Working Paper Series
FACULTY OF MANAGEMENT STUDIES)
UNIVERSITY OF )ORONTO) - -
I'STITUE OF INDUSTRIALRELATIONS LIBRARY
JUN 1 2 1973
UNIVERSITY OF CALIFORNIABERKELEY
if'
.N.EXPECTANCY THE9RY AS A PREDICTOR OF JOB PERFORMANCE,SATISFACTION AND MOTIVATION:
AN INTEGRATIVE MODEL AND A REVIEW OF LITERATURE
7Xy- Robert J. House atUniversity of Toronto
/* and 0
Mahmoud A. Wahba'//Baruch College of the City University of New York
(Working Paper 72-2l)
October 1972
Not to be quoted without the authors' permission.
ABSTRACT
The development of Expectancy theory in work and motivation is
described briefly. A model integrating these developments is proposed.
Fourteen studies are summarized in detail and evaluated, collectively,
with regard to the main constructs of the integrative model. The
distinction between two types of expectancies increases the predictive
power of the theory and the valences of intrinsic rewards are shown to be
more powerful predictors than the valences of all extrinsic rewards
except pay. Extrinsic reward has little effect on performance or satis-
faction unless multiplied by the corresponding expectancy. The implications
of these conclusions for practices are discussed. Further areas for
research are suggested.
Expectancy Theory as a Predictor of Job Performance,Satisfaction and Motivation:
An Integrative Model and a Review of Literature
Robert J. House* and Mahmoud A. Wahba**
October 1972
The purpose of this paper is threefold: First, to develop a
general model integrating various formulations of expectancy theory as
applied in work and motivation. Second, to review, collectively, the
overall findings of recent empirical research with reference to the major
constructs of the integrative model. Third, to suggest some areas in
need of further research in the theory.
Expectancy or instrumentality-valence theory has generated a
great deal of research as shown in Exhibit 1. The central concept of
expectancy theories is that the force on an individual to exert a specific
amount of effort is a function of (1) his expectations that the effort will
result in a specific outcome; and (2) the sum of the valences (personal
utilities or satisfactions) that he derives from the outcome. The theory
asserts that the function is a non-linear, monotonically increasing product
of expectations and valences (Vroom, 1964). Thus, according to this theory
of motivation, an individual chooses the behaviors he engages in and the
level of effort he asserts on the basis of: (1) the valences he perceives
* University of Toronto
** Baruch College of the City University of New York
- 2 -
to be associated with the outcomes of the behavior under consideration,
and (2) his subjective estimate of the probability that his behavior will
indeed result in the outcomes.
It has been suggested that the theory can provide the basis to
prescribe, describe and predict a wide variety of work related variables.
The theory has been proposed to predict the following variables: job
effort and job performance (Georgopoulos, Mahoney & Jones, 1957 ; Vroom,
1964; Galbraith & Cummings, 1967; Lawler & Porter, 1967; Hackman & Porter,
1968; Graen, 1969; Gavin, 1970; Goodman, Rose & Furcon, 1970; Mitchell &
Albert, 1971; Wofford, 1971); job satisfaction (Vroom, 1964; Lawler & Porter,
1968; Graen, 1969; Lawler, 1970, Wofford, 1971); organizational practices
(Evans, 1970); managerial motivation (Campbell, Dunnette, Weick & Lawler,
1970); occupational choice (Vroom, 1964; Mitchell and Knudson, 1971); the
importance of pay and pay effectiveness (Dunnette, 1967, Lawler, 1971) and,
leadership behavior and leader effectiveness (Evans, 1969; House, 1971).
In addition, Vroom (1964) asserts that the theory could easily explain the
following work related variables: occupational preference, morale, need
achievement, group cohesiveness, and motivation for effective performance.
Expectancy Theory: A Brief Description
Expectancy Theory was first proposed as an explanation of work
behavior by Vroom (1964).1 Vroom proposed three related models: the first
is a job satisfaction model; the second is a work motivation model; and the
third is a job performance model.
1 Expectancy theory derives from economic expected utility theory and frompsychological theories of choice behavior. As such, earlier versions ofexpectancy theory were applied to phenomena other than work motivation asearly as 1738. See Wahba and House (1972) for an historical review of thedevelopment of the theory and the relationship between alternative formulations.
- 3 -
First: the job satisfaction model states that the valence (or
satisfaction) of an outcome to a person is "a monotonically increasing func-
tion of the algebraic sum of the products of the valences of all other out-
comes and his conceptions of the specific outcome's instrumentality for the
attainment of these other outcomes." (Vroom, 1964:17) Instrumentality is
defined by Vroom as the degree to which a person sees the outcome in ques-
tion as leading to the attainment of outcomes. Instrumentality varies from
-1 (certainty of a negative outcome) to +1 (certainty of a positive outcome).
Second: the work motivation model states that the force on a
person to perform an act is "a monotonically increasing function of the
algebraic sum of the products of the valences of all outcomes and the
strength of his expectancies that the act will be followed by the attainment
of these outcomes." (Vroom 1964:18) Expectancy is defined as the subjective
probability that a given act will be followed by a given outcome and varies
between 0 (certain non-occurrence) and 1 (certain occurrence).
Third: the job performance model proposes that job performance
is the function of the interaction between ability and motivation as shown
in the following formula:
Performance = f (Ability x Motivation)
"It follows from such a formula that, when ability has a low value,
increments in motivation will result in smaller increases in performance than
when ability has a high value. Furthermore, when motivation has a low value,
increments in ability will result in smaller increases in performance than
when motivation has a high value." (Vroom, 1964:203)
- 4 -
Operationally, Vroom's model implies that people choose among
alternative work related actions in a manner that optimizes their expected
valence. That is, for each action, people multiply their perceived
valences of all possible outcomes, and finally choose the action with the
highest expected summation. For example, consider the case in Table 1.
Imagine an employee choosing between two actions (effective or non-effective
performance) each with two alternative financial outcomes. Suppose further,
that the employee's expectation of the occurrences of the outcomes is as
shown in the table.
Insert Table 1 About Here
According to Vroom, the force to choose effective performance is
($100 x 0.8) + ($0 x 0.2) = $80.
The force to choose non-effective performance is ($100 x 0.2) +
($0 x 0.8) = $20. Assuming that the employee is attempting to optimize his
gains, he should choose effective performance over non-effective performance.
The original Vroom model has undergone four developments in the
last few years: (1) the distinction between first level and second level
outcomes; (2) identification of intrinsic sources of valence; (3) the dis-
tinction between Expectancy 1 and Expectancy 2; and (4) elaboration to pre-
dict the effect of given additional variables in the work situation (e.g.,
the incorporation of ability and role perceptions to explain job performance,
and the concept of equity to explain job satisfaction, etc.). We will only
- 5 -
review the first three of these developments since the fourth does not
alter the general formulation of the theory.
The Distinction Between First and Second Level Outcomes
Galbraith and Cummings (1967), Porter and Lawler (1967), Graen
(1969) and H~uag,19711 distinguished between first and second level out-
comes. The first level outcome is the job behavior of the employee and is
evaluated by a rewarder (usually the employee's supervisor) in terms of its
acceptability.
First level outcomes have been given different names by various
authors. For example Porter and Lawler (1968) refer to the first level out-
come as performance, Graen (1969) refers to it as work role assumption and
House (1971) refers to it as work goal accomplishment. In general, first
level outcomes are viewed by the writers in this area as the outcome of
the subjects' effort with respect to task performance or accomplishment.
Second level outcomes are the consequences to which the first
level outcomes are expected to lead, such as reward or punishment. Accord-
ingly, second level outcomes frequently depend on someone other than the
subject himself, such as his peers or his superior. If, however, the work
system ties second level outcomes directly to first level outcomes, such as
under piece rate or commission payment for performance, the linkage between
first and second level outcomes becomes objectively certain. Thus, under
such a system the subjects' subjective probability estimate that outcome 1
will lead to outcome 2 would be expected to be higher.
Identification of Intrinsic Sources of Valence-
There has also been made a distinction between the different kinds
- 6 -
of valences associated with these outcomes. Galbraith and Cummings (1967)
extended the theory by pointing out that certain intrinsic valences are
associated with the work behavior itself. They operationalized intrinsic
valence by measuring the subject's ego involvement in his work and found
that this measure added significantly to the multiple regression coefficient
of performance of their subjects. House (1971) specified two kinds of
intrinsic valences: (1) Intrinsic valences of behavior: those associated
with task performance, such as the development of valued skills or social
satisfaction involved in interpersonal tasks; and (2) Intrinsic valences of
accomplishment: those associated with task accomplishment, such as pride
in work or the satisfaction of achieving a challenging goal.
The Distinction Between Expectancy I and Expectancy II
Campbell, Dunnette, Lawler and Weick (1970) extended the model
further by distinguishing between two types of expectancies: Expectancy I
concerning whether or not the individuals will actually accomplish first
level outcomes such as work goal accomplishment; and
Expectancy II concerning whether or not achievement of first level outcomes
will actually be instrumental in the attainment of second level outcomes.
This distinction appears to be used rather consistently by recent
investigators (Lawler & Porter, 1967; Porter and Lawler, 1968; Graen, 1969;
Campbell et al, 1970;-Mitchell & Knudson, 1971). Expectancy II is similar
too, but not identical with Vroom's construct of instrumentality.
However, the instrumentality concept lacks conceptual clarity
which is probably the reason most authors, when operationalizing the theory
(and sometimes even using the term instrumentality) merely assess the subject's
- 7 -
probability estimate that first level outcomes will lead to second level
outcomes. The conceptual problems associated with this concept is discussed
in details in Wahba & House (1972).
An Integrative Model of Expectancy
The developments referred to above, can be entegrated in the
following general model:
n nM = IVbi+ El[ IVi + i (E2i i)]
i=1. n
where:
M = Motivation to work
IVbi = Intrinsic valences associated with task behavior orperformance
IVi = Intrinsic valences associated with task accomplishment
V = Extrinsic valences associated with task or work goali accomplishment
E1 = Expectancy 1, the subject's probability estimate thathis effort will lead to first level outcomes
E = Expectancy 2, the subject's probability estimate that2 first level outcome will lead to extrinsic rewards
(second level outcomes)
The integrative model above has several theoretical and empirical
advantages. These advantages include the following:
First: The empirical validity of the model can be easily tested
by reference to the current research evidence on the subject. This is
because most of the proposed relations between the constructs are derived
from the empirical findings. As such, the model brings closer the
theoretical and empirical studies in the field.
- 8 -
Second: The model distinguishes between two types of valences;
extrinsic and intrinsic valences. Furthermore, the model distinguishes bet-
ween two types of intrinsic valences: one associated with task performance
and the second associated with task or goal accomplishment. As will be
shown later, these distinctions are empirically valid and prove useful in
the discussion of the research findings in many areas of work and
motivation psychology.
Third: The model does not utilize the concept of instrumentality.
The concept of instrumentality has been found to be logically inconsistent
with the original formulation of the theory (see Wahba & House, 1972, for
details). Rather, Expectancy 2 is used to measure the likelihood that
first level outcomes (such as performance) will lead to the attainment of a
second level outcome (such as reward). Expectancy 1 is viewed as an action
outcome association and based on direct (or simple) probability. Expectancy
2, on the other hand, is viewed as an outcome-outcome association and is
based on conditional rather than direct probability. Operationally,
Expectancy 2 is a function of the product of the probability estimation of
occurrence of the first and second outcome. The substitution of Expectancy 2
for the concept of instrumentality makes it possible to utilize the statis-
tical theory of probability for further development of the expectancy model.
In addition, this makes it possible to relate the expectancy model in indus-
trial and organizational psychology to other expectancy models in psychology
and mathematical statistics.
Fourth: The model can be used to study both job satisfaction and
work motivation. This is because work motivation is viewed as a function of
the interaction between expectancy and valence, while job satisfaction is
- 9 -
viewed as a function of the presence of various relevant valences.
Fifth: The model allows for the study of other related work
variables, such as job performance, leadership behavior, occupational choice
and others by the incorporation of some additional variables (such as ability
in the case of job performance). It should be pointed out, however, that
future studies are needed to determine the nature of the variables to be
incorporated in each case and the nature of interactions of these variables
with both expectancy and valence.
Recently, two reviews of empirical research based on expectancy
theory have been published. Mitchell & Biglan (1971) reviewed the concept
of expectancy in three areas of psychology: verbal conditioning, attitude,
and industrial psychology. In the area of industrial psychology they re-
viewed six studies. Heneman & Schwab (in press) reviewed the research
design and measurement issues connected with nine field studies.
These reviews show that the predications of expectancy theory are
generally supported. However, it is also evident that the magnitude of the
support for the theory is inconsistent from study to study. It is discom-
forting to note that the levels of the concurrent or predictive validity
coefficients (usually in the form of multiple regression coefficients) range
from .72 for predictions of satisfaction to as low as .11 for predictions of
performance. The coefficient of satisfaction is generally about .50 and the
coefficient of performance rating is generally about .30 in the majority of
the studies. Also, it was concluded by Heneman & Schwab (in press) that
research on the theory has been inadequate in three respects: the number of
independent variables studied; the measurement of these variables; and the
statistical analysis performed. It was additionally concluded that there is
- 10 -
an obvious discrepancy between the theoretical and the operational defini-
tions of the relevant variables. Mitchell & Biglan (1971) noted that the
uses of expectancy theory in industrial and organizational psychology has
been less successful than its uses in the areas of verbal conditioning
and attitude formulation.
In an earlier paper, the present authors (Wahba & House, 1972)
attributed these and other problems to some unresolved logical and method-
ological issues basic to the theory. It was pointed out that the essence
of expectancy theory in work and motivation is choice behavior. As such,
it was shown that the present formulation of the theory, and consequently the
empirical research based on it, ignores the rationality assumptions underlying
choice behavior. It was also pointed out that the major concepts of the
theory (namely, expectancy and valence) lack the necessary theoretical clari-
fication. Furthermore, it was shown that the typical formulation of the
theory is based on optimization choice criteria (most writers imply gain
maximizations, few propose satisfycing rather than maximizing). The empirical
validity of these criteria was questioned in light of recent findings in other
studies of choice behavior in general. Alternative criteria were proposed.3Whereas the papers by Mitchell & Biglan and Heneman & Schwab reviewed
some studies in detail, the present paper will evaluate, collectively, the
findings of fourteen empirical studies in light of the parameters of the
integrative model.
2The rationality assumptions discussed in the earlier paper include prefer-ence and indifference among alternatives, transitivity of preferences andindifferences, dominance, independence of irrelevant outcomes, continuity,and independence of expectancy and valence.
3Some of these alternative criteria include variance preference, probabilitypreference, the sure thing principle, the regret matrix and potential surprisecriterion.
- 11 -
Exhibit 1 summarizes the fourteen studies with the objective of
testing various propositions or parts of the theory. In the remainder of
this paper we will describe and evaluate the scientific status of the theory
in the light of these studies.
Insert Exhibit 1 about here
Expectancy 1 and 2
Despite the clarification added to the theory by the distinction
between Expectancy 1 and Expectancy 2 (El and E2) there are only two studies
which indicate that this distinction is empirically (as opposed to concep-
tually) useful (Georgopolous et al., 1957, Graen, 1969). Georgopolous et al.,
operationalized El by obtaining subject self rating of their freedom to vary
their level of performance. This inferred (as opposed to direct) measure of
El was found to be positively related to self report of performance and to
moderate the relationship between E2 and performance in a manner consistent
with the theory (i.e., subjects with high El had significantly greater rela-
tionships between E2 and performance).
Graen (1969) operationalized El by a single question to his experi-
mental subjects. He found that this measure correlated .32 with performance.
When multiplied by El, the correlation between the valence of outcomes and per-
formance ranged from .18 to .46, indicating that E1 alone generally predicted
as well as outcomes multiplied by E1. As in most other studies, the correla-
tions between the outcomes alone and the performance were not computed, and
thus the interacting effect of E1 and extrinsic valence (V) could not be
compared with the effect of V alone.3
- 12 -
Other attempts to operationalize E1 independently of the other
variables of the theory have all been based on a small number of questionnaire
items and have either failed to add significantly to the amount of performance
or effort variance accounted for in regressions4 (Mitchell & Albright, 1971)
or have been found to be very highly correlated with E2 (Lawler, 1966; Lawler
& Porter, 1967; Lawler, 1968; Porter & Lawler, 1968). These studies suggest
that E and E2, when operationalized by questions with similar format, [e.g.,
"If I work hard my performance will improve," (E2)] share common method
variance. E1 might be better measured indirectly by asking respondents to
indicate how much they personally control their own performance, or how free
they are to vary their level of effort.
E1 and E2 have been combined into a single measure, (E1+E2), in
studies by Lawler (1965), Lawler (1966), Lawler & Porter (1967), Porter &
Lawler (1968). In all of these studies this scale has been positively related
to self and supervisory ratings of the subject's performance and effort.
Lawler (1966) also found this scale to interact with subject's ability in
a manner consistent with Vroom's formulating of the theory.
The independent effects of E2 have been measured by Georgopolous
et al., (1957) and Galbraith & Cummings (1967). Georgopolous found E2 to be
positively related to self ratings of performance. Galbraith & Cummings found
it not to be a significant predictor alone, but to interact significantly with:
(a) subject's ability and valence of supervisory supportiveness and pay; and
(b) with El and valence of pay.
4Several investigators combined E1 and measures of IVa, IVb, V multiplicativelyinto a summary predictor but did not compute the effect of El alone.(e.g., Hackman and Porter, 1968; Goodman, et al., 1970.)
- 13 -
These studies suggest that new measures of E1 are needed and that
its independent effects and its independence from the other variables of the
theory remains to be determined. They also suggest that E2 is both a con-
sistent predictor alone and in combination with other independent variables.
The Distinction Between Various Types of Valences
The distinction between (a) intrinsic and extrinsic valence and
(b) intrinsic valence of behavior (IVb) and intrinsic valence of accomplish-
ment (IVa) appear to be conceptually useful. For example, these distinctions
permitted House (1971) to fit prior leadership findings together into a
meaningful and consistent pattern that could be explained by viewing the
findings from the perspective of the leader's effect on the intrinsic and
extrinsic valence and on EI and E2 of the subordinate.
House argued that prior conflicting findings regarding leader con-
sideration and initiating structure could be reconciled by considering the
amount of intrinsic valence inherent in the subject's required task behavior
(IVb) and goal attainment (IVa).Similarly, Lawler (1970) argued that prior findings showing job
enlargement to be generally more related to quality than quantity of perfor-
mance could be explained in terms of intrinsic valence increases which accom-
pany job enlargement. Specifically, he argued that such increases in valence
should, according to the theory, result in increases in vigilance and pride
of product rather than higher expenditures of physical effort. Vigilance and
pride of product should, in turn, be more related to quality than quantity of
work performed. These two applications of intrinsic-extrinsic distinction
illustrate its potential theoretical usefulness. Unfortunately, there have
been only three studies that have operationalized these distinctions.
These will be reviewed here.
The intrinsic valences. Galbraith and Cummings (1967) were the first
to distinguish between intrinsic and extrinsic valence and the first to opera-
tionalize the distinction. They operationalized intrinsic valence by use of a
self report ego involvement questionnaire which appears to be primarily a
measure of involvement in the task, or IVb. Galbraith and Cummings tested
various combinations of V and ability by use of stepwise regression analyses.
- 14 -
They found ego involvement to be moderatedpositively by the subject's ability
and the expectancy that work leads to peer acceptance. IVb was not found to
have significant effect on performance. Graen (1969) and Mitchell & Albright
(1971) compared the relative power of intrinsic and extrinsic valences as
predictors of performance and satisfaction. In both studies it was found
that effort and performance was more strongly related to valence of intrinsic
rewards such as work pride, self esteem, personal development, and experiences
of important achievement (IVa) than to valence of extrinsic rewards such as
pay, promotion or recognition (V). The implications of this finding are
important and will be discussed in the concluding section.
The extrinsic valences. Surprisingly, the effect of the valence
of specific extrinsic rewards alone on performance has been compared only a
few times with the effect of extrinsic valence in interaction with other
variables. The major exception to this statement concerns pay. Two of the
three studies in which such a comparison was made indicate that the valence
of monetary rewards bears a significant positive relationship to performance
for non-managers (Georgopolous et al., 1959) and for both government and
industrial managers from seven different organizations (Porter & Lawler, 1968).
Cummings and Galbraith (1967) also made such comparisons and found no indepen-
dent affect of the valence of money, group acceptance, fringe benefits, promo-
tions or supervisory supportiveness. They did find that valence of group
acceptance and pay is significant when moderated by E2 and that valence of
pay and supervisory supportiveness, in joint interaction with ability and
E was also significant.
When multiplied by their respective second level expectancies
extrinsic valences have been consistently predictive. The valences of such
widely varied extrinsic rewards as working conditions, company practices
superior recognition, pay, peer acceptance, and fringe benefits, when multi-
plied by E2, have been shown to be related to satisfaction (Mitchell & Albright,
1971) and performance (Hackman & Porter, 1968; Lawler, 1968; Lawler & Porter,
1967 Lawler, 1966). These correlations range widely (from .11 to .72) and are
highly inconsistent from study to study.
S-:iveral other studies have shown that the sum of the products of
several (as opposed to one) extrinsic valences multiplied by respective second
- 15 -
level expectancies, have had moderate (.30 to .60) relationships to satis-
faction (Graen, 1969) and performance (Goodman et al., 1970; Lawler, 1966;
Graen, 1969; Lawler & Porter, 1967; Porter & Lawler, 1968; Hackman and Porter,
1967; Lawler, 1968). Although these studies did not test for the significance
of valence alone, they are consistent with the predictions of the theory.
Summary and Conclusions
When viewed collectively, the studies reviewed above provide the
basis for several important conclusions.
First, it appears that the distinctions between E1 and E2 give
promise of increasing the predictive power of the theory. However, methods of
operationalizing E1 so as not to confound it with E2 are yet to be developed,and additional studies designed to test its predictive power are required.
Second, E2 and intrinsic valence have been shown to be important,
if not indispensable, constructs. In some studies E2 and IVb or IVc were
significant independent predictors. In other studies, extrinsic valences
were not predicative alone but when multiplied by E2 were predictive. In
all studies except one (Galbraith & Cummings, 1967) in which intrinsic
valence was measured separately, it was a significant predictor of
performance.
Third, valence of intrinsic rewards is shown to be a more powerful
predictor than the valences of all extrinsic rewards when the predictive
power of the two were compared. This finding is significant because it
suggests that management practices directed at providing intrinsic satis-
factions are more powerful than the conventionally used extrinsic rewards.
The fact that the finding is demonstrated with young (15 to 18 year old) part-
time female employees (Graen, 1969) as well as a higher occupational level
group (Mitchell and Albright, 1971) makes it even more significant since young
females would intuitively be expected to be motivated more by supplemental
income than by intrinsic job factors.
It is likely that intrinsic rewards are more motivational not only
because they are more highly valued but also because the receiver does not
have to depend on others for them. Rather, they are obtained directly from
job accomplishment. Consequently, the corresponding expectancy of attaining
- 16 -
intrinsic rewards as a result of working hard is also more likely than the
expectancy of attaining extrinsic rewards.
Fourth, the only extrinsic valence which is found to be consist-
ently predictive is that associated with pay. Pay has been shown to be
predictive alone and also when weighted by E2. All other extrinsic valences
have been shown to be unrelated to performance unless multiplied by E2.
These findings concerning intrinsic rewards, pay and other extrinsic rewards
are consistent with Graen's (1969) assertion that unless the relationship
between effort-accomplishment and rewards is concretely and unabiguously
established, the theory will not hold. Since intrinsic rewards follow
directly from effort or goal accomplishment their linkage to performance is
both clear and highly probable. Pay is generally the most tangible and con-
crete of the extrinsic rewards and most likely to be linked, in the perception
of the subject, to goal attainment in an unambiguous and concrete manner.
Other extrinsic rewards, not necessarily following directly from performance
and being less tangible than pay, are less likely to be perceived as
contingent on performance unless management makes a visible effort to estab-
lish such a linkage. When the researcher explicitly considers the subject's
E2$ we find extrinsic rewards rather consistently have a low but significant
correlation with performance.
Fifth, the above review clearly suggests more complete tests of the
theory are in order. A greater effort should be directed toward testing the
overall predictions of the theory. Most of the previous studies tested only
limited parts of the theory. Consequently, the overall predictive validity
of the theory is virtually unknown. This complete test of the theory should
be performed for managers and non-managers.
In summary, the evidence to date indicates that the
valence of extrinsic rewards has little effect on performance or satisfaction
unless multiplied by E2, and then the effect varies widely but is generally
quite low. The major exception to this statement concerns valence of pay
which has been shown to have a motivating effect on non-management and govern-
ment and industrial managers in several different organizations
(Porter & Lawler, 1968).
Only the studies by Graen (1969) and Mitchell & Albright (1971)included independent measures of E1, E2, IV, performance and satisfaction,
- 17 -
and V. Not surprisingly, these two studies yielded more information than any
of the other studies singly and perhaps in combinations.
Finally, the overall state of empiric knowledge can be summarized
as follows:
M = IV E r IVbi + i 2
That is, the predictive power of IVb alone is well established as
is the power of extrinsic rewards (V) when multiplied by E2. The predictive
power of E1 and IVa is suggested by some studies but the manner by which they
should be combined with IVb, E2, and V is still to be determined.
References
Campbell, J. P., Dunnette, M. C., Lawler, E. E., and Weick, K. E.
Managerial behavior, performances and effectiveness.
New York: McGraw-Hill, 1970.
Dunnette, M. D. The role of financial compensation in managerial motivation.
Organizational Behavior and Human Performance, 1967,_2, 175-216
Evans, M. G. The Effects of Supervisory Behavior on the Path-Goal
Relationship, Organizational Behavior and Human Performance,1970, 277-298.
Galbraith, J., & Cummings, L. An empiric investigation of the motivational
determinants of past performance: Interactive effects between
instrumentality-valence, motivation, and ability. Organizational
Behavior and Human Performance, 1967, 2, (3), 237-257.
Gavin, J. F. Ability, effort, and role perception as antecedents of job
performance. Experimental Publication System, 1970, 5, manuscript
number 190A.
) Georgopoulos, B. S., Mahoney, G. M., & Jones, N.W. A path-goal approach to
productivity. Journal of Applied Psychology, 1957, 41, 345-353.
Goodman, P. S., Rose, J. H., and Furcon, J. E. Comparison of motivational
antecedents of the work performance of scientists and engineers.
Journal of Applied Psychology, 1970, 14, 491-495
--Graen, G. Instrumentality theory of work motivation: Some experimental
results and suggested modifications. Journal of Applied
Psychology Monograph, 1969, 53 (2), 1-25.
Hackman, J. R., & Porter, L. W. Expectancy theory predictions of work
effectiveness. Organizational Behavior and Human Performance,
1968, 3, 417-426.
Heneman, H. H. & Schwab, D. P. An evaluation of research on expectancy
theory predictions of employee performance (in press).
House, R. J. A path-goal theory of leadere afectiveness.
Administrative Science Quarterly, 6, (3), 321-338.
Lawler, E. E. Ability as a moderator of the relationship between job
attitudes and job performance. Personnel Psychology, 1966, 19,
153-164.
Lawler, E. E. A correlation-causal analysis of the relationship between
expectancy attitudes and job performance. Journal of Applied
Psychology, 1968, 52, 462-468.
Lawler, E. E. Job design and employee motivation. Personnel Psychology,
1969, 22, 426-435.
Lawler, E. E. Pay and organizational effectiveness: A psychological
perspective. New York: John Wiley & Sons, 1971.
Lawler, E. E. & Porter, L. W. Antecedent attitudes of effective managerial
performance. Organizational Behavior and Human Performance, 1967,
2, 122-142.
Mitchell, T. R., and Albright, D. Expectancy theory predictions of job
satisfaction, job effort, job performance, and retention of naval
aviation officers. Organizational Behavior and Human Performance,
1971, in press.
Mitchell, T. R., and Knudsen, B. W. Instrumentality theory predictions of
students' attitudes towards business and their choice of business
as an occupation. Working paper, July 1971, University of
Washington, Technical Report 71-27.
Porter, L. W., and Lawler, E. E. Managerial attitudes and performance.
Homewood, Ill.: Irwin-Dorsey, 1968.
Vroom, V. H. Work and motivation. John Wiley & Sons, 1964.
Wahba, M., and House, R. Expectancy theory in work and motivation: Some
logical and methodical issues. Working paper, Baruch College,
City University of New York, 1972.
Wofford, J. C. The motivational basis of job satisfaction and job
performance. Personnel Psychology, Autumn 1971, 24, 501-519.
Table 1
A simple two actions, two outcomes work choice situation
POSSIBLE I\\O,~NSEQUENCES !
C'ALTERNAT IVEACTIONS
EffectivePerformance Bonus ($100 X Expectancy o.8 0 Bo us
A22 $
Un-effectivePerformance Bonus ($100) X Expectancy 0.2 {No Bonus ($0) X
',Expectancy 0. 8
4-i00)
0n
L44o -a
4-iwO 0
HU cc-X
14 4.1 VO V
4-4 j CuW la. UX X C0 -
cn) Q) -H
0) 1 4.U a)U0 V0 ) 14
Cu 0O 0)4.10 '-4 0 )CuU 4-4 4-4 -' 0; V .0 -C14 0 QC-
Q CULu44 0
0-4Q-C0) Q)Cu)
40n) -U >0)4J4> -L V _.0-4'---4
-,- W C: n Q) 1 10JJ O tu O a) > L C:
0 14 > 0) . -H 0) 0) 1
40-4 --4-H 0)C4-4OC U)c: a3) cs C- co u -4cu Cu) W 0 > Cu C0 C Q)
U) u >. cu > 1-
,U C CuiE -4
C0 -4 0 > 0 0 1 4-i 0-HI- ) cu V0 0) ' 0 0 14
V C0 0-3 ) V Cu 4-4 04C1c toJ4- ) C .4 0c,-4 03) -H W Q) N 1.4W 0UO-~C/)> 4-4 0-0) 0-CL0 U
4-Q IC0-H 4-4
-a > Cu U-,H Q 0Q)
Cu. 0-- 4W U) 4-4-,4 En a C: :>. 4-n Cu )Cu 0 V 0) - 0
C0 OVCuz
VI Cu'--4 0) 0) Cu
,rS. q w 3 En Q) L4-J IH- 4.J1V 0)Va0)1.4-4 0) 0: : 0 4-4u0Q0.0 Cu.0 0
4I-40
-W1.40 >0- 4-i0) -,4
0) '-40() 4-4 4.1Va .0 1.A4.*Ua0 Cu 00)00) -H- Cu Cu VO0-1.4 Cu 00) Cu 0 CO .4a> ) fa0-
l)1.4Zs
Cu V)Cu0)0)4-i
4C4 Cu,
0) UCuV 0) CL0: 4 -H4
a) -H
H :> 0
> 14'--4 Cu Cu3) 0Cu, >> 0W
0)-HO 0~ > -H 4.-,-4- -J0-4--u V x
1) >.Q) o)J CU
Wd a
0 a) Q)40- 14 0- 1.4 0 00)0 44 U1 44 -4' '--40
C 0 CW u44 0 C > E'-4 0) 00 Cu 0 14
Cl) 0.:- 444. N 0 .4
Cu44 CueW >3 Q)C/n 0 I-'- 0 0-
0) Cu
0 c 0) e0--4 0 > C00 I) -H-, H
) Cl 4-U 14
-0< u ) :uQ) 4
4- > )
-X Cu Q0 >-
Q) cn Q) C 0o oA 4u0-r : C
a -a a s a CU4. 0 Cu -4Cn u Cu14 00 S 0-
4- >,Cu 04.1
0)J * 44--H.*H
4 44-i -,A -,
UW C: LV
o4-Cu0) H
4 0) C0 b0 40 0 >, 1 Cu 4-4 -H 044 Cu 4-i w 0 H C) 44
,a UO)wC:O Cu- 0000 U)
bSO0-HX0) C44to o -, Qo
U > Cu -H >- 0- > 0CCU0 QU 4- -H
Cu C0 00 C 0 -H Cu0 CU-H-H C14 0C FA144-i 4-J En 4-1) 0-HO u U 0 u4-4 0) ) > cu C140--H *0-4-i 0C uQ) X .0 >>X -H o 4-iCl a) E U 0) r-4 0-OH U
Oa -4- C 4Cu QXa) ucu c In c) C0 4J "Q ) ouw auCu --4 0
3 C W-4'- ) 02) 30 .0 4-4 Q)0-H.0- 0-440)) 4-i'> 04 4 X )- Cu 0
-¢)4 CU 0 0) 0n L: -H- Cu
0I
C0-HO U)0 4.1 4-4 > - )-H Cu 4 3C00 04.1 -4 0) 4o 0 -H
a}4) C00 0U1.400) 4-4>4.10)
1 0 Q) 4410)a C0 -H 3 CN-H V X 0
4i w U4QCU,C0L)--4 0)4-H10) ,1 h4.i c 4-Cuc
0) 0. CU -H Z . 0H J-i Cu 0 = cC -H
1.401I.40Cu14 0-H 0 C)> 4-4 -H1.414>0)>)14 Cu
CO4-4 04O)0 ) 0
0Cu Cuo4-4 -H U -H --4 4-1 0 4-1 -H0) Cu Cu c u.C/) 1. 0 4 Cu
a)4 1.
Sw co (L)C4u0
0 > cu -H,
4-4 Cu 0 0 Cu0 4i 4i U
Cu 0) u) cu x60 0L CaC: X Ca S4 44-H 0) 0) 0 0
1i 4-144
0)
S4 U
4)c 4-i
Cu-4-i
44 1
'--4 C0 0Cu Cu Cu Cu
Cu3>
4 --
V> '- VaN CO
4N 0
0 >4-i 4.1 r--4c Q)) En 4-is $4 Cu>04
CL0
X> W -H
CL v0 M.040.,4 X) vCX*H0 4 CuQ0 Xu Eou 0
0)) 0)c
>C 1
14') > Cu. .04
CL wc waL'aJ0 4-4 0Cuo 4. cCu > 0 .d 00 UCu 0 En
140) 0)'-H.0-014LI4 aUQ) ),c
44 4s
4
o4-4C E0cn 4-41 400) 0a 0 0) 4.1 0 >1 NU4C-
NOiC 0) C- 0UL)-L)0 04 ) ~H0
a cu>£toXco>cs0NCE Cu>CuCoHb-> -H > 4-4 0) 1
" u 3 C x-5
WU i4 C Q) C)
U CuC 4-1 0 r--4 0$ 0- Cu0CuU -H 0-H 04 C4Q ) -1H - 0 44 0- 0-4 C044 H 0Cu 0Cu 0) -H.0 Cu --4 U-H> > C u 4-4. UCut-4 4-J l- H44 ^0 Cu*C0o uo Cu ua0 -w 1C
0 -HO u0C P.r C0 > -H V4i 0) 0) 4.10 CL C O 4V C0 > 0 0-H4 -H.UQ0 -4 Cu4U Cu 041 C0Cu E0 0 UU 4-4U : Q) >4 -Cu > Cu U)>a 44
J o0V 0 44- C V OcV Cwu 4. 0 -H > 0 Cu to-HCo C aM00 aJ)-H> - -H --I Cu
0)>C14
4
0) cu a~ CN iCu V 4.1 -H W Cu.,4 ,o Q)300 00) FE0)-H 0)CuQC 1) 0U)0 V) U), O
Cu Cu 0) >0v
4- Cu -H > -H ,D4.1-4 0) 4 o vJ Cu
wv X Z u0 (a
4-
4.1
0
V0Cu
0O U U00
la. 0) C w
0)>-U 0 Cu
1.r4 OC /) >
00 0 cu -H
0) -H 4 -4
VC CuO0 .i C CuCu 0 0-V V Cl)
0H U)41 4-
V0 00)0 -r4
)V0 .-4
1-x--0 0- >r
00 Cu V:0 V 0Cu> 0C >0NX g-4
0 0Cu04.0, 0)14-H4
.00) Cu -H 0-H
>J Cu4 cCu000W) CuH
-> 0 14-H -H 0a )0)
to a) c-Cu c1 laH~0 0) 4.C14 0O 00
- 40 0 )
CuCu0 0Cu)C C N - C V
.00.C0 4.1a0--V0 0U 0)Cu
CO 4C),) 0 U 0- C:-H
Q) N,4 I.UC C
x 0 + - -
u C 0) el,
U, a).0)
C >
la,0 0-H*S ) C3 ,- CSV-HU 04.1 Cu 0>
00 4.1 -> CUV'O-4 '-4-,H4 CU '- C04 Cu .Cu 4 I014- H CU 00C-i O .0 0 --H
C C 4-i 00) CN-.O Q) SW 4O 'a rUOVnO 4-lU +VOC0 4-4) H 4- '-' 000) 4-4 1~> 0) 0 Wx 4-4 -H4
W 0E-, > -C 4.I
-,H U 4.IUC a) Q)4.0 V0 :~Cu 0 Cu 1.4 0
H -H Cun0) V U Cu .0 1 DICu 0) 4-J0- H 44.00 a-4C u *-4- u
0-w .0'--40-0-0O. Lx)0) Cu0 ) -H0-40 )Cu)0L -H 1 4.1 0. 0.0,
> 14 Cu -4 -H 00 0 4.1
n> U 4 zZ 4 3
C i) la.,{ a) Q z
00 C4 -4- MD *S a) 4i0 Cu V > C
-H0) 0) Q:w 0 --
O C1 C.) CU C:4,4 > -QH0)4i
JJCLQ
04 CLCu 0
-H 0 -H 0)
t1 o4 0 1 Vsw C C r~ u swo C) *H- :: *S- Cu
OuCuO >
W4.1 ) H-
o sw-0 0
4))-4-H C
4 C0 0 4
0z 0) -H4 HC
4-H 0 C-)-m
0) a4-0-0) 0)U
Uw bD r= UC w Cj
- >0 D
4-4 U l-
w 0>014 U0.'-- 0 0s
14 Sw 4- V
I 0 >
CO) 4.Q Q0
0) 41 -HJ Cu 0 Cu44 1 > 54- 4i00Q)O0-H1..4 1.4
.0 44 4-i 0 Va0O4JO1 U44 Cu 0S
Q) :3 w Q) 5-c V 0)-4- H4.10r-4.100-
N4N0a,0'- mAH 4.1 0-SH Cu
4-4.41 . -'-I V U
t - 0 N>~O-J0 )044
- )Cu C - 0
.04-4-4 4o 14. 1 0
1. 0I 0-H-4 0440) 44 JJ -H-4-U V I Z 0 0 4.
W QJla N to0 * Sl CUO- 4-4 4iJ-HO0r-Cu.On04 U) L)u U 0-H C44 Cu
V0 -H 0 1.4 4- 4. 0- 4.1
O J Cu a. C0441 N U_I ) 1.4 0) 0-Ho1
0)0)4JU 40C0Q)CU C- m 00 0 -H Cu 0- Cu 00 0),1,0 CL Cu ) > C o Cu-
ON El C O0 C 1C 4-,-II-4 Cu Cu 0) 0) -r 0 mCu Cu
-'t- U) 044 Cu4-C 0 >
-H4-H
.r
.Ix)
0)
0)
CuOCu0l4-40 .
Cu01) '--4(000)0 >Cu03)1. .-
V4-i
1 > :0Wa 4- 4J 4J
VCJJI CWOOOOcOC C4C- acC)V C l4J 10W4-4C sC01 01
> 0 4-4 -H 0 -
H 4-14 C)u 4-4 Cu C Cla WJ-i 14
VO4-l Cu 4.J44-i0 0 014IV C 0V
a11~ CuCu4-iV' M-0E:5V> 0 X -H 0 -H r-4H.0 -r-4 (1) H4C) V
C)4-i
as0
U) 4i4 14
00 14 0 -N*H-4 -4 4-4 'Ie
44 W C40X IC: Q > Q)Cu -H0-Z
4
to cubo u
1: C0
O - S CO C
W C u0 4 C 4-lWCT
4S -- 4Cut
CU U4 0J4-i
au
O CD 4-4
4-1114sW 4-4
W44C 4W 0
0Q - H4 11
-,, Co
14, Q)0 co coC3UuCa, av 4-alZ Z E
U)r-
J 14
-H Cu
-4 10 ) 4
u
0
4
C:a la aW oO0 C
co (-r-4p Sc )
C:) > :a
C)cn a)
0CuvXCuEaJD
0)
laa14
a) a)01-W
4-iX Cu
Z01k(3 5n4
-
-H Z
H-> 0
co
4-4
4
- H44
H 44 E00 1441 C
Q 10cO4 0 CO
> (a wO 14 0: L) 01
,a H
)-4 V3> C0
4~-4 C 0: H-
Cu
c)co 1) 4-i
01
14 14
4- .4-4
14
0 (3
Z
44
W1CuC. )Cu
0C -4 0 tO)L)co 1/) 4 Cu L/)
40 IbO la W Q)C O >m ¢ i rqCw 4-i-H
C4 CUH1
)ai
facL CL Xx u
C) ) ) uCu4-4H-' -4
4-4 C -H bOC) C C E
(1 bO -rl
) -UZ =r<.-44O4 0 4A
COJ 0
00
-4- -4 04-JO 4
4-i oo Cu
-,
0a>
- H4 -H
4-4 4-4 ;>
C)0
0u4-a -H
4-4 4-
-u a)
CL U
X
0) uC.r r- a)i la cu a.a
40O CO Cu14 54X 14a
r= r=aJJPc40 0 -H -H Cu
cu >
U) a4 Q) 4JbO 4-
a.J JJ O
C C)- * CO-4 Q o W a)
> * U)
O14014-404 ~ 4
s4 > C. (3 5 CO0 C CEn -r Ct 4- C rO
> 4 w *^ N
01 0 -
4) - 4-4 '
a4 W r- $4J
En O CO COcn) ce z u u) 10
Cu) 4-H
r-4 ()--I O~CO14C C CO
> o to ajJJCEOJ
0 CO
U-4Cu C
J- C) _I C1
0 r
4-C>0 CO)0a C
Ll-4 4 r 14 C0
0*C) 4-
-i C)XO
04CO >u C: 0Oa
0-H-1
14
a u(4-1 Q 010)
14
4-1 01)> C.)
Cu -H4.0- >
CE0H4-4Cu 0 501
cu 0) C) 1),0 -mlCH Ct0-4 4-4
'H-4.0 4.0O 0-4-4g
0,14.4
Cu
4-i
01)
0-
01)
0
4-i
00
H C0-Hci ^ a) E 0 w> W W - U4J>>iW>-H tbO 0 U _I -C Cu 0-
4-i E VW C =, "U C
-,O ,1 U) WbOH SW4O U)CD UJ ^ 1 -A -r- D: 0 CUJ (U) E0 tutC 0 la 4-4 r-4 >1-4 W4CaL w bo 10 (a Wi a) 0
C *,1 sw CD U) Q) wk LHW3 14 014 Va Cu
csciJ C O 44 cua CL o1c-C c C E4 4+ CO uw H(OE 44-4E CO4) >F 4
w 4 -i Cu Cn U CuS-0 4 -4 * 00 UC o4-U-,WLH O a. 4-4 Cu 4-4-H Cu 4-4
W44 >OE4= -4 U) Z
44 -)4W4 0 14W0W 'WE1O
4-44-4 COOu14 0 4-4XO X J
3H --WW r- o cu o sW o C C 44
C/) UC C u-H
-H > C bo U4-1 D4U o H 40 0 C) 0o 4
4-i 0 4-i-UH C 4-4 C)V
W) L0 W Q W Cu
C -H u w)1 CO UC O) 4ua) Q) c C a. Q ,l- a, C
4 H-4 Cu 4-1 4i1cuaC i QW Cu Cu .0 H
14 H-- (1b O W 4-4
Cl 14o o-14 14 . 0-
*U) o>
U1 W44U4- >U) U) 4 U) o o C C.)
4u 0 Q Q0 4 Cu
.- ,9 4U C) 4-U a04-U 4U C C 0 4U
Cu Q0 -H 5 Cu 01 Cu
wrl "SJ a, . >C)WQ u1s1-4Cu X
C o 'D > 4-4 o .o U) C f: U) o-H4 -H4 4WI 4-4 4-4 >
4-1 Cu 0 C*,- 0 C) 44 > W C: >1CU C4-1 0 0
4-o a4iu0
OU CuUWOX$4 u~ 4Jaa -'o 44 Q) av W cn C1 o UJ 400 Pn CL 4U CW
4i1Cu
.,40
0 4-4 -H
4-4
o-00WXaC) >
CO.)
-A al.,1Cl 01)4-J C 4- -H
CuCu Cu V0
Cl) C0 U)(n CLcu
I0
i
L COU4-4 > 0
1W 0 UIZ 4-4 C: 4i 4U
Cl) a z Cu
4-i 0 4U
4-4 - V= 0 1(DCu 4-4 Cu 14
4-4 4-i COl -H
o X44 -4Cu -H 44 --
01 )l. 0 "CL CI w
O (
C 0CO -1 4
0 C 0) 0 UC O4 >, >Cu - H4 Cuo( CuJ Cu 4-1-H4
c Cun 1)uC) C)44 C)-H 0C 01 1W 0
14 1 0 )1I CO
- -4 -
co
41 4
p 01p5) E444
04 >1 40
0
4-1 io N r4H cu
14,-..-
00) Q0C cuc
4 14 Q0-H co CU -4
- 4.J 0 Ez 0
SJ C0 60
4-
i-H
-*H 4-
Vi C)*
U Q)0 -H Lf)-H-A4-4 4-i
0 0)4o H -J
CY
1
C.) -. 0
Cu Co
cJ a w
I >4-I C'
as ) C O
CO0
O C.)H4 Cu
4-4 14
c: C a: *A
C) 04 I
o cuC .
lz a,}
4-I
0) 0
W- 0
Wo - (Z U)4- Q
-4 Q
Ul) Q) av CU HV CO 4 40 w C w
tC
0 V) cu
4u1 -4
Q) Q>*
HIuI
CL bC 0 Ca) o C ci Ca
C/) 0- ~-4 44
b C010C Ui) Cu
)4 14 E: .4CS4 40
UCCO U ) -
a) CZ C)OC U
CO O
av C SJ CI
> 04 b
40 "
$4 4 1Ol
iO O
CO 1 r
> > wXr a,SN- o 4 ,
-
0 0
HHCsuCJ C$4 -4 4
-h 1 0 14CO
CL U 4J1CavXZX M 4
o
U)0 44 0
10 0r
-Ito ^m
4b( C. Cu
*l
r-4 c1 0)
00 -A a) CO -
ON o C Q s4r- U) C > 'A QCN- a ) Q) a)uM --
JJ CO > QC) c- H Hr
ts4 W CJ-ro Q )
HO 0r -- 01 U 4
4-4 u
C)
o-01 COO
o0 H4 -H 4-4
00 0 -It.0 0 Cu Cu
4-4 4.J CO )0 us w 'r
o X W44*H4 w 14 04i 0Cu COVl
m 14 -Cu0co x6>. a.uu C
o r-40 0z XmCu
C:0
(1)uz
W)0 4-iO -H4CO
0.40
OHCO1z)
CIu4-)C:C D U)
C U) > 444 U).14 O
U) U) Q) - Q) Q)Q)H- 004-I O C 0 0
i 'l-4 > cu o oCtl5 C'-4H)4 C-H 3s4 a)C) 0U)4- tbO V4-W 0 -O 4-^ HC -4U) o 3 c: - 0OC: aC - -Lico --
O Q) -4 .0 C U)E > C) HC 0 U) 0
uz ) .1 (- Q)UJc,) ) 4_)4-J -LJ M 0 14-J U) 0C 0-^H 0CC1 -i) Nu*
4H U) 04(3 C O- W
[ Ql0 F .Q - cu E-4
U)))av
U) -L u
'4 4 >U 0
CUC C
4L 0 4-4~ -H 0 "-4U) a) 4J 0
OC) O4-4
CO4-s (a) n l0 toa0) aJ C
0E
I, 0CQU -4
CZ U0 U)
C4 0C,4 CU3 0n 43Il E0 (
0> 4-
0 *0
CO
NCA U 1)) )
Cl) o 4-i 0=4-H 0
CO > - w
,4 U - 4) 0uCJCICl- a)
(1)
r,4 aJ, W/ 5CD o- W1vaC HQ)
CUIO C H0 )- 4i 4
m CO 00 (1) _o 5 4U)S u
Ua U) W U)
< U) U)
0-U) 0U > 4-= C 0 0 Cu
Cu()) 5: En U C
O,4Cl 4 P)Cl -a U'-'C 4C
0-H U Q O CZ4-I a C > W
t- C:(3 X ., -,- 'S) a 0 EnW= 0 ^Q
C U4 4V4 W -4CUJJ C; CC 4-O > C
CUJ - 4 -r I4
04 -4
CnwH O nCrSl5 S t
or- UCOC -J 4-
0
O 5
I-H Z04I u4-a,5 0 :0
., Jlcu ut i+O3>W
IX r
C u)*-H C -H1
44-H ,-H
03-) 4W nU)
0$ - 4
Cwur
4O 5
43)-rI U
w C U) )cu > X a)c"0S H C
C Cu 4
U) 0sC
CUX uCU> xCU N>- O Z3
IV
X a)aU1) Vw
4X , Cu
cU U)
C --
-)
C: Cu0~-> 0
U)
C:a) XU
CU 0-
>>UXHy
.4--HU
,-4 CgU) 1-4 0
C uM-H4 4-i 0 4-i
Ow X 0004-i 0
C0
H 45 0'A
Q)H4 I
C O 43 CU
.O 5X CY) CZ*
0 4- cU)o 0UC NO C )
H-4 4-44-4 4c 4-oo a) C U) 0
J4JI-H -H 0 QCU^OC CU 0o00 4-40u
~-4
4-i r U4 -i aC Lr) O)-HX Cud Cu CUr-4 4-) CU4-
4-1 : -H 4 -l4- 4 O~COCV E > E >4-u0oc -H C Q O4-40Q CU
Cl En W E 30-4 U u Cu
cnCC
0
L4-44 U) UC C u
- 0) a-4 4 -4 0
>aJOVWQ) w
J
3 -Ho 4
U) C 4H QC
a4-J
.,- . > rco>, -H
4-4 -H4 -H4 4-iC>9 O
C: 10 aCU) Cu 4--I a.0-Va 0-H14--4aC4-i 4- 44i
0c C
O3 -OCUA 04 Cua Q Q0X
Q E
(HC CU) -0
'¢ O >0) U)
- 44 U)
a U)CU O4 Cu
4-4 H4C 0400-H CU
C UW4-- Q ,
CUa4>4i > 0C( >
-H4 -H- 4-i ~-4
u U)u
CU< ~ Cu -H 4-4 Cu
rJCU
-nCuU)
0Wb >'
a4~C s00 )
*w I
4 4-
C
Cn uo C/)
4i :j 0i
C) 44 -H Cl -H
4-i 4-i
o c>-1 -4 Cl)
a Ju .J0 U) -Hi 4-I 44 Ca.
UIO) 440 5o 5 w
i o 4- . a Co o
u cCUL ::Uco- 0L U)
cu -4 J 0 HU) > 0n >
CU4ON
OC-H 0
-Cu
o) 5
ON LI- 4-4 0
LC') U) 4-I
J
> cou0
H4
Cu CUr-45 C.)
h0CU 4--i 1))U)O
CUa) 3 ) (3 -4
4uo U)X C
4) CCU--4 -
0 Cu a) CU-'--.,4 $, '> -4 > CI_C 44 0 -H HSE C - C 000 0 CN-H 0 uU 0 -)
00
0Q O
V 4 CY)O
a~-HC) 0
Cu a H0 0 O
0 4-I C
00 V
CDJJ 4J -,4~i OCULV CU4 00 -,4 CVC a, Q -L
4 00
O- CC
-i0 En HCv .,4 L
U ObU) 4-IC5CC
00 0C
C -H H 0
Cu 4-i 4-i -H
O) .H > *H
0-) 4- 005X0 0o-oHX-40-O U)
0
a3) c;
4-O U)C -H 4-iH, 4-i U)
0 0 CuCu U)
C: Cu -H *-H0: 4 U- U)
4O -, 4- Q)
-4 n -44U)C 0 0
44CI .4 0 a0-
0 -H 0 >0Q) > U;
CU U) U 4--i
C- 4- 4-1 00
0 U 0JJ U)W0-U CU
CUzuCU U)
0044 U) U
0 Ena) CO4U.H U) 1 - V>U 0 u4 U) V
U) 0 U4-H Cu r4 C
4o --i C: -4 >uU)C a
-1-* CO C4 w C-H 04i 0
C-H 0 0O
04 0 00¢ -- wO O
rz 0 V)04-00JWUCJQ
40C4-J
U) 0 0 C-HU0E Xc U)c000-vH U) 4-i
U- -H C U- :4 0 CU 4CuX Cu X Cu Cu 00 0W 0*HCuO CU0CU%0- 0-UO-H LIO4-i 0 -H 5 0 0 U)O-H 0- CUo¢C JJ 0 0 -S -H Cu O-
0D0 I-HS0 0CU)CU *V-H Cu
CU 0 CU 000U) 0 C 0
0 0 c) CU 60 w
-0) 0o
w rawkaJ a)
o C-0CU CUH S0~4 C >-j- CO toCUV -) U) 00 CU0o
0 4i a QCU U) 0 C
-40CU3 5) 1u,H F E O U3) Z 0
- HbO -HV -CV,-C a)CJ JHUCIOM.-4J 0)
-C *HU )
be 1N-C 4-C 0 a)
a S 4iEn0 Va0 4-i - O a)C CU) a) 4-C
4-4 U) a) 4-OU
-4QU) 0 U) > 4>H U S 0
C: 4- i a)-%@H Ci -H a)O
a) Q) a) -H 4- Q
E 4-4 -C > -C Q) C 4-C0 44 CU) -H U) 0 U) 0
>fiQU >
0 V 0 a) 0 ,I 04-C a) C H4o -J 0 4-C 4-4
CoO I- H< W04-CU N 4-J Uo- U)
C2 .4-HO a)u)OJ0) V4-1
a D >-H OO@a)C 0 w U) 04-i 0-a~V) >U) 0 a)O$
04J-H. C 4-4 C 0 a 0 a C0 C,-4 C v 0_ 0-a)H a))-H W-OH f.) X 4-J oa H oC JJ4-i 4 Ci4" C w- ) C cu C>z cu )0 . > w c o o404 CcU) CIU 00 a) r-C , --3 E c-> 4-I-H 4-CJ C C a)
aWr-
4av X(1
W CO< W4-4))-X 4C0*)--4a)0 CO-Xa)fiO0 44U)^Xvoa) -C O a)4 0) U) C>4-4 V 4-C4-Li *H4-4-a ()UCO C4-Ja)cq-J uC> a-HUEnla)VO uc0U) ) 4Aa)
X 0 3 >~UC 4CO w > -4-H -4 4aJ-ri U) 0oC- C> C - Ha a) -4 ) .C 4J 44 SW 4-i CC.U4J 4-4 -rHO C CL
C>,O a)4 )" -4 > c-H,> c 3 u cU) V a) *JC- boC)> 0 X la U) CO *H U) a)4NC 0 0 " J-H -
beo -Hr- --, 0 a)Oubo U 4-J r- C: w M. U)u U) CC; CC - 4-JJ .-H V J HHA Z-H W CL 0 40 COOU)to-H>O>Wa). 4-CU) 4-4 0 -H4-J.0U) 44CU )aC.)04-CO0HO O U)C-A a)5-4-I -H -N H4-O4-C 5 U) Ca0J-H1-C4iHa 0 0-J.CHvCJJ CC4Cj4 ,1> - UEn C 0-.0 4WO)00Cab CCba-HH9o> c9cf scW<:oo3uzcXU)>C U) S C a)
-Hw
H HQ D4-C > M X CD
4-4 > (X) CU)Ca U) 44 MU)O C CO C'4jC-Ja) 00::u)M >a> HO 4-4 3 WoUQa -LJ0)4-4 - U.1- M o F- . LIL o -L Q) r- 4- o Cu Cr C) ' C)HU)OCZ-C) -14 5w wQ-)OO -1 a) OCZ l Ca l O COC) > C- V 0 - .0M C: i u w toaW() L-a)X Ca-4 -H-C:a) -O a) L4-
uM - H *H U -H U) 4.J 44 4-J ' U aQ) COO U) 4- -4 00E )U) U)M Cu) OC5Q C C c > M -Ha) U 4-J 4
C C C'J]C aD O > O -H -, U 54- C 0v-,4 tU )0 O0A'r D'r/ 0 0SHO U) 0 O CaO0 C4w1 w4 75 QW~sW) a) 44 ' i-' tA -4 Q) o U a) -r4 5:- c u
-HI:> a) C > C > ) -H C CU a) 0- ra) V V a) U 0 a) U) *H 0 -H0 0 -Ho0 -H- - 5= U) 0 C/) 4-) u - U) CO 0 Z 0 t4-4 0- 4-3 V
U) W0
c oo,ci av i3~ 0
S 0 w w C 00- 'A a) 0 cu > Q)
V-)0
0 -H3) u-La
C E $v 0w X 0 *H0 000oo 0m w w 4-i
4-U O) .I--
CU)) 4 C 0o0-CVn a)0oa)aO a
)a)
' :o0 -H a) a CO -a 0 -
.4-J -4
Cn C D O 4-,4 a) 4-4 U) 44
U) 4 0 a) a) 0cU) 1- 4u >- 0
.,4 o CV0 4-C Vo - U)Da
*) a) 0, 5 U) :04J 5 o0C 4-- CD 0 0 -r
*. 4-4 -L.
*.Vaw () C a-)4-iQ) U)0-U)C
a))W 04-O aJ C C 0U)O -H 0C.04-44 C^ 4CJ Q) -H C 0
a)U ) -H C> 4-C O (a)O= a) -H- 0 4-4 0-O 0- U)O 0a)OC *QHXC;
> a) 0 C L a O- H --H O CO HA4-C 4- n c 4.J*H ^ -i-uHOU)C )C UCO0 0 4-4 a) -H 4 4-40- H cn e toU
4-i -H -4 C -HO 4-4u b- bO 4uC
(14 CU U) -AH -H -H XU0 44 U) 4-4 -C -: U)
U)
be
.*4 C4-) U)4-Ci
0
C
ES-ia) <- -H
a) 4u- obO U
U- C n aO
Uzt H o
En _ 0I- C
C4 -H0n4CCL
U)-
E r=4O)
04U
4 04
U)C o U)Co -C/lU)
-44-J-H C> bOm Uo, C: faz
U U)4 0 C>4 0C M 4)UZW4-a) C >a)- o a
C o C
a.) u- be U-a) > -H X c U) C
n > C> -H U)
a)
U)U) la
)4:4-J)--
aL) CaVa)C: -4 -H-
a) P0 C
a)3 H
C-C> 0
4-)
C r$ C O4O a)
U) U) C C -H
C C -H
-H 0,
04
CO) 4-4 CZ C:
CO W 4a )
a) ,-4.
4J
4 0) 0) .1CZ Ci
O 4-io a)H a
CL > /-- -4 4-C C:> 4 4-4
H) /--- 4- 4-C
-H 0W
CCIO
4-I 5 4-i -C 04X - UO C/I a)^3Q. C)u
CV-' 0 C 4-4
4C 0 C CN -4 3> -, CN 44
uC C) Uo r- ¢ E - 0
a)
C C
,-H
a) 4-C
U)
0- v-H a)v
- 40:0
-HU)
4-CUQ)Uu)
4-C U) cU0 C
.0
U)
4-i
C, O
-H 0-am- Q (3)0 a)
4 -Ca3) 4-C
a)as C wJJ4
COa) w a
C HS C a) C)-,4 ,H U)HnO r -C U-tl0CO C Ea)) 4- 3U) a) 4-
C 0
C -4 cU )D )ov4- l0 U 4-C 4-4 UC
X C 0 4v-r- )
4-Ca) o 5, o v-4U)O U) U) O -
a) U) 0 U) C O4-W ^-C a)
U) 0 0 CNJ-C E
H
Q) co 4-i 4- V-V U)
4-4 > > CU C 4-)
4-C C ;>1 C 0J0 U) :J r O0O
a) H4 V Ca) U) ) 0 4-1 U)V U 4-i
C 0 X 4- V-0 U)O a) > O (a
4-4 CC -H a) a) 0-H
-4 U) -H 4-V 4-C U) 4-V C U)
a) a) -H U) 0 X X )
4- 0 a) 0
-4r-414O71%4- U)na) 4 a)
U)c
-H U) 04 b
Z 0D a)Cc n 0-4c;e
r ) CL
En-H4 L-4 a
-4 U) aJ)044 UU QC OOCa) U) 0W¢Os4Z U) 0
.,4 W40 4U) C-.J C w
U) C 0 C *H Q)CCa 0 Q0 0 4- *H CH -H 0 4J cU -4J4J OO)WU)
< a.) C -4 0a) 04 U a Q)O 0 - 0* -oMiOL4i U) n0 C44 a)U -rH 0u O w U) > En
Cc)
C-0aQ)
H 0
OU)OU)S
4,-4 0J
4i aJcn
En a0
H,- 4-44-COpU)a
C: r=
:> r4
JI0-
V~~~~
a) o * ^
V a) U cU) X
0 U 4)* 1-*
uo
u cWou
a)hI U t5C
-C5)m Xa C OO bo5 s -r Q) - (j
44 C: OC -4 ) to-I coH *q -a ) c-
O
-4 0.00m
>. C)C 4 a)4n o JH 0 0 0 11 11 U) oOeZ S C CU Z ZCa)