Upload
ann-m-minnett
View
213
Download
1
Embed Size (px)
Citation preview
Internal evaluation in a self-re¯ective organization: one nonpro®t
agency's modelp
Ann M. Minnett*
Salesmanship Club Youth and Family Centers, Inc., 106 East Tenth Street, Dallas, TX 75203, USA
Abstract
Nonpro®t agencies face the evaluation dilemma of how to adequately assess outcomes and build and evaluate the program
from within using limited resources. This article presents an alternative evaluation/management model for nonpro®ts that calls
for an expanded role for the internal evaluator. For the past 5 years, a trained internal evaluator has conducted on-going
evaluations of multifaceted service programs while simultaneously serving on the leadership team of a nonpro®t agency. The
self-re¯ective nature of the agency enhances evaluation use and organizational learning, but creates unique conditions under
which to conduct evaluation studies. # 1999 Elsevier Science Ltd. All rights reserved.
Keywords: Internal evaluation; Self-re¯ection; Organizational learning
1. Introduction
Internal evaluations provide learning organizations
with invaluable information for program development
when results are available to and fully understood by
stakeholders. A major function of the internal evalua-
tor is to facilitate their understanding, and in so doing,
he or she may assume many roles within organizations
(Love, 1991). Most internal evaluator roles are rele-
gated to data collection, analysis and feedback in sup-
port of management's decision-making (Love, 1991;
Mathison, 1991a). However, internal evaluators' roles
could be expanded to include management because
evaluators are well placed for serving multiple capabili-
ties and are su�ciently sensitive to multiple perspec-
tives at all levels of sta� (Lyon, 1989). Cli�ord and
Sherman (1983) encouraged evaluators to think of
themselves as managers with experience in evaluation,
basic research and planning because those skills would
make the evaluator appealing to the organization and
would increase the likelihood of participating in pro-
gram planning.
When internalized into the organization's oper-
ational systems, evaluation guides and models individ-
ual, group and organizational learning (Torres,
Preskill & Piontek, 1996). Learning in organizations is
the interaction of members' learning and the organiz-
ation's capacity to nurture and facilitate learning and
change. Learning occurs when members are given
access to information and encouraged to re¯ect on
their own values, beliefs, and assumptions at the same
time that they process this information and act on
their newly constructed knowledge. Levels of learning
vary (Cousins & Earl, 1995a; Huber, 1991) according
to the degree of self-re¯ection engaged by the mem-
bers. Argyris and SchoÈ n (1978) described three levels
of learning in organizations. The lowest, single-loop
learning in which individuals change incrementally in
reaction to stimuli, leads to instrumental learning, but
has little to do with self-re¯ection or organizational
learning. Double-loop (generative) learning occurs
when members articulate and re¯ect upon deeply held
assumptions about their work and can lead to shifts in
organizational structure and practiceÐwhat Senge
(1990) calls mental models or maps for the organiz-
Evaluation and Program Planning 22 (1999) 353±362
0149-7189/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S0149-7189(99 )00029 -4
www.elsevier.com/locate/evalprogplan
p
A portion of this paper was presented at the annual meeting of
the American Evaluation Association in San Diego, November 1997.
* Tel.: +1-214-941-9192; fax: +1-214-946-7140.
E-mail address: [email protected] (A.M. Minnett)
ation's work. The highest level of organizational learn-
ing, labeled `duetero learning', involves the organiz-
ation's capacity to learn how to learn. To achieve
double-loop or deutero learning, it is critical that
appropriate structures, mechanisms, and systems be
developed that enable organization members to engage
in critical re¯ection and dialogueÐa shared re¯ection
(Preskill & Torres, 1999; Wheatley, 1994).
The key linkage between individual and organiz-
ational learning occurs when evaluation ®ndings are
shared with members, and they engage in a shared
re¯ection about practice. Here, evaluation becomes
part of the job itself, requiring ongoing re¯ection
about performance and resulting in enlightenment
(Owen & Lambert, 1995). Indeed, a learning organiz-
ation renews itself through self-re¯ection. Evaluators
are in a unique position to inform the mental model of
leaders at any stage of the development of a given pro-
gram (Shu�ebeam, 1983) and in doing so, in¯uence
decision-making and organizational learning. By
enhancing the quality of decisions made within the or-
ganization, the internal evaluator helps build a culture
of self-re¯ection (Torres et al., 1996).
Real organizational payo� happens when participa-
tory evaluation becomes fully integrated into organiz-
ational learning (Cousins & Earl, 1995b) because
people in the organization will learn by getting
involved in evaluation work themselves, and evalu-
ation will be one of the major instruments of learning
(Forss, Cracknell & Samset, 1994). The processes of
evaluation support change in organizations by getting
people to think empirically and teaching them to use
data-based decision-making. One advantage of this
process use of evaluation has immense potential for or-
ganizational learning in that it o�ers participants
opportunities to be re¯ective and discriminating about
their practice (Patton, 1997; Shulha & Cousins, 1997).
Another advantage is that evaluators will likely
become more involved in program planning and devel-
opment (Bickman, 1994). And ®nally, evaluation can
be woven into program practice because the two are
planned simultaneously. Patton (1997) calls this devel-
opmental evaluation where the evaluator collaborates
to conceptualize, design, and test new approaches in
long-term processes of continual development.
Unfortunately, pressing needs within the nonpro®t
sector limit evaluation use. Funders are increasingly
concerned about their grantees' accountability and are
requiring additional evidence that programs do what
they claim to do and that their services make a di�er-
ence in clients' lives (e.g., Plantz, Greenway &
Hendricks, 1997). Nonpro®t agencies have ®nite
resources, most of which are targeted for direct ser-
vices, yet their directors face the evaluation dilemma
of how to adequately assess outcomes and build and
evaluate the program from within. The Independent
Sector believes that regardless of limited resources,
evaluation is always possible and each nonpro®t
agency must be concerned with ``how to see that it
gets done and how it can best ®t within the organiz-
ation's context'' (Mayer, 1992). One way to see that
evaluation gets done is to encourage program man-
agers to assume the evaluator role in their organiz-
ations. A recent trend in evaluation treats evaluation
as a leadership function of all managers and program
directors (Patton, 1997), and certainly within smaller
nonpro®t agencies evaluation concerns typically fall to
the executive director or program administrators. As
nonpro®ts come under greater scrutiny for account-
ability for funding, we have seen evaluation increase in
importance while the evaluator's role is assumed by
practitioners and managersÐpersons who often lack
the requisite technical skills and methodological and
analytical expertise, not to mention the interest in such
matters.
For the past 5 years, a trained internal evaluator
(IE) has conducted on-going evaluations of multifa-
ceted service programs in a nonpro®t agency that
values and encourages self-re¯ection. The IE conducts
evaluations in a ¯uid environment of interaction, feed-
back, and change while simultaneously serving on the
agency's leadership team. This article is o�ered as an
example for evaluators to assume multiple roles and
help resolve a pressing administrative need in the non-
pro®t sector.
2. The nonpro®t agencyÐa self-re¯ective organization
The Salesmanship Club Youth and Family Centers,
Inc. (SCYFC) provides a variety of comprehensive
programs to children and families residing in the
Dallas area. The oldest program is the Salesmanship
Club Youth Camp, a year-round, 72-bed residential
treatment facility for adolescents with behavioral and
emotional disturbances (GAO, 1994; Loughmiller,
1965). The largest program, Outpatient Family
Therapy, serves around 2500 Dallas County children
and family members each year and also conducts a
training institute for mental health professionals. The
J. Erik Jonsson Community School serves impover-
ished children and families living in an inner city
neighborhood that is predominantly Hispanic. Social
support services and parent education, skills training,
citizenship, and ESL classes are also o�ered and
shared between the three service programs. The
Research and Evaluation Program (R&E) is separate
from, yet equal to, the direct service programs in the
organizational structure.
SCYFC's benefactor, The Salesmanship Club of
Dallas, is a self-re¯ective organization that long ago
recognized the importance of research-based program
A.M. Minnett / Evaluation and Program Planning 22 (1999) 353±362354
planning and decision making.1 The membership con-
ducts self-studies in which every member participates
via questionnaire or focus group interview to deter-
mine long-range planning for the Salesmanship Club
and the agency. For example, nearly 50 focus groups
were held in the past year before the membership
voted on (and passed) a 6 million dollar capital cam-
paign for the agency's expansion. Their thorough
re¯ection on procedures and governance serves as a
model and often challenges the agency to re-evaluate
the goals and practices of the service programs.
Because we believe that self-re¯ection, based upon
evaluation, leads to program improvement, evaluation
is built into every service program and its notion is re-
inforced by the continual presence of research and
evaluation projects and personnel. Self-re¯ection is
woven into all agency processes as we continually
question how to improve our work (``How does our
work a�ect our clients? Are we providing the best level
of service possible to our clients? If not, how can we
improve?''). Every employee is hired with the knowl-
edge that SCYFC values self-re¯ection and continuous
program development as part of our mission, and all
are invited to participate. The pervasive insistence on
accountability, growth and development, and emphasis
on research and evaluation are powerful indicators of
the agency's value for self-re¯ection and learning.
3. Internal evaluation at SCYFC
The internal Research and Evaluation (R&E)
Program has a full-time director (the IE) and one or
two graduate student assistants, an independent budget
and individual programmatic mission. The R&E pro-
gram was created in 1993 to support and facilitate
SCYFC's service program development and account-
ability through grounded investigations in collabor-
ation with service program experts. We also have a
mission to disseminate what we have learned to col-
leagues and potential collaborators. The structure pro-
vides exclusive, independent status for R&E while
simultaneously including R&E in the agency's oper-
ational organization. An External Evaluator (EE) has
conducted long term outcome evaluations of the resi-
dential treatment program since 1983 and community
school since 1996.2 The evaluators share technical
expertise and occasionally consult with one another on
methodological issues. Some data derived from in-
ternal evaluation studies are available to the EE as
baselines for his outcome evaluations.
The Internal Evaluator functions as evaluator and
administrator with responsibilities that fall into three
categories: designing and implementing evaluation stu-
dies and supervising others' research; participating in
the agency's leadership team; and representing the
agency as an administrator.
3.1. The process and role of evaluation
The IE is a participant observer who collaboratively
engages in applied social research with practice-based
experts. The evaluation process incorporates both
deductive and inductive approaches and uses multiple
methods, perspectives and measures for continuous
program development. Our model is similar to focusing
or evaluability assessment (Patton, 1997) in that the IE
often guides program sta� in their visioning process by
keeping them grounded in clarifying goals, operation-
Fig. 1. Research and evaluation process for continuous program development.
1 The Salesmanship Club of Dallas, a 75-year old group of 500
business and civic leaders, operates the SCYFC as a charity. Yearly
proceeds from the GTE-Byron Nelson Classic, a PGA golf tourna-
ment, provide for 93% of SCYFC's annual $4.5 million operational
budget. Remaining funds are derived from client fees, The United
Way and foundation grants and awards.2 J. Michael Coleman, Ph.D., is SCYFC's External Evaluator and
is currently the Dean of Undergraduate Studies at The University of
Texas, Dallas.
A.M. Minnett / Evaluation and Program Planning 22 (1999) 353±362 355
ally de®ning practices and answering the question
``What would that look like?'' or ``How will you know
when you're successful?''. With the program mission,
beliefs, values and goals in place, the working relation-
ship between sta� and R&E becomes more collabora-
tive, and the evaluator experiences a greater role in the
process. (see Fig. 1.) Typically, the evaluation loop
begins when members question the e�ectiveness of a
practice (Box 1) and believe it can be improved. The
sta� and IE collaborate (Box 2) to de®ne parameters
of a study to evaluate the e�ectiveness of the practice.
Next, the IE re®nes the design, that is, determines who
will be involved, the methods and procedures for the
data collection and data analysis, and the duration of
the study (Box 3). After sharing the evaluation design
with the sta� to insure that it melds with their practice
and that it is doable, the program practice and evalu-
ation study are implemented simultaneously (Box 4)Ð
a process that lasts anywhere from one week to as
long as one year, depending on the evaluation ques-
tion. The data are then analyzed and synthesized by
the IE (Box 5) who subsequently reports the results
back to the sta� (Box 6). When practitioners are satis-
®ed with the results, practice remains unaltered (Box
7a), and the study ends. However, when results indi-
cate that further re®nements are needed (Box 7b), they
make adjustments based upon their experience and the
evaluation results (Box 8a). We then implement identi-
cal methods and procedures to evaluate the altered
practice. The loop continues until program experts are
satis®ed with results and adopt the practice (Box 7a)
or until they decide to abandon the practice (Box 8b).
The IE designs and implements evaluation studies
for the community school to evaluate short-term out-
comes and investigate practice. Comprehensive longi-
tudinal studies are in place to track the cognitive and
social development of each student and the status of
their families across the course of their tenure at the
school. Several process evaluations of practice in curri-
culum development, parent involvement and integrated
team teaching are currently in progress. A series of
studies of procedures in our residential program are
also in progress, and we have recently created and im-
plemented an evaluation of short term e�ects of family
therapy services on clients' sense of self-e�cacy.
Other evaluation-related functions include oversight
of the agency's Institutional Review Board, dissemina-
tion of results not only to sta� but to colleagues and
potential collaborators, and information retrieval from
outside sources about exemplary practices and demo-
graphics that relate to our programs. Finally, the IE
provides seminars and workshops to other nonpro®t
agencies about evaluation techniques and designs. The
IE is not required to write grants or participate in fun-
draising activities, but would likely do so in other non-
pro®t settings.
3.2. The evaluation process within leadership
We believe that evaluation procedures should be
built into program procedures and that evaluation
®ndings should be used for the betterment of the pro-
grams. The IE uses a procedure similar to
Sonnichsen's (1988, p. 142) Advocacy Evaluation,
which states that evaluators must view themselves as
change agents and step away from the position of neu-
tral observer to that of active participant in the de-
cision making process. His model suggests that the IE
be actively involved with supervisors in the organiz-
ational process of discussion, approval, and implemen-
tation of recommendations. Contrary to those who
believe that advocacy compromises the evaluator's
objectivity (Scriven, 1997; Stake, 1997), Sonnichsen's
view neither advocates programs nor supports admin-
istrators, but acts as an ``independent advocacy for
needed change, arrived at through the conduct of an
objective, valid, defensible evaluation of the program''.
Indeed, Sonnichsen believes that advocacy begins only
after the completion of the evaluation study.
We agree. Involvement in program planning and
development is a critical feature of the IE's role at
SCYFC. We believe that objectivity and neutrality
should guide the design and implementation of the
evaluation study to the extent possible. In all candor,
it would be very di�cult, if not rather odd, for the IE
to conduct long term participant evaluations in this
agency and retain neutrality. Siegel and Turkel (1985)
call that sort of behavior ``dysfunctional'' and argue
that evaluation without advocating recommendations
renders internal evaluation ine�ectual.
The IE functions as a leader in the organization by
advocating for evaluation results and providing an
evaluator's perspective on issues. She brings evaluation
®ndings to the planning table while other leaders bring
their programmatic expertise and the practical limi-
tations that they face in implementing their services.
The leadership team (with the IE as a member) builds
evaluation into program planning and implementation
so that practitioners implement their programs while
the evaluator conducts studies that are woven into the
services. For example, family therapy clients provide
periodic assessment of the status of problems that
brought them to therapy, and the information is acces-
sible to the therapist (with the client's permission) for
re¯ection. The IE then analyzes and synthesizes the
data and provides the leadership team with thick
descriptions along with quantitative results that are de-
rived from sound studies. We believe that the IE's syn-
thesis of the data is enhanced by her thorough
knowledge of the issues and participants. As a result,
the IE is positioned to facilitate evaluation use and fol-
low-through, program re®nement occurs routinely, and
the process of re¯ection on practice is accorded im-
A.M. Minnett / Evaluation and Program Planning 22 (1999) 353±362356
portance. Issues of objectivity and neutrality do not
pose a threat to this model because objectivity and
neutrality guide the conduct of the study, while advo-
cacy guides the IE's recommendations (Sonnichsen,
1988).
3.3. Leadership roles unrelated to evaluation
Organizationally, the IE serves as ex-o�cio consult-
ant in Youth Camp planning meetings with regard to
sta� training and revisiting its organizational structure
(mission, values, beliefs and goals). That same function
exists in current Outpatient Family Therapy discus-
sions of accountability as that program collaborates
with other agencies to create a managed care network.
The most concentrated participation occurs within the
school where the IE is a member of the school admin-
istration. Finally, the IE is a member of the agency's
Leadership Council, representing the R&E Program
but helping to determine agency policy. For all three
service programs and the agency as a whole, the IE is
a program developer and decision-maker working col-
laboratively with others. The IE also represents the
agency (as an administrator) to other groups in a var-
iety of contexts such as community development com-
mittees or nonpro®t networks.
4. Issues that shape internal evaluation at SCYFC
4.1. Continuing relationships
Among the unique components of this model is the
IE's continuing, multifaceted relationship with evalu-
ation participants who are at once subjects, stake-
holders and collaborators over long periods of time.
First, the longevity of the role alone means that the IE
is accountable to colleagues for previous recommen-
dations and therefore assumes some responsibility for
outcomes. It follows that the IE has grown more
invested in and committed to the organization's devel-
opment. Put simply, the IE has a strong desire to see
the organization have positive e�ects on clients.
Love (1991) describes the complimentary, if not con-
tentious, perspectives of program managers and eva-
luators and elaborates the stereotypes one usually
holds of the other to describe when data are ignored
(in the IE's view) or when real-world constraints prohi-
bit action (manager's view). We believe that these dif-
fering perspectives are healthy in the self-re¯ective
organization. Mathison (1991b) and others have writ-
ten about the internal evaluator's vulnerability to co-
optation where administrators exert pressures on the
evaluator by relating rationale for behavior or back-
ground information for decisions they make in their
practice, usually justifying why recommendations
could not be followed. The realities of program im-
plementation and political compromise do contribute
to the IE/leader's understanding of the issues and syn-
thesis of the data. Thus, the real danger for the IE is
not co-optation from others, but a spontaneous loss of
evaluation perspective (Browne & Wildavsky, 1983). It
is unavoidable that the IE eventually becomes part of
the culture and loses some perspective in this model,
so it is especially critical that we capture multiple per-
spectives (Greene, 1997), including new sta� or visitors
who may question long-held agency beliefs or prac-
tices.
Because the R&E sta� engage in long term, colla-
borative relationships, we are constantly mindful of
gaining and retaining the trust of individual sta� at all
levels. First, we rarely use the word ``evaluation'' with
sta�Ðthey know of the program and our work simply
as ``research''.3 Evaluation has negative connotations
for many and appears to give too much power to the
IE when we would prefer to engage in a collaborative
work environment. Second, any information provided
by sta� or clients to an R&E study is treated as con®-
dential by the researchers. In program development
work, agency sta� are the IE's clients, and everyone
who contributes to the evaluation project (sta�,
families, children) are the study subjects and are
accorded rights (i.e., Joint Committee on Standards
for Educational Evaluation, 1994). All results are
reported on a group basis, and no one individual is
ever identi®ed in written reports or in verbal communi-
cation to any stakeholder group, unless the IE judges
that the individual is causing harm to another. Third,
the IE does not conduct or participate in personnel
evaluations other than for R&E sta�. Fourth, we are
respectful about the amount and type of information
we request from individuals. In this setting where most
children, families and sta� participate in R&E projects
over long periods of time, we avoid requesting a great
deal of information ``just in case'' we might need it
later. Fifth, we collaborate with sta� at all levels, in
part so that they learn to trust the process and our
work. More importantly, while the evaluation sta� is
trained in research and methodological design, we
value the expertise of teachers, therapists, social
workers, support personnel and administrators and
rely on them to help us design and conduct e�ective,
meaningful evaluations. Without trust, our entire colla-
3 A later section of this article describes the necessity for the in-
ternal evaluator to speak the practitioner's language. We use the
term research to meaningfully describe the collaborative process of
re®ning practice and developing programs because it is helpful for
practitioners to understand the work in that way. Therefore, we see
no useful purpose in discussing the internal evaluation process with
practitioners in evaluation terms, but rather focus on the process as
a means to program development.
A.M. Minnett / Evaluation and Program Planning 22 (1999) 353±362 357
borative model would fail and the bene®ts of evalu-
ation use would disappear.
We tripped along a rocky road to trust. For
example, 4 years ago, the school was engaged in exten-
sive restructuring as its focus shifted from treatment of
children with emotional and behavioral disturbances
to a community school that emphasizes academic
excellence for inner city children. The IE designed a
comprehensive study to evaluate the appropriateness
of former practices with the new population and then
enthusiastically presented the plan to the entire sta�.
Two teachers cried during that meeting, several more
were visibly shaken, and a contingent of teachers later
complained to administrators that they were unpre-
pared for such intrusions into their practice. R&E in
the school was derailed for an entire year. The evalua-
tor used that time to engage teachers individually, sup-
porting each in unrelated action research projects of
their choosing in order to regain their trust and co-
operation in the larger study (which we began the fol-
lowing year). On another occasion, we were required
to initiate an outcome evaluation of family therapy
services, which are notoriously di�cult to evaluate
(Seligman, 1995). After engaging in rather heated
debates about the need for and merits of outcome
evaluations, the IE and therapists came to an under-
standing that we would collaboratively design the most
palatable study as possible. As the pilot study
unfolded, therapists suggested procedural changes,
which we made. When the pilot data were presented to
the group, they could see the value of the information
and were more receptive to the design revisions and
subsequent evaluation study.
Di�culties such as these forced the R&E program
to be more collaborative and to respect participants'
opinions and expertise (Cousins & Earl, 1995a;
Greene, 1988; Preskill & Torres, 1999). We have found
that this attitude engenders their trust and that they
become more participative in our work. ``Trusting the
process'' as we do would not be possible without the
support of a very secure leadership team which has
opened the agency to constant, pervasive evaluation
and self-re¯ection (Senge, 1996).
4.2. Communication roles and functions
The evaluator translates evaluation results into prac-
tical applications in three ways. First, the R&E team
investigates, summarizes and presents exemplary prac-
tice from a variety of disciplines to our sta�.
Practitioners decide its value and then assimilate those
®ndings into their practice to re®ne the program.
While this appears as instrumental learning on the sur-
face, this information sometimes precludes an evalu-
ation study and self-re¯ection.
Second, and most often, the IE communicates
results from internal studies to our self-re¯ective prac-
titioners using the professional language that is most
meaningful to them. For example, SCYFC's family
therapists mull over decisions based on theoretical con-
structs, abstract concepts and empathic acceptance for
clients. They eschew labeling diagnoses and are most
receptive to hearing results about clients' perspectives
and how clients managed to change their own lives.
On the other hand, community school teachers are
grounded in sequential events, details and goals for
learning. They are exceptionally interested in strategies
that will help them improve practice in the classroom,
such as integrating curriculum, including parents in
their children's learning and time management, for
example. The action-oriented residential treatment
team, deals with a volatile population and makes criti-
cal split-second decisions constantly. They espouse a
solution-focused approach to practice with children
and families, and want direct, succinct information
drafted in language that focuses on the solution rather
than the problem. Any good evaluator should alter
communication strategies to ®t their audiences' con-
cerns (Torres et al., 1996). SCYFC's model simply fa-
cilitates the quality of information provided by the IE
because the IE is well grounded in the work and
languages of the agency (Goering & Wasylenki, 1993).
The third way in which the IE functions as commu-
nicator in this multi-disciplinary agency is by acting as
translator between programsÐusually at the time feed-
back is presented and always during program plan-
ning. The IE who works with all programs and has
leadership responsibilities is able to provide feedback
in the context of organizational culture and in so
doing, informs others about larger issues in the
agency.
There are also opportunities to make smaller con-
nections. Recently, the IE conducted telephone inter-
views of selected families who had dropped out of the
admissions process for our residential treatment pro-
gram. Over 1100 families express interest in the pro-
gram each year, but the camp is able to serve about
one-®fth that number. The remainder are referred else-
where, ®nd other placement, or simply withdraw from
the process. We were interested in the reasons that
families decide to discontinue the admissions process
and what happens to them afterward. We found that
some families dropped out of admissions because their
child had a ``scared straight'' experience regarding the
possibility of residential treatment. Their family pro-
blems subsequently lessened in severity, and they
reported that they needed something less intrusive
than residential treatment for their child. These
families are now a source of referrals for our
Outpatient Family Therapy program.
There is an additional component to communication
that has no name and is not included in a job descrip-
A.M. Minnett / Evaluation and Program Planning 22 (1999) 353±362358
tion. Self-re¯ective organizations bene®t from a ``dis-
cusser of the undiscussables'' (Senge, 1997)Ðsomeone
who can ask hard questions. Mathison (1991a, p. 177)
stated that internal evaluation can maintain its integ-
rity only when evaluators see the big picture and dare
to ``pursue potentially unpopular issues''. SCYFC's IE
is perfectly positioned to pose questions that appear
naõÈ ve to everyone but to those for whom the issue is
undiscussable. This role is permitted because the IE is
both part of the system that is questioned and yet ade-
quately set apart to be moderately objective.
The evaluator in this model acts as one of several
liaisons/communicators to the Board and other fun-
ders. Occasionally, this includes formal presentation of
program outcomes but most often involves describing
and explaining a complex set of programs and services
to Board members and others. Our funders are
business leaders who are comfortable with succinct,
bottom-line bullet points about client status, organiz-
ational process, and the status of education and social
services. The external evaluator creates annual reports
of long term follow-up evaluations of the residential
treatment and school programs. The internal evaluator
submits occasional reports about the residential pro-
gram's development and annual reports of the
Outpatient Family Therapy outcomes evaluations, the
school's continuous program development and the sta-
tus of research and evaluation at SCYFC. The
research and evaluation report addresses accountability
issues for R&E and recounts the year's evaluative pro-
cesses.
4.3. Evaluation studies and information are accorded
importance
When our board of directors formed a separate
R&E program and placed the IE in a leadership pos-
ition, the agency ``walked the walk'' and reinforced to
sta� and others that R&E is crucial to the agency's
success. In order to compensate for the scope of our
work and size of R&E sta�, we have established a
community of researchers, similar to Preskill and
Torres' (1999) community of inquirers, in which each
person's expertise and perspective are valued and real
e�ort is made to nurture their self-re¯ection (Minnett,
1998b, 1999). In the spirit of collaborative learning
(Torres et al., 1996), we believe that evaluation
research is everyone's responsibility, and the commu-
nity of researchers allows for that expression. Indeed,
individual sta� are encouraged and recognized for
their research e�orts and participation. We believe that
individuals who feel ownership in the research process
are more likely to assume research responsibilities and
understand the information. We have found that they
also are more likely to make multiple uses of the data
(see below) as well as become more self-re¯ective. The
sta� is also invited to make sense of the ®ndings,
which acknowledges their expertise and likewise
enhances the accessibility of the data.
4.4. We ask meaningful questions
An internal evaluator is more likely to know which
questions are most appropriate and meaningful and
when to ask them than is an external evaluator (Lyon,
1989). The IE in our model has not only the advantage
of an insider, but the added perspective of viewing
issues in terms of larger organizational systems and
processes. This also allows the IE to provide for conti-
nuity of project development where one research ques-
tion builds upon another. And ®nally, the evaluator
provides some evaluative continuity by coordinating
the next set of questions that the leadership asks about
the agency or that program directors ask about their
programs.
An example of the value of continuity in evaluation
for continuous program development has occurred in
the school. The community school is developing a
team-teaching model in which each self-contained
classroom has two certi®ed teachers who integrate all
subjects throughout the day.
. Our ®rst study identi®ed two successful teams and
examined in depth their perspectives on their work
together (Minnett, Kaye, Bryant, Wetzel &
Camacho, 1997). Teachers were interviewed about
philosophies of education, strategies, training, atti-
tudes about teaching and working relationships with
their partner and the administration.
. The second study used what we learned about tea-
chers' views and observed how they interacted in the
classroom (Minnett, 1998a). Each pair of teachers
was observed repeatedly for one semester as they
engaged in a variety of classroom activities to see
how teachers communicated with one another, nego-
tiated, amount of time spent with students and div-
ision of responsibilities. Teachers were also observed
as they planned.
. A third study examined qualities/characteristics and
criteria for partnering teachers using qualitative
interviews of all school personnel.
. The fourth study compared what we have learned
in-house with what others have done vis-aÁ -vis team-
ing and integrated curriculum, and we concluded
that our model of teaming was inappropriate for
junior high grades and warranted change. New
teaming practices were developed to acknowledge
junior high teachers' expertise (they are not general-
ists as are elementary teachers on whom the model
was patterned) as well as adolescents' developmental
needs for broader social a�liations.
. The next study investigated the e�ects of new pro-
A.M. Minnett / Evaluation and Program Planning 22 (1999) 353±362 359
cedures in 7th and 8th grades, and results showed
continued di�culty with the model for this age
group. As a result of our work and the agency's
increasing interest in early intervention, grades 7
and 8 were dropped from the school at the close of
the 1997±1998 school year and a new K±3 program
was installed in the school.
. Our current study of teaming includes all teachers
and administrators in creating a Teaming Pro®leÐ
the operational de®nition of teaming. They have
identi®ed processes associated with teaching (sche-
duling, meeting special needs of students, etc.) and
described the necessary skills, relationships and
resources to accomplish each activity in a teaming
environment.
. Once the Teaming Pro®le is complete, we will video-
tape indicators of good practice in teams and study
concurrent e�ects of teaming practice on student
learning. The entire sequence has taken less than
three years and has incorporated several small stu-
dies into the larger investigation of teaming. This
work will contribute to making teaming more e�ec-
tive for everyone by applying the research process
for continuing program development to re®ne the
practice. And ®nally, the pro®le will form the basis
of a new personnel evaluation system currently
under development.
Not only are questions more appropriate and perhaps
more planful as the example above demonstrates, they
are more creative because we are able to adopt an
inductive, generative approach to investigating prac-
tice. Social service programs are moving targets, and if
evaluation data are meaningful, they must be current.
Because R&E is part of program planning, our readi-
ness to ask questions puts R&E into the mix at an
early stage of planning which is critical to the process
of evaluation and program planning. Further, self-
re¯ection is encouraged and study questions are
actively solicited from all participants. We therefore
bene®t from multiple perspectives from the very begin-
ning of the research and evaluation process.
4.5. Multiple uses for meaningful data
Our evaluation process often, but not always, results
in greater utilization of the data. It stands to reason
that data, which are understood by practitioners and
administrators, are more likely to be utilized, and that
having access to a researcher who understands the
data will increase the likelihood that program sta� will
understand the results. It is possible that the manner
in which the data are framed will increase their use,
and again, the evaluator in leadership will have greater
insights into the potential use of the data and can an-
ticipate the need. More importantly, the IE and each
participating sta� member may make practical appli-
cations of the data in distinct ways. For example, for
the past 5 years every person on the school sta� has
recorded their contacts with parents as one measure of
parent involvement at school (e.g., Minnett, 1996).
School personnel record who initiates the contact, how
it is made (note, phone, incidental meeting, conference,
participation in a school activity), and the tenor of the
conversation (positive focus, problem focus or infor-
mation exchange). Everyone contributes to the data-
base, which provides a wealth of information about
parent participation and the e�ects of certain practices
on parent involvement, the intended uses for the data.
The IE originally provided feedback to teachers pri-
marily as a courtesy to acknowledge their e�orts in
this large, on-going project. However, teachers
revealed that they used the organized printout (by
child, showing contacts in sequence, and the nature of
the contact) to determine which families had been
overlooked or to detect troubling patterns in school-in-
itiated contacts with their students' families. The data
have shaped their practice.
Another use for data involves client data feedback
to practitioners. The IE always adheres to con®dential-
ity when acquiring information from clients but o�ers
them the option of agreeing to release the information
to their practitioner. Clients often agree to release this
information and it is surprising how important these
data are to practitioners. We ®nd that both positive
and negative feedback enhance practitioners' re¯ec-
tions and are welcomed.
4.6. Evaluator skills
Internal evaluators working in this environment
must have technical and methodological skills and
should be able to assume a mentoring role with the
community of researchers. Because we engage in par-
ticipant evaluation, IE's conducting this type of work
should be motivated to participate in organizational
activities and have tolerance for imperfection (Cousins
& Earl, 1995b). Due to our emphasis on collaboration,
few skills are as essential as ``people skills'' (Ayers,
1987). Social skills allow internal evaluators to build
connections and trust, especially with those who are
skeptical of their presence. We have found those qual-
ities are essential in building trust among participants
at the agency, gaining cooperation and buy-in from
the sta�, and communicating negative ®ndings while
maintaining long term relationships with colleagues.
We are a social service agency that places high value
in respect for clients. It is therefore important that the
IE engages clients, student, or family members in the
same respectful manner espoused by the practitioners.
That means knowing a little about the services, how
A.M. Minnett / Evaluation and Program Planning 22 (1999) 353±362360
clients access them and being able to interact with
them informally.
Goering and Wasylenki (1993) argued that substan-
tive program expertise is also an asset because it facili-
tates the evaluator's reconciliation of a host of
program and organizational context variables that
impact the use of evaluation information. Our IE had
previous experience in educational settings but was less
adept in mental health and community development
issues. It was imperative for the IE to learn their
language, observe their work, and acknowledge the
fact that she would conduct evaluations on their turf.
5. Conclusion
Our evaluative procedures lend themselves to non-
pro®t settings where employees may assume multiple
roles. However, our example is unique because of the
substantial support of our funders for R&E, their
insistence upon continuing program development, and
the pervasive nature of evaluation in our organization.
Self-re¯ection is an isomorphic process permeating our
agency through every organizational level and pro-
gram. We are encouraged to explore alternatives and
re®ne practice through research and evaluation, which
further nourishes our self-re¯ection. This article is
simply another exampleÐan internal evaluator re¯ects
about R&E and the unique opportunity to function in
the dual role of internal evaluator and leader.
Acknowledgements
The author would like to thank the reviewers who
made substantial contributions to this article and
thank Richard Sonnichsen for his comments on an
earlier draft.
References
Argyris, C., & SchoÈ n, D. (1978). Organizational learning. Reading,
MA: Addison±Wesley.
Ayers, T. D. (1987). Stakeholders as partners in evaluation: a stake-
holder-collaborative approach. Evaluation and Program Planning,
10, 263±271.
Bickman, L. (1994). An optimistic view of evaluation. Evaluation
Practice, 15(3), 255±259.
Browne, A., & Wildavsky, A. (1983). Should evaluation become im-
plementation?. In: A. J. Love, Developing e�ective internal evalu-
ation. New directions in program evaluation, vol. 20). San
Francisco: Jossey±Bass.
Cli�ord, D. L., & Sherman, P. (1983). Internal evaluation: integrat-
ing program evaluation and management. In: A. J. Love,
Developing e�ective internal evaluation. New directions for program
evaluation, vol. 20). San Francisco, CA: Jossey±Bass.
Cousins, J. B., & Earl, L. M. (1995a). The case for participatory
evaluation: theory, research, practice. In: J. B. Cousins, &
Bradley, Participatory evaluation in education: studies in evaluation
use and organizational learning. London: Falmer Press.
Cousins, J. B., & Earl, L. M. (1995b). Participatory evaluation in
education: What do we know? Where do we go?. In:
J. B. Cousins, & Bradley, Participatory evaluation in education:
studies in evaluation use and organizational learning. London:
Falmer Press.
Forss, K., Cracknell, B., & Samset, K. (1994). Can evaluation help
an organization to learn?. Evaluation Review, 18(5), 574±591.
Goering, P. N., & Wasylenki, D. A. (1993). Promoting the utilization
of outcome study results by assuming multiple roles within the
organization. Evaluation and Program Planning, 16, 329±334.
GAO Government Accounting O�ce (1994). Residential care: some
high-risk youth bene®t, but more study is needed. Report to the
Chairman, Subcommittee on Oversight of Government
Management, Committee on Governmental A�airs, US Senate.
Greene, J. C. (1997). Evaluation as advocacy. Evaluation Practice,
18(1), 25±35.
Greene, J. C. (1998). Stakeholder participating and utilization in pro-
gram evaluation. Evaluation Review, 18(5), 574±591.
Huber, G. P. (1991). Organizational learning: the contributing pro-
cesses and the literature. Organizational Science, 2(1), 88±115.
Joint Committee on Standards for Educational Evaluation (1994).
The program evaluation standards: how to assess evaluations of
educational programs, (2nd ed.). Thousand Oaks: Sage
Publications.
Loughmiller, C. (1965). Wilderness road. The University of Texas,
Austin: The Hogg foundation for mental health.
Love, A. J. (1991). Internal evaluation: building organizations from
within. Newbury Park: Sage Publications.
Lyon, E. (1989). In-house research: a consideration of roles and ad-
vantages. Evaluation and Program Planning, 12, 241±248.
Mathison, S. (1991a). Role con¯icts for internal evaluators.
Evaluation and Program Planning, 14, 173±179.
Mathison, S. (1991b). What do we know about internal evaluation?.
Evaluation and Program Planning, 14, 159±165.
Mayer, S. E. (1992). Common barriers to e�ectiveness in the inde-
pendent sector. In: Proceedings of Independent Sector's Annual
Meeting and Assembly of Members, Minneapolis, MN.
Minnett, A. M. (1996). Longitudinal study of teacher-parent contacts
and parents' subsequent involvement at school. In: Proceedings of
the Annual Meeting of the American Educational Research
Association, New York.
Minnett, A. M. (1998a). Two-teacher teaming in elementary class-
rooms: a pilot study. National Association of Laboratory Schools
Journal, XXII(1), 9±11.
Minnett, A. M. (1998b). Establish a community of researchers and
transform the process of internal evaluation. In: Proceedings of
the Annual Meeting of The American Evaluation Association,
Chicago.
Minnett, A. M. (1999). Building a community of researchers to
improve teaching practice. In: Proceedings of the Annual Meeting
of the Association for Supervision and Curriculum Development,
San Francisco.
Minnett, A. M., Kaye, B., Bryant, H., Wetzel, M., & Camacho, G.
(1997). Two teachers in the classroom: process and outcomes of
collaborative teaming. In: Proceedings of the Annual Meeting of
Texas Association for Supervision and Curriculum Development,
Houston.
Owen, J. M., & Lambert, F. C. (1995). Roles for evaluators in learn-
ing organizations. Evaluation, 1(2), 237±250.
Patton, M. Q. (1997). Utilization focused evaluation: the new century
text, (3rd ed.). Thousand Oaks: Sage Publications.
Plantz, M. C., Greenway, M. T., & Hendricks, M. (1997). Outcome
measurement: showing results in the nonpro®t sector. In:
A.M. Minnett / Evaluation and Program Planning 22 (1999) 353±362 361
E. Chelimsky, New directions for evaluation, vol. 75 (pp. 15±30).
San Francisco: Jossey±Bass.
Preskill, H., & Torres, R. T. (1999). Evaluative inquiry for learning in
organizations. Thousand Oaks: Sage Publications.
Scriven, M. (1997). Truth and objectivity in evaluation. In:
E. Chelimsky, & W. R. Shadish, Evaluation for the 21st century
(pp. 477±500). Thousand Oaks: Sage Publications.
Seligman, M. E. P. (1995). The e�ectiveness of psychotherapy: the
consumer reports study. American Psychologist, 50(12), 965±974.
Senge, P. M. (1990). The ®fth discipline: the art and practice of the
learning organization. New York: Doubleday Currency.
Senge, P. M. (1996). Leading learning organizations: the bold, the
powerful, and the invisible. In: F. Hesselbein, M. Goldsmith, &
R. Beckhard, The leader of the future (pp. 41±57). San Francisco:
Jossey±Bass.
Senge, P. (1997). Comments during satellite discussion of issues in
education at the Annual Meeting of the Texas Association for
the Supervision of Curriculum Development, Houston, TX.
Shu�ebeam, D. A. (1983). The CIPP model for program evaluation.
In: G. F. Maddus, M. Scriven, & D. L. Shu�ebeam, Evaluation
models: viewpoints on educational and human services evaluation.
Boston, MA: Kluwer±Nijho�.
Shulha, L. M., & Cousins, J. B. (1997). Evaluation use: theory,
research, and practice since 1986. Evaluation Practice, 18(3), 195±
208.
Siegel, K., & Turkel, P. (1985). The utilization of evaluation
research: a case analysis. Evaluation Review, 9(3), 307±328.
Sonnichsen, R. C. (1988). Advocacy evaluation: a model for internal
evaluation o�ces. Evaluation and Program Planning, 11, 141±
148.
Stake, R. E. (1997). Advocacy in evaluation: a necessary evil?. In:
E. Chelimsky, & W. R. Shadish, Evaluation for the 21st century
(pp. 470±476). Thousand Oaks: Sage Publications.
Torres, R. T., Preskill, H. S., & Piontek, M. E. (1996). Evaluation
strategies for communicating and reporting: enhancing learning in
organizations. Thousand Oaks: Sage Publications.
Wheatley, M. J. (1994). Leadership and the new science. San
Francisco: Berrett±Koch.
A.M. Minnett / Evaluation and Program Planning 22 (1999) 353±362362