according to problems which surf ace
wíth the completion of the scale wilI fail to change the
label of many respondents from "delinquent". However, if
many of these respondents feel that they and others can
ríghtIy consider themselves as less delinquent, which can be
indicated by responses to these items, support for the
efficacy of therapy can be obtained"
Of course, one must remember that these items reflect
the perceptions of the individuals completing the
questionnaire, not "reality" " It would be ínformative to
determine, in future research, whether others' opinions of
an individual as delinquent or schizophrenic confirmed the
indívidual's ansu¡ers to these items. More confidence could
then be placed in these items for that individual, and much
information could be obtained from them, such as specific
77
thoughts, behaviors and emotions which the individual feels
are abnormal or would seen by others or ¡rarticularindividuals as abnormal"
Although Entropy Scale scores may correlate
significantly with an individual's score on items requesting
self assessment ( items 8I and 82) , they may correlate
significantly hígher with the indivídual's score on the
items (79 and B0) relating to what he/she perceives as being
others' perceptions. This information may ultimately be
important in the planning of treatment" For instance,
interviews may determine that some respondents over-estimate
family members perceptions of them as schizophrenic or
delinquent. If this is known, it may be possible to focus
therapy to increase the accuracy of their perceptions, in
order that, they may begin to feel less estranged from their
family and others"
Temporal StabilitvThe analyses performed to determine test-retest
reliability indicate that the Entropy Scale is reasonably
stable for the one-month period examined. A correlation of
"4974 is signÍficant, although not as high as one rnight
wish" The low temporal stability of some of the items Ìras
depressed the overall temporal stability of the scaIe.
7B
Future research should also determine the test-retest
reliabitity of the scale using longer intervals"
fhe t,est-retest reliability analyses of individual items
on the scale indicate that 22 out of 33 Entropy items had
significant test-retest coefficients at the.05 level or
better. ThÍs indicates that subjects' responses were fairly
stable over the one-month period wittr respect to many
items" Those 11 items with relatively low tenporal
stability may be badly worded items, or are perhaps unstable
for other reasons. fhey or the scoring of them should thus
perhaps be deleted from the Entropy Scale, in order to see
how this influences stability. Rewording them míght also be
explored "
Many of the items with relatively low test-retest
coefficients are located near the end of the questionnaire"
It is possible that fatígue effects influenced the responses
to some of these last items on the questionnaire" To test
for this possibility, the order of the items on the
questíonnaire caR be reversed and the test-retest analyses
can be performed again with different subjects" The
hypothesis that subjects respond more randomly with
increased fatígue would be supported , if the iterns which
previously had low test-retest coefficients increased the
size of their correlations and the temporal stability
coeff ici ents
decreased "
which might
analyses to
factor "
of
Thi s
be
79
the "new" items placed at the end
would imply a need to shorten the scqle,
done by deleting items found by factor
overlap with similar items loading on the same
A few of the items with relatively low temporal
stability coefficients do not occur near the end of the
questionnaire. Item 26 lacks ternporal stabílity, which mâlr
again, be because of misinterpretations due to its wording
or some other known reason(s). Specifically, responses to
"Things are peaceful at home" may be based on an immediate
temporal framework" Some respondents may feel that the item
refers to the family environment at the time of
admínistration of the scale, which nay differ from the
family environment as perceived at the next administration.
Were respondents to instead respond in terrns of the general
degree of peacefulness in the family, which may be perceived
as relatively constant, this item might attain a higher
temporal stability coeff icient"
Responses to item 39, "People in my famíIy seem to
disagree in their opinions.on most things", may be similarly
influenced by fluctuating environmental condítions, thus
explaining its relative lack of temporal stability. In
particular, the number and severity of disagreements between
family members may be expected to fluctuate somewhat over a
80
one-month period, which might render responses to this item
unstable.
Responses to item 6L, "In my farnily we try to keep time
for each other", may have been influenced by the two tines
at wt¡ich the test-retest data was obtained" The fírst test
administration occured at the beginning of the first term of
university, a time at which it. may understandably be easier
for a university student to keep time for his/her family.
The second administration occured one month into the term, a
time during which increased academic responsibilities may
preclude time being easily available to spend with the
family" Thus the nonsignificant, negative correlation found
between the mean Entropy Sca1e score obtained on item 6l at
Time I and Time 2, one month later, perhaps becomes more
underst,andable" It may instead be that the case study
reports of entropy as involving some degree of
noncommunication among famíly members need revision. Wit,h
future studies the reasons for this as well as other items'
instability may become clearer"
Yet another possible explanation for the relatively low
temporal stability of some items is that the "normal" sample
studied gave different responses to the items than would a
"clinical" group, whose responses may more closely conform
to the clinicatly based Iiterature; especially if the latter
group comes from families more consistently "entropic". of
81
course, some or aII of t,he relatively unstable items may be
useful despite t,heir instability. The very variabilit.y of
the responses of some subjects to these items over time nay
permit, t,he items to aid in the diagnos is of maladapt,ive
interact ion patterns; or ¡ to discr iminate betr¡reen normal and
dysfunctional families. As well, their possible suscept,i-
bility to various changing circumstanees over time may render
Ehem useful in assessing the effects of treatment interven-
tions. Only time and further research can clarify this"
Int,ernal ConsisLency
The internal consistency analyses resulted in a
Cronbach's Alpha of " 8078, which signifies a fairly high
degree of internal consistency" The it,em-total correlations
support. t,his st.atement, as only a f ew items had negative
correlations with the t.otaI Entropy Scale. These it,ems were
item 10r'In my family it's clear who has Lhe most powern,
item 39, rPeople in my family seem to disagree in theiropinions on most, thingsn, item 63, nIn my home outsiders
make social visitsn, item 65, ol4y family is afraid to be
seen as dif f erent by ot,hersn , it,em 68, 'With respect to
variety my family would be considered boriflgn, and item 74,nPeople in my family express their anger openlyn. With the
exception of items I0 and 39, t.hese items occur near the end
of t.he scale" Hence the effects of fatigue may again be
82
responsible for these items' low item-EotaI correlations '
In fut,ure administrat,ions of t,he Entropy scale they will be
placed near the beginning of the questionnaire, in order
that, they may be revised or deleted if the'ir associat'ed
item-tota1 st,at.istics do not improve.
Responses to item 10, 'In my family it's clear who has
the most povrer", may have been influenced by bhe clarity of
porr¡er and rules in subjects' families" The clariLy of po!''er
and rules in the family relates to entropy as viewed in the
Family systems I iterature, as case study descr ipt ions of
'entropicn families often portrai Lhem as lacking
information about who is'the most influential in decision
making" As well, nClar ity of Pov,ter and RuleSn was f ound in
the factor analyses to comprise one of t'he six factors into
which bhe scale items víere grouped. Howeverr responses to
item t0 may also have been influenced by the degree of power
which subjects perceived as being possessed by particular
family members. If this latter variable is not related to
entropy¡ unlike the nclarit,y of Power and Rules" variable,
the relatively low correlation of this item with the rest of
the scale may be explained. Specifically, responses to the
item would then have been inftuenced by an aspect of family
power dynamics, i . e " who in the family is perceived as
having the most povüer and,/or how large a difference is
perceived between the degrees of pov¡er possessed by various
B3
family members, which nay have littIe or no relation to
aspects of entropy covered by other scale items. If a
higher item-total correlation is obtained in the future when
this entropy item is reworded so that only the clarity of
pourer and rules is determined by it, this hypot,hesis would
gain credibility"
Factor AnaL ses
The factor analyses of the Entropy Scale were best
accomplished, i.e" led to the simplest solution, when an
oblique rotation was performed" Some degree of correlation
between the six factors was thus indicated. This
correlation might be expected from both the Famíly Systerns
literature dealing with entropy, which suggests that a
constellation of "entropic" characteristics nay permeate the
interactions of some families rather than a single aspect of
entropy, and from the fact that entropy is a complex,
multifaceted variable.
Tnterpretation of factors
Factorl-Thisfactor, "Togetherness and Tranquility",
is outlined in Table 7 " The greater the score obtained on
the items comprising the factor, the greater the interchange
of information within the boundary of the family system may
exist" An individual scoring high on the factor may
generally describe his/her family as being closer, more
B4
cohesive, relaxing to be with and considerate of its members
tÌ.an would an individual scoring lower" The possible
utility of the factor stems from the potential of its items
to explore the perceived "quality" of relationships among
family members " The general feeling of family members
toward one another is an aspect of family life which, ifdetermined, will aid in both diagnosis and treatment
planning. Scores on other factors' items will be more
understandable if interpreted ín conjunction with scores on
items comprising this factor. The degree to which a lack of
closeness and/or "good feelings" among family members as
opposed to other aspects of family life contributes to
maladaptive behaviors will become cleareri as will the
degree to which closeness and/or "good feelings" may
contribute to problems. The most appropriate diagnoses and
points of intervention wiIl become more evident with
follow-up interview data.
Because most items of Factor I (with the except,ion of
items 26 and 71) contribute highly to both the test-retestrelíability and the internal consístency of the Entropy
Scale, these items are important to it. However, furtherrefinements of the factor may be made on the basis of future
tests with "clinical" groups" One possible means of
refinement may involve rewording item 26, which has been
mentioned abover ërs we1I as item 7L, if placement at the
85
beginning of the scale faits to increase item 71's
stabílity. Again, with a "clinical" group different results
might obtain"
Factor 2 - Factor II, "war and Peace in the Family", is
outlined in Table 8" It is comprised of seven items ( items
33, 23,9,26, 32,39 and 30), six of which have negative
loadings" Upon first inspection the negative versus
positive loadings do not all make sense. Therefore, before
factor II is interpreted it is necessary to describe and
give possible reasons for the direction of its factor
loadings. ftre direction of these loadings seems at first
contra-intuitive, hence the interpretation of the factor nay
make greater sense giverr some comment on them"
Four of the seven items have negative loadings and imply
a "warlike" atmosphere in the famÍly if responded to
positively" However, item 26, "Things are peaceful at
home'0, has a negative loading but implies peacefulness if
positively responded to" A possible reason for the
contraintuitive negatÍve loading of ítem 26 may be that
responses of some subjects to the item have been based on
immediate rather than ongoing conditions in the family
environment. "Things are peaceful at home'o may imply to
many respondents that the degree of peacefulness in the
immediat.e (present ) family environment is in question.
Hence responses given by respondents to the item may not
B6
conform to their responses to those items in the factor
which they perceive as per,taining to longer term, gleneral,
ongoing conditions" This possibility, combined with the low
test-retest reliability coefficient obtained by this item in
the temporal stability analysis, makes the item a príme
candidate for change or elimination"
Item 30, "Our family tries to keep things calm", has
peaceful implications despite its negative loading" This
can perhaps be explained in the following t¡ray: A positive
response to the item, which reads: "Our family tries to
keep things calm", ñêy indicate a perceived need in some
respondents' fanilies to keep things calm because of a
general state of unrest. and being "close to the edge of
blowing up", rather than an actual state of calmness in the
family environment" If this is so the negative loading is
explained as it becomes clear that the item does not have
peaceful implications" This makes it similar to the other
items with negative loadings on this factor. Direct
observation of a family and/or questioning of respondents
should provide information
case" Both thís ltem and item 26 will be scrutinized in
future administrations of the questionnaire to "clinical"groups and wiIl be deleted or reworded if similar
incongruities remain or emerge"
87
Fínally, item 39, "People in my famity seem to disagree
in their opinions on most things", has a positive loading on
this factor, Yet it seems to relate to a warlike
atmosphere, as do the four items which have negative
loadings" Perhaps this incongruity is because a positive
response to this item signifies sornething different from the
state of "rnrar" in the family which the other items
describe" A positive response may instead índicate an
ability or tendency of family members to openly discuss
thoughts and feelings without risking "hrar" " It should be
noted that this item, like item 26, possesses a retatively
low test-retest coefficient. Again, there nay be a need to
reword or elirninate this item in the future" The
possibitity might also exist that with a "clinical" group
both items would be more stable, and would "fit in" to a
greater degree with their respective factors.Thus, Factor II, "War and Peace in the Family", seems to
relate to the degree of peace versus unrest which isperceived in the family environment" Possible reasons for
the direction of the loadings have been speculated upon
above. An individual responding positively to the items of
this factor may generally describe his/her family as less
peaceful, in terms of length of arguments, Ievel of
organization, noise level, need for rules, amount of general
agreement among members and protection from the outer
BB
environment, than an individual scoring lower. Items 92 and
39 seem to have relatively low item-total and test-retest
reliability respectively" Item 26 has relatively low
teét-retest reliability" These items should be flagged for
elimination or future refinement, âs discussed above"
Perhaps the terms "ru1es" and "chaos" in item 32, "Without
rules my family is in chaos", could be more specifically
defined. If the item is understandable to most if not all
subjects the responses to it may conform more to subjects'
eotal Entropy Scale scores. Item clarity may be even more
crucial to stable responses i f a i'"Iinical " group with
limited reading abili.ty is tested. Similarly, in item 39,
"People in the family seem to disagree in their opinions on
most things" , the word "most" could be elirninated. Its
inclusion produces two quantitative terms in the statement
once a subject responds on the Likert scale, these being
"most" and the subject's selection from the Likert
continuum" This may be confusing to the subject. If
similar results are found with "clinical" groups, these
refínements may be advisable.
Factor 3 - This factor, "Transparency of Family
Members", is detailed in Table 9. It relates to the
openness of family members, their tendency to express
thoughts and feelings, êrs well as the clarity of those
thoughts and feelings to others in the family" The aspect
B9
of entropy related to may be the amount of freedorn of
information exchange within the famíly system boundaries,
especially information relating to indÍvidual components
( menbers ) of the system" This factor thué ultimately
relates to an aspect of "metacommunication", the degree to
which members are open about their relationships with each
other¡ ârd goes beyond dealing simply with their own
feelings. As such it is diagnostically important and may
aid in the planning of treatment. The FamiIy Systems
literature dealing with entropy relates that an irnportant
aspect of entropic information loss in the famity involves
statements about relationships among its members, and the
emotions involved
Factor 4 - This factor, "Rigidity and Conformity", is
detailed in Table 10" It seems to relate to the degree of
spontaneity which a family exhibits. Spontaneity in this
sense refers to the family's willingness and tendency to
depart from existing schemes or plans, with or without
apparent reasons for modifying them, and from perceived
social norms as well " The positive or negative affect
associated with changing plans is included in this. fhus an
individual scoring high on this factor may describe his/her
family members as more apt to alter plans, less conforming,
and less concerned about being perceived this way than an
individual scoring relatively lower" Responses to items in
90
this factor, coupled wittr affective information, would be
helpful as an index of the possible difficulties of treatinga particular individual or family. WhiIe positive responses
to item 29, "My family hates to change plans" and item 35,
"My family always sticks to its plans no matter what" may
indicate that promoting change in a family may be difficult,a positive response to item 65, "My famity is afraid to be
seen as different by others," may work to. a therapist'sadvantage" The use of social pressure in treatment may
sometimes be used in a manner beneficial to a family which
is accurately perceived by its members as basically"conformist".
E+ç!er S This factor, "Clarity of Power and Rules", isthe only doublet factor" Only two items loaded on it but itis not "trivial" as Gorsuch (fggs) has written about such
factors, âs both of its loadings are above "30. The factorappears to involve the clarity of power and rules in the
family" The perceived clarity of information regarding
which member hotds the most power, ês well as the perceived
clarity of "rules" to be adhered to by famiì-y members, are
the aspects of fanily interaction dealt wittr by thäse
items" The implications of both high and low scores on the
items comprising this factor are many. To the degree thatboth the cont,ent of f amily rules and who holds the most
pohrer in the family are perceived as being "clear", a family
9l
could range from being flexible and adaptively "organized"
to being destructively and intolerantly rigid in its power
and rule structure. The family could be tyrannized by one
or more members who create irration.lr' inconsiderate rules
of a "black/white" nature, ol it could be eff icientty "run'",
according to members' changing needs" Other possibifities
exist. In contrast, to the extent that power and rule
issues are perceived as unclear, complete confusion and
"chaos" may reign, leaving family members feeling frustrated
and necessary tasks undone; oi:, the family may be perfectly
satisfied with a "democratic" process of negotiation, which
precludes the need for rules or a power structure. The many
interpretations of this factor indicate a need to supplement
information obtained from its items with that obtained by
interviews and/or direct observation. In this way the
interaction of power dynamics with problems experienced will
become clearer. Freedom of information exchange is an
important aspect of entropy as shown in the literature, and
an important, source of information in the family relates to
pou¿er and rules. Thus this factor is an important one.
Factor 6 - This factor, "Social Appearance and
Interpersonal Enjoyment", is outlined in Table L2" Items 7L
and 47 have been included in the description of this factor,
as the corresponding coefficients are only slightty below
" 30 and their inclusion helps to define the factor "
92
ïnterpersonal enjoyment in this context seems to involve the
degree of enjoyment or lack thereof which is experienced by
family members according to the respondent" An individual
responding positively to item 68 or 47 nay describe family
activities as more lacking in excitement or as more orderly
( respectively) than would an individual scoring relatively
lower. Item 7L has a positive loading, which is
understandable because a positive response to this item
signifies interpersonal enjoyment, while a positive response
to items 68 and/or 47 signifies an absence of enjoyrnent"
lfith a "clinical" group respondents from more "entropic"
families may perceive that an orderly socÍaI appearance is
essential to interpersonal enjoyment, if case study reports
in the lit,erature dealing wíth the apparent conformity of
"entropic" families are to some degree accurate" How the
need for "correct" social appearance hinders enjoyment, as
well as which aspects of farnily life are viewed as most and
least enjoyable, can be further explored through
interviehrs " The best means of intervention night then
become more evident"
Factor summarization"
The factors extracted correspond fairly well to many
aspects of entropy as viewed in the Fanily Systems
literature" Those aspects involve the amount of information
exchange within and without the family system boundary,
93
especially as this infornation exchange affects the degree
of enjoyment experiencedr pêrcêived behavioral/at,titudinal
rígidity, the perceived clarity of family norms, knowledge
of individual members' thoughts and feelings, perceived
calmness and peróeived cohesion. The factors which have
emerged from the analyses thus seem to indicate that the
Entropy Scale may in fact allow the gathering of information
on many of the aspects of entropy deemed important in the
literature. Famities nay profit from completÍng the scale'
Their position on various continua relating t'o entropy can
perhaps then be more validly and reliably ascertained, and
both diagnosis and treatment can proceed on the basis of
scores obtained on the items of particular factors. In this
way t,reatment can be "tailored" to fit the need of specific
families and individuals within them. Treatment of this
sort may be much more efficacious than that designed simply
to decrease a hard-tO-interpret, overall "entrgpy" Score, or
that designed to decrease the problems indicated by scores
on individuat ítems whose factor composition is not known.
The former treatment would lack direction due to not having
a firm basis from which to begin, such as the score obtained
on particular factors " fhe latter treatment, based on
problerns perhaps identified by individual item scores, would
lack an integrating theme and thus would also lack direction"
94
Rel aL ionsh ip of bhe Entropy Sca1e to the Social oesirabilitv
Scal e of t,he Personal it Research Form AA (Jackson 1967 )
A significant correlation was found in this study
between t.he mean Entropy Scale Score and the mean score
obtained on Lhe Personality Research Form" A correlation
just as significant or larger may or may not be found when
nclinical" gtoups are tested in the future" The imporLance
of Lhe PRF in Ehe present study is that the responses of the
presenL sample have been found similar to t'hose of Jackson's
original normatiVe sample. Because no significant'
differenceshlerefoundbetweentheseresponsesasocialdesirability nSet" can be assumed to not have been operating
in the present sample" The items of the Entropy Scale thus
do not appear to facilitat,e the development of such a
* seÈ n . Demonstrat ion of t,he f ormat, ion of such a set' woul d
have invalidated the Entropy Sca1e.
Concl us i ons and Future Research
This study has been an attempt to develop a scale
measuring entropic famity interaction patterns as perceived
by the respondents. The items ltere derived from a review of
the Family Systems I iterature on eneropy. The review
revealed that this form of interaction may exist alone or in
conjunction with cenLripetality/centrifugality or other' as
yet unnamed interaction pat,terns, and may promoLe
95
significant pathology in family nembers. Further research
should follow the psychometric validation which has been
performed" This research should be based on the
administration of the questionnaire to a "clinical" groupr
composed of adolescent delinquents and schizophrenics, to
determine which of the reTationships and psychometric
properties found in this study repeat themselves and which
would differentiate between these "cIinical" groups and
"normal" groups. The Entropy Scale has shown itself to be
fairly reliable over time, to have a high degree of internalconsistency, and a factor composition which makes much sense
on the basís of the lit,erature. Knowing its psychometric
properties nakes it possible to refine the scale.
Those entropy items which did not attain a sígnificanttest-retest reliability coefficient will be flagged as
possibly "bad'o or "unreliable", and may be elininated f rorn
the questionnaire" However, the normal subjects who took
part in the study may not be typical of "cIinical" groups of
either sehizophrenic or delÍnquent adolescents, and greater
test-retest reliability may be found with a "clinical"group" As well, even though an individual's response to a
particular iten may vary across a one-month period, the
information thus obtained at both administrations may be
important in the planning of treatment" Both the general
"traits" of individuals and their families and the various
96
"states" experienced across time are important in diagnosis
and the planning of treatment. In particular the revisions
of diagnoses and the ongoing modification of treatment
depends on'knowledge of changing states in the Índividuals
concerned, as well as knowledge of which circumstances these
changing states covary with. However, those items which do
remain relatively constant across time may promote greater
understanding of some of the more stable characteristics of
entropic families than do those which are less stable.
A "clinical" group nay be even more subject to fatigue
than was the sample of university students who took part in
this study. Thus it, may be wort,hwhile to make the scale
more concise by eliminating items which repeat information
already obtained, if subjects' fatigue indeed seems to be
influentíal. Another influential variable may have been the
time at which subjects completed the scale, which no doubt
influenced not only the quality and quantity of experiences
with their families just prior to the scale's administration
but their perception of these experiences as well.
Specifically, with the subgroup of 47 subjects the first
administration of the scale came at the beginning of the
school term, before academic responsibifities may have
greatly affected them" The second administration occurred
one month into the term, ât the time of midterm
examínations " Some iterns with the lowest test*retest
97
reliabilíty, for example items 26, 39, 57 , 68 and 7L,
pertain to disagreements, adequacy of communication, amount
of routine present in family life and enjoyment of family
activíties; all of which may have varied great'ly brith
changes in the amount of respondents' schoolwork. The
non-significant, negative correlation found between Time I
and Time 2 fot item 6L, an item relating to the amount of
time kept by family members for each other, provides a
concrete example of the possible variations in family life
occurring over time. With a "clinical" group time of
questionnaire completion may be important also, although it
may be less so or more so, varying differently with the
calendar" For example, âñ approaching holiday may
positively color respondents' perceptions of their
families. The time elapsed. since admission of an individual
to aR institutíon may also be crucial in determíning his/trer
responses "
Certainly, test-retest reliabitity may not be examined
adequately Íf a variable which has potential to influence
responses is present only at one administration of the
Entropy Scale " Thus, events occurring in the lives of
respondents at the time of administration of the scale
should be carefully explored in the future, in order that
responses at these times t^¡ilI not be misconstrued as being
"typical'u " It should be noted that this applies to most
98
tests and is a variable that should be explored even wit,h
Lhe more widely used and more "standardn tests. Wit,h
respect to t.he Entropy Scale, valuable information may be
gained if the influences on questionnaire responses of
particular happenings can be explored through interviews.
Various other types of validat,ion will eventually be
required if the Entropy Scale is to be most effective, in
terms of directing therapy and aiding in differential
diagnoses. v0ith respect to construct validity¡ ârl
instrument has been developed to assess the degree of
centripetality and centrifugality in families (toff, L982,
1984 ) . Thus with the present measurement of psychometric
prope rties of the Entropy Scale it wilI now be possible t,o
test Olsonrs theoretical model (1983), as well as the models
of Stierlin (L97I, I973, I974) and Beavers (L977t 1982). Itmay be possible to show that cenLripetality, when combined
wit.h entropy, Ieads to the development of schizophrenia in
the offspring of some families; whereas centrifugality, when
combined with entropy, leads to delinquency in the offspringof some families. The degree of entropy, centripetality and
centrifugality which ís necessary for the development of
cert.ain problems¿ âs well as some of the predisposing
conditions which when combined with any or aII of these
three variables increase the Iikelihood of occurrence of
problemsr lnây become more evident. The results of the
99
latest study of the Centripetality/Centrifugality Scale
(Loff, I984.) suggest that centripetality and centrifugality
lie on one contilluum, with extreme centripetality at one end
and extreme centrifugality at the other, âs Olson (fggE)
suggests. This hypothesis, that entropy increases with both
a high degree of centripetality and a high degree of
centrifugality, nay thus be correct" Future correlational
studies of the interaction between these three variables
among "clinical" groups may increase the evidence for this
curvilinear relationshiP"
On this topic, it may be possibte to ultimately
identify, on the basis of responses to the items of the
Entropy, Centripetality and Centrifugality Scales, fanily
interaction patterns which relate adaptive "closeness"
versus those which relate maladaptive centripetality"
Similarly, it may be possible to identify families which may
produce independent, children, versus those which may produce
"delinquents". Specifically, it, may be that a relatively
high score on the Entropy Scale combined with a high
Centripetality Scale score wilI be obtained by respondents
whose families prove through direct observation and
interviews to be "dysfunctional" in some way; whereas a high
score on the Centripetality Scale but a relatively low score
on the Entropy Scale may not be found among respondents of
dysfunctional "centripetal" families, but rather among
100
respondents of nclose" families. Similar research may be
performed with respect to the Centrifugality ScaIe, in order
to ultimately differentiate between the interaction patterns
comprising dysfunctional n centrifugalityn and those
comprising adaptive "independencen in families.
FinalIy, predictive validily of the Entropy Scale can be
explored. It may be found, for example, that particular
scores obtained on the scale lead to specific problems in
the respondents involved. In this wayr it may eventually be
possible to predict future problems on the basis of present
Entropy Scale scores, and thus to plan interventions to
prevent those problems
Presently an attempt is being made to contact future
clinical subjects, in order that the above psychometric
testing may be replicated with these subjects. In addition,
it. is well known that university students do not represent
the average individual, as they generally come from families
of above average socio-economic status, they are generally
of above average intelligence, they are presumed to be more
nliteraten, etc. Thus nnormaln subjecLs in the future wilIbe randomly selected from the general population" An
attempt. can be made to include Native Canadians in futureonormaln groups as wel1" These individuals are often overly
represenLed in groups labeled "delinquent", and it will be
r01
important to compare "delinquents" with nnormaLn individuals
who are culturally as similar to them as possible.
If part,icular it,ems are responded to differenLly by
f ut,ure n cl inical n groups than by the present and,/or f uture
normative samples, these iEems can be used as aids in
treatment planning" For insLance, it may be found Lhat many
nclinicaln respondents find their famil ies [o be
ndisorganizedn in various v¡ays, and for this reason do not
Iook to them for support for personal problems. Determining
Ehe ways in which these respondents wish tLrat their families
would increase in norganizationn, as weIl as the vüays in
which t,hese families can be helped to come closer to the
respondentsr ideals or the ways in which the respondents can
be helped to alter t,heir expectations r Ítay then be the most
effective treatment.
In order to preclude the confounding of ent,ropy with
various characteristics of families such as et,hnicily,
family size , teligion, educational 1eve1, occupational
status and socio-economic status' information should be
obtained from respondents regarding these variables. IL
w iII perhaps then be l ess probable that t,hese
characteristics and their effect,s will go' unnoticed and be
mistaken as indicating dysfunctional interaction patterns.
Tn particular this possible confounding may occur due to
unacknowledged ethnic status. For example, an Ir ish
L02
family's general stoicism in the face of a sick member may
be misconstrued as a sign of an nentropicn Iack of
communication rather than a sign of a ntypical" reaction to
illness in Irish families, which may in fact be perceived as
adaptive among Irish people (ZoIat L966)" A family such as
this can be helped through treatment to deal nstoically'
with the problem, wit,hout hiding all thoughts and feelings
from each oÈher r âs the information on the ethnic Status of
t,he family may permit, a more informed approach t,o treatment.
Qualitative analyses of family interaction patterns,
especially Ehrough direct observation of and/ar intensive
interviews with families who obtain especiatly high or low
entropy scores, may help to attach more concrete, 'clinicaln
pictures to the scores received on Lhe Ent.ropy Scale. It
may be possible to determine whet,her increasing entropy
scores correlate with increasing overt, evidence of various
problems in or dysfunctional aspect,s of family interaction
paLt,erns in the f amil ies st,udied ' which may or may not
conform to Lhe I iterat,ure. Individuals scor ing high on
various entropy items can be asked to complet,e Lhe scale
both before and after treatment, and the effect on future
scores obtained with these items (and these same
individuals ) of various intervention techniques can be
determined. In this way perhaps more effective treatments
foc parLicular ent.ropy-related problems of particular
r03
individuals will be found" The meaning of different members
of the same family having different or similar scores might
also be explored.
Additional knowledge may be gained through retrospective
analyses involving older subjects. Hovüeverr Prospective,
longib,udinal studies, involving young subjects who are
followed after completing the Entropy Scale in order to
examine future problems which are encountered by Lhemr fiìâY
have greater credibilit,y. This is because retrospective
information, being dependent on memoryr maY be subject to
many information processing errors and more fraught with
inaccuracies than is prospective data. As weIl, prospective
data may be used in the prevention of future difficulties in
" entropie n famil ies r âs it may alert treat.ment special ists
to potentially problematic interaction patterns before bhey
become eoo "crysLallizedn for "easyn modification. Thus a
critical, discerning stance wiII be taken with respect to
any data obtained. Only in this v/ay will entroPYr a perhaps
widespread but as yet largely unexplored phenomenon of
individual psychological pathogenesis "or family breakdown,
be more thoroughly explored. Entroy is thought of as
inevitable in the physical sciences, but is perhaps either
preventable or more amenable t.o change in the social realm.
r04
Refer ences
Aftanas, M. S" (1983 ) .psychologY. Manuscri
Anastasi, A . (f976) " Psychological testing, AEh ed. New
York: Macmillan eubÏsEîñg co. ÙÈd.
Standard s stems of measurement, inÞ m Ite or pu cat on.
Arthurs, R. F., and MacCrimmon' D. J. (1978).performance deficit in clinicallylchizophrenics: A marker of schizophrenia.Abnormal PsvcholoqY 87 , 597-608.
& Weakland'Behavioral
Res idualremit,ted
Journal of
J. (r9s6).Bateson, G.,Toward a25L-264 "
Jackson, D. D., Haley, Jtheory of schizoPhrenia, Sc ience r,,
Beavers, w. R. (L977 ) .Brunner Mazel, Inc.
Psychother a & qrowth. New York:
M. N. ( 1983 ) . ramilY Models:the Olson circumplex model with
22, 85-98.
a schizophrenia.
Beavers, W. R" and VoelIer,Comparing & contrasEingBeavers' sYstems model. ramily Process t
Birchsbaum, M" S. (L977.). PsychophysiologySchizophrenia Bulletin, 3, 7-L4.
Fanily therapy in clinical practice. NewBovten, M.York:
(1e78).Jason Aranson.
Cloward, R. A. & Ohlin, L. E
opportunitY. New York: Free(L977 )
and VüiAbnormal Ps chol o
ns on.
(1960 ) .Press. InToront,o:
Essentials of Þ chol oHarper ovü u 1s ers.
DelinquencY &
B. Martin ( Ed. )
Holt, Rinehart
Cronbach, L3rd ed.
Faunce, E.scales.
J" (1970).New York:
ical testin f
E. & Riskin, J. (1970 ) . FamiIYArchives of General Ps 22,ch iat ry
interacLion41-65 "
Gorsuch, R" L.Hillsdale' N.J
(1983).Lawr ence
Factor anal sis,Er aum Assoc ates,
2ndInc.
ed.
Gottesman, I. & Shields, J. (f912) "genetics: A twin van!-age Point.
Press .
Schizophrenia elqXew York: Academic
Haley, J. (1964). Research on f amily patt"erns:Process, 3 t 4l-65.instrumenL measurement . Fam il y
AN
rril,in
r" M" (tott¡ "
schi zophrenia "
,Jackson, D. N. (fOeZ¡"Goshen, New York:
Lewis,V"in
Un
Loff,
Mark, J" C. ( f953 ) . The attitudesschizophrenics toward childAbnormal and Social Psychology,
r05
Quatitative & quantitative EEG findingsSchizophrenia Bulletin, 3, 6L-79"
Personalit research form manual.Researc
personal andto observedJournal of
Fami Iyfami 1y3-L2
"
syc o sts Press, Inc"
Jacob, T" ( fgZS) " Family interaction in disturbed andnormal families: A methodological and subst,antivereview. Psychology Bulletin, 82,33-65"
Johnson, S. M. and Lobit.z, G. K. (tol+¡ " .Themarital adjustment of parents as relatedchild deviance and parenting behaviors "Abnormal Child Psvchiatry , 2, L92-2O7 "
Kelsey-Smith, M. , and Beave.rs, W. R" ( f 98f ) .assessment: Centripetal and centrifugalsystems. Arnerican Journal of Family Therapy, 2(4) ,
Lennard, H. L. and Bernstein, A.interaction. San Francisco:
(f969) " Patterns in humanJossey-Bass Inc.
J. M., Beavers,A. (tølø¡ . No
!rl" i R., Gossett, J. T. and Phitlips,single thread. Psychological health
nerfMazel-, Inc.family systems" New York: Brun
Lídz, T., Fleck, S. and& the family. New
cornelison, A (fOAS¡. schizophreniaYork: International Universities
Press.
Loff, C"D. (f982). Development of an instrument t,o agF€tterc tions of fam centr I etá1it and centrifu a1i t
s members" Unpu S pre-Master es s,vers r- tyo Man itoba, Winnipeg.
st
C"D. (1984)" Development of an instrumen! to eÊteteerc tions of fam centr 1 etal i t and centrif ali t
s members. Unpu s Master s es s, vers vô n nnÍpeg "
of the mothers of rnalebehavior. Journal of48, 185-189 "
Martin, B. (L977). Abnormal psychology: clinical andscientific perspectives , 2nd ed. New York: Holt,Rinehart and Winston"
McCord, hI., McCord' J" and Jola, I. K. (1959)"crime. New York: Columbia University PresEãffin (Ed.), (1977). Abnormal PsychologY.
106
Origins ofs. In B.
ToronLo:HoIt,, Rinehart and Winston.
Mednick, s. A. and schlusinger, F' (1968) ' "some premorbidcharacteristics related to breakdown in children withschizophrenic mothers. rn D. Rosenthal and S. S. KetY
of schizophrenia. New York:(nds.¡, The transmissionPergamon.
Mednick, S. A. and Schlusinger' F.theory of schizoPhrenia: ThirteenHammer, K. Slazinger, and
(1973)" A learningyears later" In M.
E. Sutton (Eds. ) ,PS cho atholo Contr ibut,ions from bhe social
e AV ora & o og ca Ðv
Mishler, E. G. and Waxler, N. E"families. New York: John WileY
(r968)"& Sons.
Interaction in
Family env i !enlnen!__eç 31 s
ences. evJ Yor w ey.
Moos, R.H.,manual.Press.
and Moos, B.SPal o Al t.o ,
(r98r).Cal. : consutting PsYchologists
Mosher, L" Rtr eatmentHospital
. and Menn, A. ( 1978 ) .f or sch i zophr enia: Tvio
Community residentialyear f oIIo\,r-up data.
29, , 7L5-723.
K.,the
and Community Psvch iat ry
. H., Jenkins, J. G.' Steinbrenner'Nie, N. H.' Hul1, C
and Bent ' D. H
social sciencesStat is t ical cka e for( r975 ) .
(2nd Ed.).ComPanY.
Nunnally, J. c. ( 1967 ) .McGraw-HiIl Book co.
New or McGraw- oo
Psychometric theory" New York:
Introduction to chol o icaLr avr- H Boo Co.
NunnalIy, J. C
measurement.(1970).
New York:
Olson, D. H., Russel1,Circumplex Modeltheoretical uPdate.
Pope, W. ( 1976 ) .Chicago: Uni
C. s., and Sprenkle, D. H. (1983) "
of mar iEal and famiIY sYstems: VIpamily Process , 22' 69-83.
Durkheim's suicide: A classic anal zed "
ver s Yo .c cago press "
F. K.acute
32,
Post , R. M., Fink | 8", Carpenter, W. T. and Goodwin,(f975). Cerebrospinal fluid amine metaboliIes insch i zophr enia .
1063-r069.Archives of General Psvch iat rv
thenet.
r07
etiology ofSchi zophreni a
Reiss, D" ( f975 ) " Families andschizophrenia: Fishing withòut aBulletin, L4, 8-lI"
Rutter, M. ( 197I ) .effects on theand Psychiatry,
Rosenthal, D" (197I)"York: McGraw-Hi11.
Genetics of psychopathology" New
Rosenthal, D., Wender, P. H., Kety, S" S", Schlusinger, F.,Wilner, J" and Rieder, R. (f975). Parent-chifdrelationship & psychopathological disorder in thechild" Archives of General Psychiatry , 32, 466-476.
Parent-child separation: Psychologicalchildren. Journal of Child Psychology
12, 233-256"
Speer, I. F. ( 1970) .morphogenesis, orProcess, 9, 259-278"
Family systems:'Is homeostasis
morphostasis andenough?'" Family
Stabenau, J. R., Tupin, J., Werner, M., and( f965 ) . A comparative study ofschizophrenics, socÍopaths and normals"45-59
Pol1in, W"fani Ii es
Ps chiat
A.of
28,
Stierlin, J"runaways.
(re73).Archives
Aof
on29,
family perspectiveGeneral Psychiatry,
adolescent56-62.
Stierlín, H. (tOl+¡ " Separating parents & adolescents. NewYork: QuadrangLe/ The New York Times Book Co.
Stierlin, H., Levi , L. D" and Savard, R. J.Parental perceptions of separating children.Practice, I0, 4LI-427 "
(1e7r)"Fami ly
Stierlin, H., Levi , L. D" and Savard, R. J. ( fOZa¡ .Centrifugal vs. centripetal separation in adolescence:Two patterns and some of their implications. In S.Feinstein and P. Giovacchine (Eds. ) , Annals of theAmerican Societ for Adolescent Ps hiat , 2, 2LL-239"
Thurstone, L.L. (1947). Multiple factor analysis. Ctricago:Universit,y of Chicago Press.
Von Bertalanffy, L. ( I969) . General systems theory Anoverview. In w" Gray, F.S. Duhl and N"D Fti-zzo (Eds. ),General systems theory 6. pgy . Boston: LittIeErown.-
lvertheim, E. (rgzs)"and typology of361-376
"
unit therapy .andsystems. Fami I
IOB
the scienceProcess , 12,
Familyfami ly
Werthein, E. (IOZS¡" The science and typologysystems II " Further theoretical andconsiderations. Family Process, 14, 285-308"
of familypract ical
Wynne, L" C., , Ryckoff, I" M",(fSSS1 " Pseudomutuality in
Daf, J" and Hirsch, Sthe family relations
2r, 205-220"
L"of
schi zophrenics " Psychiatry,
Zola, I.K. ( feee¡ " Culture and symptoms AnAmerican
analysis ofSoc i olog ica Ipatients' presenting complaints "Revíew, 3I, 615-630"
r09
Appendix A
On the following pages we are going to ask you to answer
some questions. Some are about your family; some are about
yourself. The reason we are asking you to do this is to
find out how members of differ'ent families get along with
each ot.her. We'd like to f ind out whether dif f erent kinds
of families produce different kinds of individuals" Since
some of these individr¡als experience problems, $Je' re doing
this in order to See how general these problems are among
people. We hope this information witl help us to find the
best way to help people in families with problems.
This survey has four parts to it". Two have been
combined in one questionnaire so some of the questions may
appear to be similar, if not the same. Be sure to ansþJer
all the questions, even if they seem to be the same as other
questions. We would appreciate your cooperation in filling
this out the best You can"
l{e are passing out a sheet on which we are asking you to
put your name and also a code name, whích you will use on
your questionnaires. The reasons for using the code name
are So no one you don't want to wi11 have access to your
results, and also, in a monèh or sor bre will be asking you
to filf out these questionnaires again. This is so we can
see how similar the answers are from one time to another,
110
because '¡¡e would like to better understand these
questionnaires. If people answer a question differently at
different times, then the question is may be no good and we
need to know this so we can get rid of it.
we would be glad to come back and talk to you about the
general findings of the study. Although we are more
interested in group results, if any of you are interested in
your scores, we wilI also be glad to talk to you about
them. These scores will not be used individually and no one
will be able to find out about your individual scores unless
you want them to. If you do, check the box at the bottom of
thi s sheet "
Yes, ï
give my
1"
want others to know my scores- The people you may
scores to are:
2"
3"
4
5
6
Your code name:
111
This questionnaire concerns your feelings and attitudes
about yoü, and your family" Read each statement carefutly.
The Statements are answered with the terms "Always",
"Usua1ly", "Frequently" , "Sometimes", "occasionally",
"Rarely", Or "Never". When yOU have decided which answer
best fit,s the statement, underline the word that most
accurately describes your situation"
Example:
My famíly thinks I'm "tTte black sheep".
Always Usually Frequentlv Sometimes Occasionally Rarely Never
The.statements begín on the next page. Answer all the statements,
underlining the word which best fits the stat,ement for you. For
example the word "Frequently" is underlined above.
Thank you for your cooperation"
Check here if you would like your institution to
know your scores.
LT2
I lfhen I feel down, I go out and do something about it.
AIways
?" My
Usually
family has
RareIy
of others "
NeverFrequently Sometimes Occas ionally
values/attitudes similar to those
Always UsuaIIY Frequen tly Sometimes Occasiona IIy Rarely Never
3 I value members of my family more than I value anyone else"
Always UsualIY Frequent y Sometimes Occas ionally RareIy
enjoyable "
Never
9" Times my whole family spends with outsiders are
Always Usually
5 " In order to
Frequently Sometimes Occasionally RarelY
out "
Never
have fun I have to go
Always Usually
6 " People in my
Frequent Y
family are
Sometimes
up-front "
Occasionally RareIy Never
Always Usually
7" I hate being away from home.
Frequently Some times Occasional ly Rarely Never
Always
g We
Usually
do things
Frequently
together as a family.
Some times OccasionallY Rare1y Never
A1ways Usually
2" Going out is
Frequen tIy Sometimes Occasionally Rare ly Never
an escape from the disorder in my family.
Always Usually
LO" In my family
Frequently
it's clear
Sometimes
who has the most power.
Occasiona Ity RareÌy Never
Always Usually FrequentlY Some times OccasLona
Note¡ Entropy items are underlined.
IIy Rarely Never
II3
11. It's hard to trust Your familY.
AIways
L2" When
Usually Frequent y Sometimes Occas ionalJ-y Rarely Never
the chips are down I can count on my familY"
AIways Usually
t3 " It is easiermy family"
Frequently
to discuss
Sometimes
emotional
Occas ionally
problems with
Never
outside
RareIy
people
Always UsuaIIy Frequent ly Sometimes occas ionalJ-y RareJ-y
make
Never
childL4" Talking andbehave.
understanding is the best way to a
AIways
l5 " You
Usually
have to
Frequent v
scare a kid
Sometimes
to make him
Occasionally RareIY
behave "
Never
A].ways
16" My
UsuaIIy Frequent y Sometimes Occas ionally RareIy Never
family views me as a pain in the neck"
Always UsuallY FrequentlY Some times Occas j-onally
only be talked
RareIy Never
about with ny17" I believe personal problems shouldf amíly"
AIways Usually FrequentlY
18" I daydream about being
Sometimes
independent
Occas iondll-y
of my family"
Rarely Never
AÌways
19" rrUsually Frequently Some
seems to me my family witl
times Occaslonally
stay together"
RareIy Never
AIways Usually FrequentlY Sometimes occasionallY Rarely Never
lr4
20- People in the family have'the same opinion on things.
Always
2L" You
UsuaIly
can tell
Frequently
what people
Sometimes Occasionally Rare ly Never
in the family are thinking"
AIways
22" The
Usually Frequently
rules in the family
Occas ionally RareIy NeverSomet imes
are clear"
AIways
23" My
UsuaIJ.y
family is
Frequently Sometimes Oceasionally Rarely Never
in an uproar"
Always Usually
24; People in my
Frequently Sometimes Occasionqlly Rare1y Never
family want to get away"
Always
25" The
Usually Frequently Sometimes Occasionally
sooner people leave home the better off they
RareIy
will be.
Never
Always UsuaIJ-y
26 " Things are
Frequently Sometimes Occasionally RareIy Never
peaeeful at home"
AIways
27" My
Usually
family is
Frequently Sometimes Occasionally
not very close "
Rarely Never
AIways
28" My
Usually Frequently Sometimes Occasionally Rarely Never
family reatty sticks together"
AIways Usually Frequently Sometimes Occasionally Rarely Never
farnily hates to change plans.29. My
Always Usually Frequently Sometimes Occasionally RareIy Never
115
30" Our famlly tries to keep things calm"
AIways UsuaIIy
31" People in my
Frequently
family have
Sometimes Occasionally
the same opinions on most
Rarely
things.
Never
Always UsuaIIy
32 " lVithout rules
Frequently
my family
Sometimes Occasionally
is disorganized"
Rarely Never
Always Usually
33" Arguments are
Frequently Sometimes Occasionally
unending in my family"
Rarely Never
Always Usua-tly
34" What peoplefamily thinks "
Frequently Sometimes
I know think of me is
Occas ionally
more important
Rarely Never
than what my
AIways
3ã" Mv
Usually Frequently
family always sticks
Occas i ona ll-y Rarely NeverSometimes
to its plans no matter what"
AIways
36" r
UsuaIIy
can count
Frequently Sometimes Occasionally Rarely
on my friends when I can't count on my family.
Never
Always Usually Frequently Sometimes Occasionally Rarely
37 " People outside my family or rnyself províde for mostemotíonal needs.
Never
of my
Always Usually Frequently
3_8" People apologize after
Occasionally RareIy
my family.
Sometimes Never
arguments in
Always Usually Frequently Sometimes Occasionally Rarely Never
tr6
in my family seem to disagree in their opinions on mostæ. Peoplethings "
,cl-hrays
40" My
u sua lly Frequent Iy
family provides for
Sometimes
my emotional
Occasrona lly Rarely Never
needs.
Always Usually
4L" Tf I get inconcerned.
Sometimes Occasiona tly Rarely Never
my famity is
Frequently
trouble, I 'm on my own as far AS
AIways
42" My
Usually Frequently Sometimes OccasionaÌIy
family views me as incompetent"
Rarely Never
A1ways Usually Frequently
43" I think children shouldposs ible "
Occas iona 1ly Rarely NeverSometimes
live with their families as long as
Always UsuaIly
44 " Opportunity
Frequently Sometimes Occasionally
lies in staying close with one's ohrn
Rarely
fami ly "
Never
Always UsualIy Frequently
1Þ" People in my family do
Sometimes Occasiona lly Rare1y Never
things to bug each other.
Always
46" To
Usually
succeed,
FrequentJ-y Sometimes Occasionally Rarely Never
one has to rely on oneself.
A1ways Usually
47 " From outside
Frequently Sometimes Occasionally
my family would look orderly.
RareIy Never
Always UsuaIIy Frequently Sometimes Occasionally Rarely Never
LL7
48" I don't let my family know what I am feeling"
AIways UsuaIIy Frequently
49 " I guard my feelings from
Sometimes Occasionally RareIy
people outside my family.
Never
ÃIwats Usually Frequently Sometimes Occasionally RareIy Never
50. My home is a shelter from the chaotic world"
AIways
51. My
Usually Frequently Sometimes
family sees me as very much like
Occasionally Rarely
them "
Never
Always UsuaIIy
52" Members of
Frequently Sometimes Occasionally RareIy
my family don't care about each other"
Never
Always
53" rrUsualLy
is easier
Frequently
to discuss
Sometimes Occasionally Rarely
emotional problens with my family.
Never
AIways
54" In
UsuaIIy
my family
Frequently Sometimes Occasionally RareIy
we care about each other's feelings.
Never
Always
55 " Wtrenpasses "
UsuaIIy
I feel
Frequently'down, the
Sometimes
best thing
Occasionally
to do is to
Rarely Never
wait until it
Always
56" You
Usually Frequently Sometimes Occasibnally Rarely Never
can't trust people outside my famíIy"
Always Usually Frequently Sornetimes Occasionally RareIy Never
2L. PeopIe in the family consult others they are making decisionsabout "
Always Usually Frequently Sometimes Occasionally Rarely Never
118
58" HolÍdays are for the fanily to be together"
Always Usually Frequently Sometimes Occasionally Rarely
me is more important t,han what59" What my familyI know think"
thinks of
Never
people
AIways
60" My
Usually Frequently Sometimes Occasionally Rarely
friends are more reliâble than my family.
Never
Always
61" InUsually Frequently Sometímes Occasionally RareIy
my familli we try to keep time for each other.
Never
AIways
ç.2" r
Usually Frequently Sometimes
am much like other members of my
Occasionally RareIy
fami ly.
Never
AIways
63" In
Usually Frequently Sometimes Occasionally Rarely ñèver
my home outsiders make social visits.
Always
64" Sex
Usually
is best
Frequently Sometimes Occasionally Rarely
when you have no emotional ties.
Never
AIways
qq" Mv
UsuaIIy
family is
Frequently Sometimes Occasionally Rarely
afraid to be seen as different by others"
Never
Always UsuaIly Frequently
get to know
Sometimes Occasionally Rarely Never
someone very well before you have sex66" You shouldwíth them"
Always
67 " IrUsuaIIy Frequently Sometimes Occasionally Rarely Never
is easy to meet new people"
AIways Usually Frequently Sometimes Occasionally Rarely Never
68 " tvi rh
119
respect to variety, my family would be considered boring"
AIways
69" rthink "
Usually Frequently Sometimes Occasionally
have taken non-prescription drugs to change
Rarely Never
how I feel and
Always
70. IUsuaIIy Frequently Sometimes Occasionally Rarely
ny life"
Never
feel frustrated and angry about
AIways
7!" Mv
Usually Frequently Sometimes Occasionally Rarely Nevei
family has a good time when we're together.
AIways Usually
72" I can telt
Frequently SometÍmes
what makes people in my
Occasionally Rarely
family angry"
Never
AIways UsuaIIy
73" T feel down
Frequently Sometimes Occasionally Rarely Never
about my tife"
AIways Usually
74" People in my
Frequently Sometimes Occasional-ly Rarely Never
family express their anger openly"
AIways
75" InUsuaI Iy
answering
-b'requently Sometimes
the above questions I
Occasionally Rarely Never
was thinking of my mother"
Always
76" In
Usually
answering
Frequently Sonetimes
the above questions I
OccasionalJ-y Rarely Never
was thinking of my father.
AIways Usually
answering
Frequently
the above
Sometimes
ques t i ons
Occasionally Rarely
I was thinking of my
Never
whol-e77" Infami ly
"
Always UsuaIIy Frequently Sometimes Occasionally Rarely Never
L20
78" The one with the most poþter in my(Describe your relat,ionship with this
79 " Some of my actionsschizophrenic "
family is myperson).
have been considered by some to be
Always UsuaIIy
80. Some of mydelinquent.
Sometimes Occasionalry Rarely Never
some to be
Frequently
act ions have been considered by
Always Usually
81. Some of theschi zophrenic.
Sometimes Occasionally
have engaged in I
Rarely Never
consider to be
Frequently
actions I
Always UsuaIly
82" Some of thedelinquent "
Frequently
actions I
Somet imes Occasionally Rare1y Never
have engaged in I consider to be
Always Usually FrequentlY Somet imes Occasionally Rarely Never
Parents' Copy"
I hereby consent to allow
to participate in a study toconducted by graduate students andof Manitoba, Psychology Department"
L2L
(Name of Child)test this questionnaire,faculty of the University
Date:
ChÍldren's Copy"
ro(Name)
in a study to test this
students and facultyPsychology Department "
Date:
Signature:(Parent or Guardian)
Witness:.
, hereby agree to participatequestionnaire, conducted by graduate
of the University of Manitoba,
Signature:
Wítness:
( Name )
L22
Appendix B
On the following pages you will find a series of statementswhich a person might use to describe himself/herself. Readeach statement and decide whether or not it describes you"Then indicate youf answer by circling either rtTrr or "F""
If you agree with a statement or decide that it doesdescríbe yoü, answer TRUE. If you disagree with a statementor feel that it is not descriptive of you, answer FALSE"
ANSWER EVERY STATEMENT either t,rue or false, even if you arenot completely sure of your answer"
1" ï always try to beof my friends.
considerate of the feelings
In the long run humanity will owe a lot moreto the teacher than to the salesman"
I am seldom ill"
My memory is as good as other people's"
Nothíng that happens to me makes much differenceone way or the other"
I have a number of health problems.
Most of my teachers were helpful "
I often have the feeling I am doing something evil"
I almost always feel sleepy and Lazy"
I am not willing to give up my own privacy orpleasure in order to help other people.
My life is full of interesting activities.
We ought to let the rest of the world solve theirown problems and just look after ourselves.
I often question whether life is worthwhile"
I am able to make correct decisions on difficultquestíons.
TF
TF
TF
TF
TF
TF
TF
TF
TF
TF
TF
TF
TF
2
3
4
5
6
L2"
13 "
L4"
r5"
7"
B"
o
10"
tl"
I often take some responsibility for lookÍng out T Ffor neeJcomers in a group.
TF
16"
L7"
18.
t9"
24.
L23
Rarely, if ever, has the sight of food made me ifl. T F
I believe people teII lies any time it is to T Ftheir advantage"
I find it very difficult to concentrate" T F
Many things make me feel uneasy T F
ï am always prepared to do what is expected of me" T F
APPENDIX C L24
ENTROPY SCALE FREQUENCY HISTOGRAMS
VALUE ONE SYMBOL EQUALS APPROXTMATELY 4"OO OCCURRENCES
ITEM 2
COUNT
5
t1867
63
t6l0
2
I2B
107
4B
53
22
19
3
r"002 "OO
3"004"005"006.007.00
"oo1.002"003"004.005"006"007 "OO
***********************************************Je
** * ******* ** ************
ï ï ï40 80 120
HISTOGRAM FREQUENCY
**** * ****************************** ******** ******** * * ******************
I
I0
ï160 200
ï0
VALTD CASES 281 MTSSING CASES 0
ITEM 4
COUNT VALUE ONE SYMBOL ESUALS APPROXTMATELY 4.OO OCCURRENCES
ï ï I T ï40 B0 L20
HTSTOGRAM FREQUENCY
VATTD CASES zBL MTSSING CASES 0
r60 200
TTEM 6
COUNT
3
37
65
58
69
26
2L
2
VALUE
.001.002 "003.004.005.006.007"00
ï0 15
HISTOGRAM3 45 60 7\
REQUENCY
VALTD CASES 28I MISSING CASES 0
VALUE ONE SYMBOL EQUALS APPROXTMATELY 1.50 OCCURRENCES
********************************************************************************************************************************************** *
*********************************************
1"002"003.004.005.006. 00
7.00
L25
ONE SYMBOL EQUALS APPROXTMATELY 1.50 OCCURRENCES
****************************************************************************************************************************.******************* ***********************************
ï0F
I ï...., T
TTEM 8
COUNT
13
30
51
70
50
6l6
ï0
]STOGRAM
I5H
,ï30
I ï ï1
FREQUENCY
0VALID CASES 281 MISSING CASES
45 60 75
ITEM 9
COUNT
4L
78
61
47
2L
20
I3
TTEM 1O
1
59
67
28
51
29
34
L2
VALUE
1.002 .003 " 00
4.00s " 99
6"007.00
L26
ONE SYMBOL EQUALS APPROXTMATELY 1.50 OCCURRENCES
***********************************************************************************************************************************************
ï
ï0
.ï20
ï0F
T ï ï0 4 60
HISTOGRAM REQUENCY
VALTD CASES 281 MISSING CASES 0
COUNT VALUE ONE SYMBOL EQUALS APPROXTMATELY 1.50 OCCURRENCES
*
*******************************************************************************************************************************************************************************************
.00t. 00
2 .043"004.00s"006.007.00
80
T
100
Tï I ït5 30
HISTOGRAM FREQUENCY
VALTD CASES 2BI MISSING CASES 0
45 60 75
TTEM 20
COUNT
43
59
90
56
27
6
I65
88
77
35
6
2
VALUE
2"003.004.005"006.007.00
1.002 "003 " 00
4"00s.006.007.00
L27
ONE SYMBOL EQUALS APPROXTMATELY 2.OO OCCURRENCES
**.**** ***** ********** *
************************************************************************************************************************
ï0
.ï.". .r..."....,720 40 60
HÏSTOGRAM FREQUENCY
ï80
T
60
100ï
T
VALTD CASES zBL MTSSING CASES 0
ONE SYMBOL EQUALS APPROXIMATELY 2.OO OCCURRENCES
****************************************************************************** ******************************************
ITEM 2L
COUNT VALUE
****
ï0
,I15
,T30
ï45 75
HÏSTOGRAM FREQUENCY
0VALTD CÃ.SES 281 MISSING CASES
L2B
VATUE ONE SYMBOL EQUALS APPROXIMATELY 4 " OO OCCURRENCES
TTEM 22
COUNT
I61
118
37
31
20
L2
I
I34
LL256
42
25
7
3
"001.002"003.004"005.006 " 00
7.00
"001.002 .003.004.005.006.007"00
**********************************************************************
ï0
ï0
ïBO
, 'ÏL204
ï40 80 r20
HTSTOGRAM FRESUENCY
ï160
,,r,160
20n
..ï200
T
HISTOGRAM FREQUENCY
VALTD CASES 281 MTSSING CASES 0
TTEM 23
couNT VALUE ONE SYMBOL EQUALS APPROXIMATELY 4.00 OCCURRENCES
************************************ *
**********************************
ï ï,",,, T
0
VALTD CASES 28O MTSSTNG CASES 1
TTEM 26
COUNT
19
r2550
49
19
16
3
Ï TEM
COUNT
6
55
29
101
54
30
6
29
VALUE
1"002 .003 " 00
4 " 00
5.006.007.00
VALUE
r.002"003 " 00
4"00
L29
ONE SYMBOT EQUALS APPROXTMATELY 4. OO OCCURRENCES
*********************rk** ***** *******
***********************************
I0H
I0
ï0F
..ï120
I160
T
2Dn4 IISTOGRAM REQUENCY
0VALID CASES zBL MISSTNG CASES
ONE SYMBOL EQUALS APPROXIMATELY 4 " OO OCCURRENCES
5.00
**************************************************************
**6".00
7"00r..... " "..ï0 40 r20
REQUENCYB
I0F
ï..., T
160,.ï20n
HTSTOGRAM
VALTD CASES 28I MISSING CASES 0
r30
ITEM 30
COUNT
22
107
62
43
22
2I3
TTEM 32
COUNT
2L
97
57
63
20
20
2
VALUE
1.002"003.004. 00
5.006.007"00
1"002.003.004"005.006"007 " 00
ONE SYMBOL EQUALS APPROXTMATELY .4
" OO OCCURRENCES
ï0
*************************************************
************
***********
.ï" "
40I0F
,,ÏL20
I160
ï200I
HISTOGRAM REQUENCY
IVALTD CASES 28O MTSSTNG CASES
VALUE ONE SYMBOL EQUALS APPROXIMATELY 2. OO OCCURRENCES
**********************************************************************************
ï ï00 2HÏSTOGRAM
ï0F
I4 69
T
80ï
100REQUENCY
IVALID CASES 28O MISSTNG CASES
13r
VALUE ONE SYMBOL EQUALS APPROXTMATELY 4 " OO OCCURRENCES
TTEM 33
COUNT
37
103
61
31
20
18.10
TTEM 35
5
48
42
68
61
49
7
1"002.003.004 " 00
s.006"007 " 00
1.002.003.004"005.006.007.00
*********************************#*
************************************
I0
r.. " ". ".. ï0F
T ï160
I20040
HTSTOGRAMI20
REQUENCY
I45
REQUENCY
1
I
VALTD CASES 280 MISSING CÀSES I
COUNT VALUE ONE SYMBOL EQUALS APPROXTMATELY 1 " 5O OCCURRENCES
******************************************************************************************************************************************************************************************
,ï15
HISTOGRAM
I0
I0F
ï ï3 60 75
VALTD CASES 280 MISSING CASES
L32
VALUE ONE SYMBOL EQUALS APPROXIMATETY 4.00 OCCURRENCES
TTEM 38
COUNT
I29
60
33
49
35
54
t9
ITEM 39
COUNT
3
L2
22
66
97
76
5
.001.002.003"004 " 00
5"006.007.00
*
*****************************************************************Ji********
* ** *** * Tk * * * * rt * rk* * * * * * * * *** **
***********************************************************************************************************************************
ï0
ï2
T ï ï
ï
T".1 24 36
FREQUENCY48 60
HISTOGRAM
VALTD CASES 280 MISSTNG CASES I
VALUE
1"002"003"004"005"006.007"00
ONE SYMBOL EQUALS APPROXIMATELY 1.50 OCCURRENCES
**********************************************************************************************************************************************
ï ï ï ï ï0 20 40
HTSTOGRAM FREQUENCY
VALTD CASES 28I MTSSTNG CASES 0
60 80 100
TTEM 45
COUNT VALUE
133
ONE SYMBOL EQUALS APPROXTMATELY 1.50 OCCURRENCES
*************************************************************ìb******************************************* *
********************************************************************************
2
23
65
7I69
28
15
I
I39
r2744
31
13
1B
I
.001" 00
2.003.004.005.006"007.00
" 00
1.002.003. CIo
4.00s.006"007.00
I0
ï
ï5
T Tï. "
T
IHÏSTOGRAM
30 45FREQUENCY
60 75
VALTD CASES 281 MTSSING CASES 0
TTEM 47
COUNT VALUE ONE SYMBOL EQUALS APPROXTMATELY 4. OO OCCURRENCES
***********************************************************************
ï0
ï,, r0 120FREQUENCY
T
0 4HÏSTOGRAM
VALID CASES 28T MISSING CASES
I
0
',Ï160 200
ITEM 50
COUNT
2
2T
33
44
79
53
36
13
TTEM 52
COUNT
I151
B1
19
19
4
3
3
VALUE
.001.002 .003.004.00s. 00
6.007.00
134
ONE SYMBOT EQUALS APPROXIMATELY 2. OO OCCURRENCES
*
**********************************************************************************************************************************************
ï0 2 40 60
]STOGRAM FREQUENCY80
VATID CASES 28L MTSSING CASES 0
VALUE ONE SYMBOL EQUATS APPROXTMATELY 4. OO OCCURRENCES
T
0H
ï..... f
100I T
"001.002 .403.004.005.006.007"00
***********************************************************************
ï0
ï ï ï ,ï20n40 80 r20
HISTOGRAM FREQUENCY
VALTD CASES zBL MTSSTNG CASES 0
..ïI60
TTEM 54
COUNT.
2
77
86
39
41
r9t3
4
VALUE
.001.002 "003"004.005.006.007.00
135
ONE SYMBOL EQUALS APPROXTMATETY 2. OO OCCURRENCES
*
**********************************************************************************************************************************************
ï ïï ï0F
ï0
I1000 80
VALTD CASES 28T MTSSING CASES 0
VALUE
.001" 00
2.003.004. 00
5.006 " 00
7"00
ONE SYMBOL EQUALS APPROXTMATELY 2. OO OCCURRENCES
*
*************************************************************************** *
******************************************************************
204HTSTOGRAM
6
REQUENCY
TTEM 57
COUNT
I25
79
46
67
29
2B
6
ï0 20
HISTOGRAM
ï...".... ï T
REQUENCY
0VALTD CASES 2BT MISSING CASES
4ï0F
60 80T
100
ITEM 58
COUNT
43
87
38
74
25
11
3
VALUE
1"002"003.004"005"006"007"00
r36
ONE SYMBOT EQUALS APPROXTMATELY 2. OO OCCURRENCES
******************************************************************************************************************************************
I .I20
,ï40
ï T I1000 60 80
HTSTOGRAM FREQUENCY
0VALTD CASES 28T MISSING CASES
COUNT VALUE ONE SYMBOL EQUALS APPROXIMATELY 1.50 OCCURRENCES
*
**********************************************************************t*********************************************
*********************************************************************
TTEM 61
)
L7
47
37
74
51
47
6
.001.002.003.004 " 00
5.006.007. 00
ï ï.. .
5HTSTOGRAM
I30
I ï ï0 1
FREQUENCY
0VALTD CASES 2BL MTSSTNG CASES
45 60 75
TTEM 62
COUNT
2
16
57
59
68
40
28
11
ITEM 63
VALUE
.001.002 " 00
3.004. 00
5"006"007"00
.001.002.003.004.00s.006.007. 00
137
ONE SYMBOL EQUALS APPROXTMATETY 1 " 5O OCCURRENCES
*******************************************************************************************************************************************************************************************
T I5
I ï I ï0 30 45
FREQUENCY60
HÏSTOGRAM
VALID CASES 28L MISSTNG CASES 0
COUNT VALUE ONE SYMBOL EQUALS APPROXTMATELY 1.50 OCCURRENCES
********************************************************************************************************************************************************************************************
1 75
3
14
20
38
6¿
7L
51
22
I0 15
HTSTOGRAM
ï.,,, ï. r30 45
FREQUENCY
T ï57
VAL]D CASES 28L MISSTNG CASES 0
60
TTEM 65
COUNT VALUE
r3g
ONE SYMBOL EQUALS APPROXTMATELY 2. OO OCCURRENCES
***********************************************************************************************************************************************
5
18
2La1¿¿
3B
33
97
47
.00r " 00
2 "003.004.00s.006.007 " 00
.001.002 "A03.004.00s.006.007.00
ï ï ï40 60
FREQUENCY80
*
********************** ** * * *** *** ** * * * * * * * * *.* * ************************************************************************************************
I I ,I I In 20 40 60 80
HISTOGRAM FREQUENCY
2BT MTSSTNG CASES O
I0
ï T
0 2HISTOGRAM
VALTD CASES 28T MTSSING CASES 0
TTEM 68
COUNT VALUE ONE SYMBOT EQUALS APPROXTMATELY 2.50 OCCURRENCES
100
2
3
16
22
49
6L
76
52
I
VALÏD CASES
100
r39
VALUE ONE SYMBOT EQUALS APPROXTMATELY 4. OO OCCURRENCES
ITEM 7L
COUNT
2
36
107
53
42
23
t62
.00r" " 00
2 "003.004.005.006.007 " 00
*
*********************************** *
***********************************
ï ï0
ïïBO I2O
FREQUENCY
I160
ï20t0 4
HÏSTOGRAM
VALTD CASES 28T MTSSTNG CASES 0
.001.002 " 00
3.004.005.006.007.00
ONE SYMBOL EQUALS APPROXIMATELY 4. OO OCCURRENCES
********************************************* *
**************************
ï..." T.".....,,ï, r0 40 80 120
HÏSTOGRAM FREQUENCY
ITEM 72
COUNT
2
46
131
35
4T
16
I2
VALUE
I20n
VATID CASES 28T MTSSTNG CASES 0
."ï160
TTEM 7 4
COUNT
16
31
87
54
53
27
VALUE
.001.002 " 00
3.004.00s.006"00
140
ONE SYMBOT EQUATS APPROXTMATETY 2 " OO OCCURRENCES
*********************************************************************************************************************************************
r""" .ï,", .ï.....""".r0204060
HTSTOGRAM FREQUENCY
ï80
..ïr00
VALTD CASES 275 MTSSING CASES 6
APPENDTX D
ROTATED FACTOR MATRIX, VARIMAX ROTATION
Factor l- Fact.or 2 Factor 3 Factor 4 Factor 5 Factor 6Item No"
6lB
7T38503054
4582A57
3323
9265¿45
72632I65
.7 5897
.68634" 66504. 5r8r_ 2.49485.466LA.4660r.452LA.42439.4L460.28394.30r63.27054.30966.5LO7 4
-. 01067" 23503
.07510- "17 407
.t9445- "16278
"r2285.L7 3t_9
-.0r300.0693r
-. 07559.37 466
" 18395"IO944" 36138.2L499. I 5486" 42038.30905.o8222" 08037. IBB9O" 19650
.670r 3
.66688
.58653
.54L57
.36772
.34795
.r2250- . 09239
.o4a7 4- .25652
.2r868
.l-2255
-"02072.05645
-.12236. r1908
.01134"09259" l_ 10r_ 5
" 08469. OOBBB.07465.r9764.09207"20534.L7 448.06468. 06633" 09514
-.00309.07797. I 5405.194L2.7L556
- .47 22V.44IO9
-. 3B8r 7
.18273
. 07 22L
.a2246
.00840
.2LL65
.o6602
-. o2L87
.oL692
.1 52Br
. 05562
-.0182r_. 05960" l_6839.L4287. 06100" 09109.39061.08305. 0991 3.TI7T6.L7032.L4332.l-797 4.o5629" 20608.03182.25489
" I 5BBO- "1,5467
.02430-.2AA2B
.77 258
.36427
.a2614
. oooo4
-. 0565r.LO27B
-.00990-.058t2
.L6094-.03041_
-.aL377.roo22. ott63
-.06692.02255.r70r9
-. 04424.03518.02051.06433" oB3B3
.oo47r
. o9823
.02897-.01975-.24077-.00975-.aI712'.l-6922
. r7895"3324L.oo426"0257 4
.7L64r
.48344
-. a3947.03034
"oL232.46526
"r2944. 05596
-.01663" or614. rBB06.16375.0323r
-.r3782.o977L.0208r
-. oL796.05086.22925
-.ooa24-. 0Br 0B
" 14316" 11806
- " 03492-.04380
.r637L" 0576r_.1BBO4
-.03369" 02626
- "oL296-. o429l-
" 03037
" 66399.49183
" 1601_B
-.08173
" 2L943
-.15242
5247
2935
746
2
10
22
68
.L4782 .r9058-.06276 -.11683
" 23382 . 08018
-.28243 -.L7766 HÀts
APPENDÏX E
ENTROPY SCALE FACTOR LOADTNGS, OBLTMTN ROTATïON, PATTERN MATRTX
It.em No" Factor I Factor 2 Factor 3 Factor 4 Factor 5 FacLor 6 Communalit.y
6l_I
50VL3858623020
4
3323
9263239
v46
2L
293565
1022
68
257
"75924.62689.5s440.51-823.43483.35986.3s850.3327 2.32324" 29683
" a4709- .007 42
.08393
.32257- "a6277- "24246-.140s6
.2527 6
.r2663- "0927A-.00444- "o7 255
.02854
. r6036
-.13389- " 05135
.092r0
.03125-.07884-.L2427
.2r063" 09164.03 49 4
- "02498" 061_83
-"0911_r-.19010-"08598
.o4647-.21940-.31016-.L2639
.LL7 44
- .65447-.54639-.54284-.43550-.4I361
.3877 L
.07 693-.01544-.0t132
.03545- "03737
" 239 44
" 04 711-. û2018
"071-l_5
- "09242- " 05043
-.0030r-. Ir0lt- " 03694-.05243
.00511_-.24262
- "04226-.00294
" 00 567
" 16415"14606.00366
-.0I 65s-. I 5889
.0374r-.04773-.05628- .11 018
" 0877s"0827 4
"oL757"09916"70665.4s433.30531
- "02082.01 736
-. t 2318
-"07390" 2r1 65
-.09r78"a657 2"L7 97 4
.0r58r" 33193
-.0087r.0sI63
-"04520- "03729
- . 041_ 70.o9822
-.03562.03652
-.06486.02551.03249"r77 32
"L2466"03170.07334. r_ 4523.06334.00821
-.r7398-"03338-.04246
. 003 66
. r8938
.75400" 4BI2A" 2647 s
"o2557.08651
"aL629- "02696
,07815.L27 7 A
" 03235
.04137-.04826
"000r4.o2209
- "05822- " 001_03- "Ls707
.o2435"1651r"0s603.02997
-.07528" 24667"L3236
- " 00952-.00042
.03935-.00s40
.09r98-.160s3- "0L7 67
. 0 6700
.00248
"02375-.06037
"03570-.58205- " 48549
" 0r 435
.0101_7"04432"47 L9 4
- " 14074
.09341
.0r675-"0s008-.06323
.06393"t7 862.0086s" 26833
-. 04 538.00123.L5477
-.08054-.0s705
.r9907
"Lr57 2. r 3209.08556.02420
-.04096.00132.L4L7 2
-"01234-.18738
.02588- "0257 2
- " 02380
-"0r348.o37 32
-.68635-.03874
"L269L-"06062-"r_0697
.r2950
. 0r 208- " 22888
.07 286
.6189
.5447" 32L4"6985" 4052.240L" 42L2.4796" 4053" 4142
.6339"5724.5136.6300.2357.4005.5330" 4439"2848"547 5
"2544"4770"3648"438s.5564
"37 45.2692
"6250" 5416
.5706" 6004"2698"2963
6372
52544745
FÞ¡\)
143APPENDIX F
SUMMARY STATISTICS
RESPONSES OF SUBJECTS TO JACKSONIS PRF
SOCTAL DESIRABILITY SCALE
FREQUENCY PERCENTVALID
PERCENTTOTAL PRF
SCORE
5.008.009 " 00
10.00I1.0012"0013"0014.00r5.00I6.0017"0018 " 0019"0020 " 00
TOTAL
IT
49I
t1161825204038222L47
"4.4
L"73.83"44"76.87"7
10 " 78.5I7"IL6,29.49.0
4442I9749I25I57
1.3.2"3.5.6.8"7.
r4.13.7.7"
16"
CUMPERCENT
L4.2r.29"39.48.65.81.91.
100 .
1
6"9"
4964I54I734600
281
MI.SS ING
100"0 100.0
Standard Deviation
2 "962
Mean
r5 " 880
APPENDIX G
CORRELATTON MATRTX F'OR ENTROPY SCALE
9 l_0 2A 23, 22 23 26I
z46I9
IO20
ïtems
2L222326293032333538394547505254575B6I6263656B7T7274
2
" 6L52.L297" 0680" 3060
-. 081 5-. II09
.227L
.2008- " 1894
.LO67-" 0845
. r_004
.0600
.0073" o2B4.o602.1852" l-690.0035
-.L29L-.0676- " o73r
.1806
.2299
.3666
.0846
.3071-.o973
.2L67
.t238
.0388-.0056
. o555
4
-.l-392"61_13.L628.27 67"0220.01_50.397 9.t647.LIg4.L248.4928.1884.0553
- " 053r-" 0251-.1_630-.o783-"1_846-. o716.I454.1803.3523.27 5L. TB2B
" 3281-.0544
.37 04-.2831- .2A62- .2\37
.2339
.47 7A
.0729
6
- " l_l-80" l_o7r_
" 4443" 15r1. oo83
- " o9r-0.3019.37 52.2446
- " 0949" 3646
-"0r_16.0911
- " 0927-. o23 5
. r3B4
.r733-.l_552
.1r45"L7T6.l-862.2665. r_ 357"L37 2.0299.0014.r267.oo76
-. r 934-. 0078
.2L42
.32LT
.1846
-.L6L2.4947"l-320.6009.2666
- . o97L" 3640.22IL.1 751.284A.t626.2027"2824
-. oo8t"I3BB"oL26. OBB4
- .2602.IO94
- . 0594.4585.2262.27 20" 2BO3.4404.2848.3624
-.0463-. oo90-. 331 3
.3565
.3541-. 2111
- " 2IL2.240r" 0899.L282" 4400
- .267 0" 3103.28TO.2028.5006"4156. r683" 2183.1-648.27 47
- "2622-.1028- " 2089
.l-458
. 0347
.3928
.24L7
.3r08
.257 3
.37 02
.T228
.3220-.1625-.o6L2-.3258
.3098
.3638-.0646
- "a523- .4l-76-" o58B- " toBT-.286L
.5863-.264A
.0426
.L765- .4409-"0762
.064L" 14BO
-.3199- " 4435
" 5000.0703.2635
-.t618-. 0tB0=.0336- .27 48-.1342-.0582j. 2BO5
.o657-. r 566
.2202
.2930
.3460-. 1017-.1731_-.0123
"2L2I"IOI3.3556.32A9"r379
- .2599.5068"28L6
- "r42A.27 52.2168.0130" I9OB
-.oo22.38I4
-.l_048" 3BB2
" 0529-. 0830
.0924
.0591" 0435.324L"2375"1664. TB3B
"4702-.3108-. 1 606-.r154
.2407" 1r15
-. ooo8
" o24L.o92A"422A"2477" 0226. 0449"2963.5413"1812
- " 0966" 2954
-.0843.2063
-.oo47-" 0512-"0323
.2646" 0783" 3388.0796
- " 0598.1 233.0048. r830"o6L4. l_ 681
-"0215-" 3368-.L206-" 0539
.0323
.l-669" 3Br_ 3
-. o92L- "1258-"1_01_3
.1326
. o356
.2964-.2L1,L
" 0692.3930
-.1493.o44A.r529. IOB2
- " 2632-. 0790
" 1665.r953. oo67" 1106.307 2"TOI2. o254
-. oo93-"0772-" 0r60
" 2266-.045r-
" 2793-. o4B4
.2142" o235
-.1424-.3182
-,1-993" 0128
- " 0789- " 0987
.1485-.1985
.T2L2
.r073-.1 598
.3389
.0139
.0806-.1087
.2595
.2596-.3433-. 32Br
.0227- . 0648- " l- 2Bt--" 1552- "o375
.0870
.1069" 0792
- " 2060.156I
- " 0394. o5r2
-.1328-" 0052
"o452.07r5
- " 0367.3022.1345" 2857"T76L
-.2543" 2635" 1935
-. to02.3493.2339.2246.1733.r705.3366
- . 1919-. 0117
" o036. o5B0
-.11I 3. l_ 013.0922.2L24.3997.27 57
-.0428.2444
-.o7r2" IIB2
- .2220.27 37"1307.0536
HÞ,Þ
Iterns 29 30
-" 0493" 2243.030CI"r97 2
-.0310-.L241,
.1804"1454
-.t473.2530" 2554" ATL4" 44Il_
- " I018.2450. t252.0527
-"0097- " 2L36
.0301
.254I
.067r
.Is63" 2307.l-431.0375.I848.1091.r050
-.0598.23L6.L637
-. I s9I
32
-"2494-"1007-.i-715- " 236V
"t692- " 23t6-.I965- " 2822-.3083
.0038-. 0r 59- .267 2- "L929
" 4984-" 055i_-.3157-"I593
.a766-.0037-.0378-.0831
. 1511
.0520-"1205-.056r-" 2818- "0s44-.0412
.0494-.II7 2
- .0s27.02r2.0149
33 35 38
-.2861_-, 09 7r_
-.023r."L7 2A
-.2050- "L97 6
.0055- "057 2
- " 2624" 0837.139t.0099"I450.0220
- " 0I29-.0908
" 3129-"0170
,0067-" 0334
.04r3"0191.0606.0244.2406
-.I882"0977.2062.2264
-.0543-. 0tl 0
.07 29-.I427
39
- "0464-. 0570-"0792
"01_82.447 5
-.0082-"40v6
.01_08
. Is75-.2662.I379
- "L27 4
-.0496-.0418-"L777
.3011
.29L6" 0392"2552.l_89s.Ì618
-. 01 31-. i_874- .226L-.0930
.2628-.0565
.248L
.0036
.I823
.0956-.0218-.2137
45
-"1_17I" 3229"1163" 41,22.Ig 44
-" 3502" 2799.2077
-,0379.4398" 2043" 2023
-.0136.2088" 2842
-.29 40-" 0189
.0168
.3058
.0439
.0396
.2358" 3120" 4439" 3348
- " t71,2"27L7
-.2105" 1006
-. 4l_ s5.327 5.2002.0961
47
.L967-.I98s
.1566" 0234.227 9
" 0758.1344.2029,L27 2. ao24" 2369
- . 0111.3269.1816.1098
-.0039"I077
-. 0049.0875.4443.2678.02I 4
" 0860"I106. OII2.1363
-.0r89.0413,0200.0875"r760
-" 0653.0352
50 52
246I9
t02A2I2223262930323335383945475CI
52s4575B616263656B777274
" 2437- "L4A2
" zLL9"1285"I462"2727
-.0316.0082.l_401.0996.3396.3539" 3256
-. l7 s8.3056" 4345" 3606
-.0321.0836.2I95.1,629.0619.L737.0552.026L" 3202.1 552
-.0s68-.1,77 4
.L964
.2608
.0117- "I782
-.0060.rs79.3!27" 5l_ 70.36s6
-.3170.4314" 2963
- "0422.7050" 3423.3s78" 2357" 2524.6411
-. tI 7s.l_25I
-. 0570. L325
-.1934.0L62.19s0.3423.33I3" 3032.2124.3413
- " 2124-.0696-.3562
.3546
.3055
.0s76
-,1_366" 0149
- "2254-. 34s8- "L7 84
"2278-.4436-, 3081-"0015-.22I4- " 02L6
.I505-,0393-. 17 s8- "17 48
" 3463.0360
- " 2029- " 0924
.0935" 0986.0415
- "1,22A- " 3209
.0486-" 2330
.025r" 3002
-.0373" 3888
-" 07s0-"01r9- "2692
- " 1014" 2002" 0186" 3022"I543
-.0254.1093.07 52.0322"4076"1863.3205.368t.L232. t19r.0568" 0046
-.I226-.0165-.0161
" 4925. 179i-" 3l-I3"3779" ¿oro.0242" 4596. L592" 2342
- .2566" 3861" 2L57.
-.0656
-. I 685.2487" 026L
-" 0136.I085
-. 19 3r"L487.2653.2066.2235.0046.1710.0075.1349.0819
-. 150 2-.1624
" 0834-.1_194-.0279
.0191"L569.16T2.3190"LL7 2
- "L521" 0493
- " 0697" 0444
- " 2902"0759" 2048.r0s0
PÞ(Jl
ILems 54 57 61 62 63 65 7T t¿
-.I67 6.387 4. 1181"2797" 2260
-.3412.2423.4199.0800" 4045.2908.1618.2099.LL7 2. 1310
- .2554.1148.0790.1538.1900.261 4.2I75.3I86.4L99.40s5
-.0067.3836
-.1537-.0932-"2377
" 23r7.4203
-.0368
-.2526" 1580
- "2268-.0424- "L547-.1609
.0057-.0420-. 0t s9
.IO5B-"0786-.0324-.L662
.07 82-.0832-.2629-.297 2
.07 39-.0738
.04L2
.0772
.0078
. 18I0
.2126
.L77 6
-.2600.0988. tBlL.2008
-.068s.0146.IA77
- .0626
58
.0305" 2L6T
-.1607.207 3
- "07 43.0038.0215.0436.1510.1480
-.0340.10 35.0840
-.0370.II98.0591.r083
- .22LI.0467.0804.2217.I170" 2501.0413.6183.0I 42.2958.0520.0308
-.0245.0783" 1418
-.1933
- " 2295. I915" 0247.297 4
" 0308-.1443
.1049
.0893-.1635
.0088
.1350
.1639
.2892-.0433-. I069
.4237
.0319-.3029
.0s70-.0r32
.2018
.0448
.0190
.0845
.1 201-. 081 6
.1896
.1528
.11 71
.r270
.1503
.2190-.2I55
" t229.3493"I407" 52r l-.0607
-.2077"2768.202I
-. 0 071;0591.I340.0I22.L524
-. 0556.097r.0619.I506
- .02I 4.0080.0438.3434
-.0951.237 4" 4435.2847.37 64.4657
-. 04 58.1884
-.I988.3531,I097
- .037 9
"2464- " 28s7- "17L4-.2501- .07 22
.1 695- .296r-. 3004-.I99s-.237 5-. I52B-. 1161- .037 4
.1096-.0981
.2677
.0583
.07 37-.084r-.0803
.034r-.1s9 3
-.0653- .0442-.3231-,0833-.2209
.5663" 2L29.380Ì.0190
-.4323-" 0s0B
"2878-" 3465-. i.2 3l- .2487-.239 4
.2L24-. 3 6I3-. 3584-.2243-. 2831-.0985-.1546
.0 432
.0355-.L962
" 2662.I795.r137
-.0220.0453.r128
- . t457.0140
-.0s81-.1838-.0451-.1366
.3102
.I957
.3652-.0856-.293r-" 0560
68
" 3948-.1613
.0087-.01_75
. 114I
.2929-.0r43-"0r58
.0963-.0604-. 0 316-.067 4
"1773- .07 28
.0601
.0734
.r243
.0863
.0002
.2357
.2282-.1005
.0305
.0287- .0562
" 3616.0803.0247.L704,0954.1387
- .3254.0580
-" 30r4" 3579.L347" 2226.0670
- "LI97.2129.151I.0043.1 559.L449.2775.IL97
-.0019" 0343
-.0435-.0926-" 0983-.1455-"1064
.1103
.1364
.101I" 2347.0517
-.1_484.2404
-.0480" 4347.1,827" 2355.300s
- .057 2
-.23L4.047 6.44r8.t051" 3197.0227.134I.34r8" 4832.0794.4608.0438.2790.027r.0788
-.0112.3535. OTB2
" 2556" 4520.4592.299L.L457.2636" 2362.3659.1r21
-"4494- " 2283-" 2055
" 2854" 27r8"0724
74
-. I 553- .2452
.2327- .07 64
.2255
.0397- .0367
.L704" 2351
-.1378"L592
-.0399-.0r85
. 1410-. 0349
.047 5
.0321-. 0r 21
.2227- .0242
"0727" 0102
- .0297.0355.0348.00r9
-.1930.1366
-. 01 3l_
" 249r.047 6.1595.2157
246I9
l_0202I22232629303233353839454750525457586I6263656B7L7274
tsÈOr