Theory-Based Evaluation

In: The Bulletin of The National Center for Zero to Three, Feb/March 1998, p. 30-36.

Theory-Based, Participatory Evaluation:

A Powerful Tool for Evaluating Family Support Programs

by

Beth L. Green, Ph.D1.

Portland State University

and

Carol McAllister, Ph.D.

University of Pittsburgh

In: The Bulletin of The National Center for Zero to Three

April 24, 2002

1
For more information contact: Beth L. Green, 317 Cramer Hall, Portland State University, Portland OR, 97207
ph: (503) 725-3968, email: [email protected].

Green et al., p. 1
4/24/02
 Theory-Based Evaluation

Acknowledgments: We would like to thank the families, staff, and directors of Family Foundations,

Partnerships for Family Support, and First Steps for their dedication in working with us to improve

evaluation methods for Family Support Programs. This work was supported in part by the following

grants: DOE: P252A20082, DHHS: 90YF0002, and support from the Howard Heinz Endowments.

Green et al., p. 2
4/24/02
 Theory-Based Evaluation

Theory-Based, Participatory Evaluation:

A Powerful Tool for Evaluating Family Support Programs

by Beth L. Green, Ph.D. & Carol McAllister, Ph. D

For the past several years, we have been engaged in evaluating community-based family

support centers. These programs present many challenges to evaluators, due to the complex nature of

the services delivered, the programs’ philosophy of family-driven and individualized services, and the

broad range of outcomes that such programs are designed to impact. To address these challenges, we

have combined two evaluation approaches: (1) theory-based evaluation, which is based on careful

articulation of the program model and use of this model as a guiding framework for evaluation; and (2)

participatory evaluation, which involves close collaboration between evaluators and program

administrators, staff, and families in developing, implementing, and interpreting the evaluation. We

believe that these two methods, used in combination, represent a powerful tool for developing valuable,

program-friendly evaluations. In this paper we will begin by describing the nature of these two

approaches to evaluation. We then provide examples of how we have worked with family support

centers to develop and articulate their program “theories” and to use these theories as the basis for

evaluation.

What is a Theory-Based Approach to Evaluation?

In the field of program evaluation, it has long been acknowledged that it is important to

understand the intervention process, including the type, nature, and intensity of services delivered, how

services are experienced by program participants, and how services are expected to lead to outcomes.

Unfortunately, program evaluation that focuses on the intervention process, sometimes called “process

Green et al., p. 3
4/24/02
 Theory-Based Evaluation

evaluation,” often takes a back seat to evaluations focusing solely on program outcomes. Often,

evaluations focus only (or primarily) on participants’ status on some measure or set of measures (such

as knowledge of child development) both before and after the program, and not at all on what happens

to participants during the program. While this outcome-focused approach can be useful for

demonstrating that a program “worked” or “didn’t work,” such evaluations don’t provide any

information about why and how an outcome was achieved (or not achieved). In order to understand

why a program worked or didn’t work, the evaluation has to be informed by an understanding of what

happened during the intervention.

One method of doing evaluation that is based on developing a clear understanding of the

intervention process is called theory-based evaluation (Chen & Rossi, 1983; Weiss, 1972). Theory in

this usage doesn’t always mean a grand theory in the traditional social science sense (e.g., “attachment

theory”), but instead simply refers to a program logic model, or “theory of change” that represents a

“plausible and sensible model of how the program is supposed to work” (Bickman, 1987). Theory-

based evaluation involves identifying the key service components and expected program outcomes, and

working with programs to make explicit the underlying assumptions about how these service

components will lead to the desired outcomes. These services, outcomes, and the hypothesized links

between them are the basis for developing a program model or theory. This program theory becomes

the framework to guide the development, implementation, and interpretation of the evaluation.

What is a Participatory Approach to Evaluation?

The description of theory-based evaluation provided above implies that program theory is

developed in close collaboration with program administrators, staff, and families. Not all theory-based

evaluations, however, rely extensively on this collaborative process (Chen & Rossi, 1983). To promote

Green et al., p. 4
4/24/02
 Theory-Based Evaluation

this collaboration, we have explicitly incorporated aspects of another powerful evaluation method, called

participatory evaluation, into our evaluation approach (Greene, 1987). Participatory evaluation involves

active collaboration between key stakeholder groups in designing, implementing, and interpreting the

evaluation. Stakeholder groups include all those who have a vested interest in the program and its

evaluation, such as funders, program directors, line staff, families, and community members. Our use of

participatory methods has evolved directly from our ongoing evaluation work with family support

programs, which we describe below.

Understanding and Evaluating Family Support Programs

The family support programs that we have worked with share a loosely defined conceptual

basis, but differ substantially in the specific kinds of services that they deliver. Generally speaking, these

programs are physically located in the communities that they serve, are open to the entire community but

targeted to families with young children (age 5 and under), provide a combination of center-based and

home-visiting services, and offer a comprehensive array of services tailored to individual family needs,

such as case management, information and referral, parenting education and support, child development

services, health services, transportation, adult education and employment services, and recreational

activities. Although the specific services may differ in nature, intensity, and mode of delivery, family

support programs do share an explicit value system—an approach to service delivery that is guided by a

shared service philosophy. This philosophy suggests that services are collaborative (characterized by

staff and families working in equal partnership to address family goals); family-centered and driven

(focused on the family as a whole and allowing family members to choose the specific services they will

receive); strengths-based (focused on building on family members’ strengths, rather than on identifying

their problems); and comprehensive and flexible (offering an array of services focused on family’s

Green et al., p. 5
4/24/02
 Theory-Based Evaluation

individual goals). We will refer to this set of beliefs about how services should be delivered as the

“family support principles” (Kagan & Shelly, 1987).

In developing evaluations for these programs, we have tried to address several problems that

we identified in past evaluations of family support programs (Green, Mulvey, Fisher, & Woratschek,

1996). First, many evaluations used methods and approaches that violated the family support principles

described above. Second, these evaluations often failed to account for program comprehensiveness

and for the dynamic nature of services and families. Finally, many past evaluations were simply not very

useful to programs, researchers, or policy makers for increasing our understanding of how these

programs function (Weiss & Jacobs, 1988).

In response to this, we have taken the stance that evaluations will be more useful if the

evaluation shares the program value system (Green et al, 1996). A family support program

evaluation, therefore, should include the following features:

(1) Collaboration among researchers, program administrators and staff, and family members,

with all parties contributing equally to development of the evaluation;

(2) A family centered and driven approach, in which family members have decision-making

power in regards to the evaluation;

(3) A Strengths-Based orientation, focused on building the program’s strengths and

continuous program improvement;

(4) Comprehensiveness and flexibility, to capture the programs’ multifaceted nature and

ongoing changes in program goals and services.

The use of participatory and theory-based evaluation techniques provides a useful framework

for translating these principles into evaluation methods. For example, the use of participatory methods

Green et al., p. 6
4/24/02
 Theory-Based Evaluation

facilitates collaboration and family-decision making power. The theory-based approach provides a way

to capture the comprehensive nature of the program within an organizing framework, and facilitates

flexibility by outlining the likely developmental sequence of program changes. Using a theory-based

approach combined with participatory methods thus provides a means for organizing collaborative

efforts aimed at building consensus about the program’s theory, and for translating this theory into

evaluation activities. Below we outline the key steps towards implementing a participatory, theory-

based approach to evaluation.

Conducting a Participatory, Theory-Based Evaluation

Step On: Developing the Program Theory.

There are many different ways that we have worked with individual programs to develop

program models. Whatever the process used to develop the theory, however, the end product should

answer the following questions: (1) What are the important program goals? (2) What are the program

services offered? (3) What are the implicit or explicit assumptions about how program services will lead

to program outcomes? This last question is key to developing a program theory, as it addresses the

question of why and how the services provided are expected to lead to change in families.

One successful method we have used to develop the program theory relies on a “grounded

theory” perspective (Miles & Huberman, 1995). In this approach, our goal was to represent the

perspectives and implicit understandings of participants and staff, rather than imposing our own research

theory onto the program. This is critical to the development of a truly useful theory-based evaluation

that is rooted firmly in the way that participants and staff understand and interpret the program goals,

services, and outcomes. This approach also clearly reflects the family support principles that we use to

guide the evaluation.

Green et al., p. 7
4/24/02
 Theory-Based Evaluation

In one evaluation, we conducted separate focus groups comprised of program participants and

staff to address a series of “framing questions”. Our goal in this process was to gain an understanding

of the specific outcomes program families and staff viewed as most important, and what steps they

believed were important in facilitating these changes. We asked parents questions such as “How have

you (or your child) changed since being in the program?” and “What are the most important things that

this program has done for you”? We asked staff questions like “What kinds of changes have you seen

in participating families” and “What are the most important things the program does to help support

families?” We took extensive notes during each meeting, and organized these notes around key issues.

When possible, we then returned to the group with this synthesis for more discussion, verification, and

revision of the synthesis, until eventually we reached some consensus about the program model.

Another approach, which is helpful for programs that lack available staff or families for

participating in this process, involves starting with the program’s written proposal, which will often have

explicitly stated goals and a description of the program services. After initial examination of these

documents, we derive a tentative program theory. We then use feedback from “key informants” (e.g.,

proposal writers, funders, other program administrators) who have knowledge about the program to

verify and modify the theory. In order to remain consistent with the family support principles, however,

it is critical in these situations to involve program staff and families as soon as possible, and to work with

them in a manner which does not imply that we are telling them what the program should be. Rather,

we come to such discussions ready to abandon completely the theory that was previously developed, if

necessary.

Involving staff and families as early as possible after program start-up is not simply a matter of

consistency with our principles, but also helps ensure that the program model is an accurate reflection of

Green et al., p. 8
4/24/02
 Theory-Based Evaluation

how the program unfolds in the “real world”, which is often quite different from the original proposal. It

is quite possible for major changes in program models to occur between the time proposals are written

and the time services are implemented. Program theories developed solely on the basis of initial

proposals cannot adequately reflect these changes. While this method (starting with written program

documents) is frequently used for developing or inferring program theory, we think this should be done

with caution.

Regardless of how the initial model is developed, the program model should not be assumed to

static. Programs often change and adapt to political and social contexts, policy changes, and to the

demands and needs of constituents. For program theories to be useful they must keep up with these

changes. Thus, we prefer to think of any program theory as a “working model,” that is, one which is

constantly scrutinized, evaluated, and modified as needed.

Figure 1 presents a simple model of one program’s understanding of how their services might

lead to a set of positive program outcomes. This model suggests that through implementation of

services using the family support principles, parents and children will experience two important changes:

(1) an enhanced sense of mastery, control, and competence; and (2) improved relationships, including

both formal and informal social support, intra-family relationships, and parent-child attachments. These

two intermediate outcomes are seen as key changes that families experience en route to other kinds of

positive outcomes, such as healthy child development, improved parenting, improved physical health,

and economic self-sufficiency.

It is important to note that this is just one way of depicting the program’s theory of change, and

represents a very simple description of basic program assumptions, intended outcomes, and program

services. We have also developed other, more complex models, using the basic program model as a

Green et al., p. 9
4/24/02
 Theory-Based Evaluation

starting point, in order to more specifically understand specific outcomes and to make more specific

decisions about evaluation procedures. We should also make it clear that although our priority as

researchers is to surface the implicit understandings of program staff and families, we also shared our

knowledge of social science theory and research during this process. The key in this process is

developing a partnership between researchers and other stakeholders so that no one perspective is seen

as “true” or “right” but that all perspectives could contribute equally to development of the program

theory.

Step Two: Using the Program Theory to Design the Evaluation

Although the process of developing a program theory is worthwhile in and of itself, we are

particularly interested in using the program theory to guide the evaluation. Consistent with the

participatory approach, we have relied on extensive collaboration among evaluators, program staff,

directors, and families throughout the evaluation process, typically in the form of an “evaluation oversight

committee” (Green, et al, 1996). This committee works together to make central evaluation decisions

based on the program theory.

Using the model to develop the evaluation questions. One of the most important functions

of a theory-based evaluation is to provide a basis for generating and prioritizing evaluation questions

(Weiss, 1995). For example, key questions that emerged from the program theory depicted in Figure 1

included: “Do the family support services lead to increases in parents’ feeling of mastery?” and “Does

this mastery then lead to improved parenting? Improved self-sufficiency?”

Further, development of the theory often surfaces important differences of opinions between

stakeholders about how the program functions. Such disagreements can be parlayed into key

evaluation questions. For example, one question that emerged regarding Figure 1 was “are mastery and

Green et al., p. 10
4/24/02
 Theory-Based Evaluation

relationships really stepping stones to other outcomes, or can families reach desired outcomes without

experiencing these changes”? For purposes of the working model, the group decided to assume that

mastery and relationships were, indeed, stepping stones; however, we as evaluators were able to

facilitate this decision by suggesting that this could be a central evaluation question.

Using the model to develop evaluation methods. Once questions about the program are

generated, the model can facilitate discussion of how to gather information to address these questions.

In one program, discussion with staff and families led to the decision to focus on three sources of

information, which we have found to be extremely valuable and have used in other family support

evaluations. First, to assess changes in participants on key intermediate and longer-term outcomes,

evaluation staff conducted face-to-face interviews with families at baseline (program entry) and annually.

Second, staff used contact summary records to document the nature and content of services received

by families on an ongoing basis. Finally, families and staff develop individualized goal plans and used

these to plan services and document families’ progress towards their goals.

Methods for assessing individual goals were designed to address a common issue that emerged

during conversations with families and staff, which was that in family support programs, not all families

have the same goals. Therefore, focusing solely on over-arching program goals would overlook an

essential component of the service program. For example, not all families might identify adult education

as a goal, so looking at overall program effectiveness in increasing the level of adult education could

mask program effectiveness. However, a more sensitive set of analyses could be conducted on the

subset of families who identified a goal in this area. The documentation of individual goals is, therefore,

an essential component of family support evaluations, as it allows us to address the question of “what

works for whom” based on the individualized goals set by families. This also made us realize that an

Green et al., p. 11
4/24/02
 Theory-Based Evaluation

important part of a theory-based evaluation of family support programs is the ability to develop

individualized, family-specific “theories of change”.

Using the model to develop evaluation measures. After determining the basic evaluation

methods, the model can be used to help guide decisions about the specific measures. In one program,

we spent a considerable amount of time developing a tool that would record the nature and content of

services delivered (the “contact summary”). To do this, we started with a particular desired program

outcome, for example, parenting competency. Then, using the program model that outlined how

services were expected to lead to this outcome, an extensive list of all services relating to parenting

competency was generated. This list was then used to ensure that the contact summary would reflect all

relevant services. The process was then repeated for other program outcomes. In this way, we

ensured that all services that were believed to relate to key program outcomes could be easily coded in

the contact summaries. Such decisions, about how to assess service delivery, are critical to a theory-

based model that aims to understand the intervention process.

In another program, there was a strong interest in looking not just at what kinds of services

were delivered, but instead at whether services were, in fact, consistent with the family support

principles. This challenged us as evaluators to develop ways to assess the program’s approach to

service. To do this, we worked closely with program staff and administrators to develop the Family

Support Practices Instrument (FSPI, Green & McAllister, 1998), a measure of families’ perceptions of

staff’s behavior and attitudes.

Another example of how the theory-based participatory approach, has facilitated decisions

about evaluation measures concerns how decisions were made about instruments to include in the

interview protocols. We (the evaluators) had suggested that the program include the Parenting Stress

Green et al., p. 12
4/24/02
 Theory-Based Evaluation

Index (PSI, Abidin, 1986) to assess reductions in parenting stress. The evaluation oversight

committee, however, reacted strongly to the PSI’s negative wording. As parents said, “its everything

about how bad and hard it is to be a parent, and nothing about what makes a good parent or a happy

relationship—this program focuses on what’s good so, let’s measure that”. The program theory, in this

situation, had included improved parenting as a key outcome, and although reducing parenting stress

was an important way to enhance parenting, these parents urged us to follow the program values and

focus on measuring family strengths.

Step Three: Continuous Program Improvement

The program model has also served as an important framework for sharing information with

these programs and for guiding continuous program improvement. Often, too, changes and

improvements in the program have led to changes in the evaluation. This iterative process helps to

ensure that the evaluation remains integrally tied to the program, and helps to maximize the usefulness of

evaluation information.

The program model can guide program improvement because it serves as a reference point for

understanding information generated by the evaluation. For example, in one program, early descriptive

data detailing the kinds of services that were being delivered indicated that while many recreational and

social activities were being provided, there were fewer services related directly to parenting.

Comparison of this information with the program model led to discussions of the importance of

social/recreational activities, as well as the realization that more parenting services were needed if

desired parenting outcomes were to be achieved. Thus, the program began to integrate recreational

activities with parenting information (for example, hosting card parties which included a guest speaker,

and developing games that could be both recreational and educational).

Green et al., p. 13
4/24/02
 Theory-Based Evaluation

In another program’s model maternal depression was seen as something that was a significant

barrier to a broad range of positive outcomes. This led to the inclusion of measures of depression in the

parent interview. Initial baseline data suggested that the program had been correct in assuming that

depression was a significant problem, but had underestimated the extent to which this was true;

baseline data indicated that upwards of 60% of the mothers interviewed had near-clinical levels of

depression. The program responded by hiring an on-site mental health specialist.

As we stated above, sometimes changes in the program can result in changes in the evaluation.

For example, in one program, although the original program model didn’t mention support groups for

parents as a way to enhance parenting, support groups were added about a year after program start-up,

and were quite successful in attracting large numbers of participants. Because of our ongoing

discussions with program staff about the program model, we were able to capture this change in the

evaluation by incorporating this service into our measures.

All of these programmatic activities were a natural part of the evolution of the program which

were not part of the original program model: this underscores the importance of continued dialog

around the program model and accompanying evaluation activities to ensure a “match” between the

program and its evaluation. Moreover, we have not only modified assessments, but also added new

evaluation questions as programs have evolved. Changes in the program model have changed the focus

of the evaluation questions as appropriate, and the evaluation has attempted to be responsive to the

needs of the program in by producing timely, useful information.

Step Four: Disseminating Evaluation Findings

Finally, it should be mentioned that the program theory can serve as an important conceptual

framework for disseminating evaluation results. The model can be used to focus on specific questions,

Green et al., p. 14
4/24/02
 Theory-Based Evaluation

as we mentioned above, and then brief, timely reports can be generated based on a particular aspect of

the model. Such reports are often much more useful than long, complicated summary documents

produced at the end of the evaluation period.

Benefits of Theory-Based, Participatory Evaluation

In sum, we believe that using a theory-based, participatory evaluation approach results in a

number of immediate benefits to service providers and evaluators, including the following:

(1) Providing a format for evaluators, practitioners, and family members to share their

assumptions about how and why a program is supposed to work;

(2) Development of evaluations that more accurately reflect actual program processes

and intended outcomes;

(3) Providing a systematic, commonly understood method for prioritizing evaluation

activities and allocating limited evaluation resources;

(4) Providing a framework for sharing information with programs for continuous

program improvement and for dissemination of evaluation findings.

In addition to these benefits, we believe that theory-based, participatory evaluations further our

understanding of program functioning at a broader level. These kinds of evaluations foster the exchange

of ideas, information, and assumptions among researchers, practitioners, and parents, which can lead to

a richer and more complex understanding of how and why these programs work. Further, by definition,

theory-based evaluations focus directly on understanding the mechanisms underlying program

functioning and thus can address complex conceptual questions about the nature and effectiveness of

interventions. Finally, Weiss (1995) suggests that, because theory-based evaluations focus on

explanation of program effects (rather than just documentation) an increased use of this method may

Green et al., p. 15
4/24/02
 Theory-Based Evaluation

lead to an improved ability to integrate evaluation results into a larger body of theoretical and program

knowledge (Weiss, 1995). Thus, using a theory-based, participatory approach has both immediate

benefits to the program, as well as enhancing usefulness of evaluation results on a broader level.

Challenges in Theory-Based, Participatory Evaluation

Despite these benefits, evaluations using theory-based and participatory approaches are still less

common than more traditional outcomes-focused evaluations. We have identified several factors that

may contribute to this fact: Time and cost, control and power, lack of appropriate methods, and

questions of evaluator objectivity.

The challenge of time and cost. It is probably clear that this method of evaluation involves a

large time commitment from evaluators, program staff, and families. Because of this time commitment—

especially the initial time spent developing the model - both programs and evaluators have to be

committed to the benefits of this approach. A common compromise in many participatory evaluations is

to involve stakeholders in the initial model development and evaluation design, and to decrease their

involvement over the course of the evaluation. We believe, however, that one of the most important and

valuable aspects of the evaluations we have conducted has been the continued, long-term involvement

of stakeholders in generating changes in the program model, refining data collection systems, and

interpreting and using the information collected. Thus, to maximize the benefits of this approach, a

commitment to work together throughout the course of the evaluation is needed.

Related to time are issues of cost. Bickman (1989) suggests that theory-driven evaluations are

almost always more expensive than less comprehensive approaches, and indeed, a participatory,

theory-based evaluation, can be quite expensive in terms of purchasing services from an evaluator.

Such costs can be minimized, however, to the extent that theory-based evaluations result in better

Green et al., p. 16
4/24/02
 Theory-Based Evaluation

prioritization of a small number of focused research questions, rather than in full-scale comprehensive

evaluations that attempt to test a program model in its entirety. Thus, although theory-based evaluations

can result in extremely comprehensive, complex, and expensive evaluations, we do not think this is

necessarily the outcome of adopting this approach.

The challenge of control and power. At its best, a participatory evaluation approach can

help to mitigate the power differentials that exist between researchers, program staff and administrators,

and program participants, and facilitate true sharing of decision-making power among stakeholders.

This power sharing, however, this doesn’t happen automatically - bringing stakeholders to the table is

necessary but not sufficient for building collaboration. Evaluators are challenged in this approach to

resist being the “experts” – to share expertise when asked or as appropriate, but to let final decisions be

made by the group. This is often a bitter pill for evaluators to swallow, but we believe that the benefits

of obtaining significant and meaningful input from program staff and families about the intervention

process far outweigh any costs resulting from lack of evaluator control over methodological decisions.

The challenge of methodological inadequacies. Programs are not stable, and neither are

their theories of change. Therefore, the evaluations of these programs have to be able to adapt to

evolving program processes. Most researchers, however, especially those trained in quantitative

research approaches, are not used to measuring “moving targets.” In using a participatory, theory-

based approach we have continually struggled with this issue. Proposed changes are frequent, and each

requires a consideration of the costs and benefits of that particular change, in light of the goals of the

evaluation. For example, adding questions is minimally intrusive to a research protocol; changing the

way questions are asked, for example, is more problematic, although at times we have done so, often

sacrificing standardization in order to make sure that the data were meaningful.

Green et al., p. 17
4/24/02
 Theory-Based Evaluation

Finally, we should note that in order to fully understand programs’ intervention processes and

how they relate to program outcomes, the use of both quantitative and qualitative data is required.

Qualitative methods, in particular, are much better able to adapt to constantly changing program

contexts. In the evaluations we have conducted, we have always relied on a combination of these

methods, and we believe our evaluations have benefited from systematic integration of both kinds of

data.

The challenge of the objectivity of the evaluator. A frequent concern about using

participatory evaluation methods relates to the perception of the evaluator as being an advocate for the

program, rather than an “objective” evaluator. However, we would suggest that the problem lies in our

understanding of the term “evaluation”, which is too often defined as a judgment about whether a

program works or doesn’t work. The theory-based evaluation approach does not work well for

programs that want an evaluation to “prove” to their funders that they are doing a good job. Rather,

because it is explicitly oriented towards explanation and program improvement, theory based evaluation

relies on an objective, open, reflective, and self-critical approach to the evaluation process. We would

suggest that the close working partnership we have discussed can lead to greater objectivity and

openness in addressing program strengths and weakness than does a traditional, hierarchical evaluation

approach.

Summary

We have presented just a few examples of how a process of working collaboratively with

program staff and participants to develop a program’s working model can be a powerful mechanism

for furthering our understanding of family support programs. Clearly, it is not an approach that can be

used indiscriminately for any evaluation—it requires a commitment to depth of understanding, sometimes

Green et al., p. 18
4/24/02
 Theory-Based Evaluation

at the expense of breadth, and a willingness to sacrifice tight methodological control and to be flexible in

data collection. It is, however, an extremely useful approach when the goal of the evaluation is to gain

a richer, more complete understanding of how the intervention process works, and to collect data that

is useful and meaningful to researchers, program practitioners, and families.

References

Abidin, R. R. (1986). Parenting stress index, manual (2nd Ed). Charlottesville, Va: Pediatric

Psychology Press.

Bickman, L. (1987). The importance of program theory. In L. Bickman (Ed.), Using program

theory in evaluation: New directions for program evaluation, pp. 5-18.

Bickman, L. (1989). Barriers to the use of program theory, Evaluation Practice, vol. 12, pp. 387-

390.

Chen, H. & Rossi, P. (1983). Evaluating with sense: The theory-driven approach, Evaluation

Review, 7(3), pp. 283-302.

Green, B., Mulvey, L., Fisher, H., & Woratschek, F. (1996). Integrating program and evaluation

values: A family support approach to program evaluation, Evaluation Practice, 17(3), pp. 261-

272.

Green, B., & McAllister, C. (1998). The Family Support Practices Instrument. Unpublished

manuscript.

Greene, J. (1987). Stakeholder participation in evaluation design: Is it worth the effort?

Evaluation and Program Planning, 10, pp. 379-394.

Kagan, S.L., & Shelley, A. (1987). The promise and problems of Family Support Programs. In

Green et al., p. 19
4/24/02
 Theory-Based Evaluation

S. L. Kagan, D. R. Powell, B. Weissbourd, & E. Zigler (Eds.), American’s Family Support

Programs. New Haven, CT: Yale University Press.

Miles, M.B., & Huberman, A. M. (1995). Analyzing qualitative data, 2nd Ed. Thousand Oaks,

CA: Sage.

Weiss, C. (1972). Evaluation research: Methods of assessing program effectiveness. Englewood

Cliffs, NJ: John Hopkins Press.

Weiss, C. (1995). Nothing as practical as good theory: Exploring theory-based evaluation for

comprehensive community based initiatives for children and families. In Connell, J. P., A. C.

Kubisch, L. B. Schorr & C. H. Weiss, (Eds.), New approaches to evaluating community

initiatives: Concepts, methods, and contexts. Washington, DC: The Aspen Institute.

Weiss, H. & Jacobs, F. (1988). Evaluating Family Programs. NY: Aldine De Gruyter.

Green et al., p. 20
4/24/02