4

INTERVENTION AND EVALUATION

ADAM LoDZINSKI

MICHIKO S. MoTOMURA

FRANK W. SCHNEIDER

CHAPTER OUTLINE
Design of Interventions Developing a Solution:
Nature of Interventions Forming the Intervention
Key Tasks in Intervention Hypothesis
Design and Delivery Goal Seting and Designing
Evaluation of Interventions the Intervention
Reasons for Evaluating Implementing the Intervention
Interventions Evaluating the Intervention
Ineffective Interventions Further Applications of Social
Types of Evaluation Norm Theory
Importance of Research Other Interventions
Design in Evaluating Examples of Other
Interventions Interventions
Evidence-Based Complex Problems Require
Interventions Complex Solutions
An Intervention Example: Influencing Social Policy
Reducing Alcohol Intervention Issues
Process Issues
Problems on Campus
Ethical Issues
ldentifying
the Problem Summary

61
 62
PART I FOUNDATIONS OF APPLIED SOCIAL PSYCHOLOGY

ontributing to the development of inter behavior of people for the purpose of improving
C ments in thelves of people is thedefining
vention strategies that lead to improve their functioning with respect to some social or
pYactical problem, Some interventions might not
feature of the field of applied social psychology target people's behavior directly, for example,
and, in our view, is the most exciting thing about those that are designed to increase knowledge or
being an applied social psychologist. Imagine the awareness (e.g., of the environmental benefits of
great sense of accomplishment and gratification recyclingof are designed to change attitudes (e.g»
that Sherif (1966b) and his colleagues must have becoming more supportive of recycling).Howeve,
experienced when they, in fact, did reduce conflict the ultimate go¡l of most interventions is behavior
between rival groups of boys by using a strategy change (e.g, increased recycling). As pointed out
that involved having the groups work together in Chapter 1, interventions can be conducted at
toward common goals. Imagine how rewarding different levels of analysis (e.g., individual, group,
it is for those who are involved in designing organization, community). Although interven
community-based interventions that influence tions can be directed at different levels, it can be
people to adopt healthier litestyles (e.g, engaging argued that ultimately they are directed toward
in regular exercise, employing safe-sex practices), individuals in that for changes to occur, such as in
thereby reducing the incidence of serious health a group or in an organization, the individual
problems (e.g., heart disease, sexually transmitted members must change in some way.
diseases). Throughout the remainder of this book, Moreover, to understand the focus of this
you will be provided with many examples of suc chapter, it is helpful to distinguish between two
cessful interventions that are grounded in social broad types of intervention: personal and pro
psychological theory and research. Likewise in grammatic. Personal interventions are those that
this chapter, several social psychologically basedpeople carry out in the course of their daily lives,
interventions are reviewed. In particular, the that is, when they use their khowledge of social
chapter focuses on an intervention strategy that psychology to improve their own circumstances or
has been implemented on many college
campuses those of people around them. Examples of per
to deal with the high rates of alcohol abuse. sonal
Given that interventions are so central to the where interventions
"personal
were provided in Chapter 1
uses
field of applied social psychology, it is important of social psychology were
considered
for you, as a student of the field, to have a good tips). (e.g., waitpersons improving their
Clearly, all of
understanding of their nature. The major goal of social psychology inusourcan benefit from applying
this chapter is to consider how applied social psy lives. However, personal
interventions
chologists draw on their ünderstanding of theory, which deals with are not the focus of this chapter,
methods, and research evidence in the design and However, they are programmatic interventions.
evaluation of interventions The chapter takes includes considered in Part 3, which
chapters that
you through the steps involved in developing and sonal interventions. focus specifically on per
evaluating interventions, using examples of actual In
interventions to illustratè the steps. It also recog with the current chapter, our primary concern is
nizes social psychology's role in influencing social programmaticinterventions, which com
policy, which ís another very important way of monly are referred to simply as programs. In this
applying social psychology and is closely related chapter (and in the book as a whole), the terms
to the application of the field through the intervention and program are used
interchange
imple- ably.
mentation of interventions. Near the end, the detinedRoyse, Thyer, Padgett, and Logan (2006)
chapter considers some of the practical and a program as "an
cal issues that are confronted by ethi- of activities designed to reachorganized collection
people who p. 5). In the context of appliedcertain objectives"
practice applied social psychology. the activities that comprise a social psychology,
toward addressing a social program
or
are directed
with the objective of practical problem
DESIGN OF INTERVENT0ONS preventing, reducing,
eliminating its negative consequences. In someor
instances,
forcing andinterventions
may be directed at rein
Nature of Interventions
.A An intervention may be defined as astrategy (e.g., strengthening a positive situation
improving the productivity of an already
eftective work group). Except for the personal
(or procedure) that is intended to influence the interventions that have been
mentioned, all of the
 Chapter 4 Intervention and Evaluation,

interventions discussed so far in this book fall Step 1: Identifving a problem.!Programs are initi
under the category of programmatic interven ated to address social problemsor practical prob
tions. For example, one program involved the lems. The first step in program design is to
procedures and actiities that Hodges, Klaaren, identify the kxistence jand yseverity}of a problem.
and Wheatiey (2000)employed in getting female Aproblem usualy is identifed and defined by
college students to engage in comfortablesafe-sex stakeholders. Stakeholders are individuals or
discussions, and another example entailed the set groups who have a vested interest in the possible
of procedures and activities that Sherif (1966b) development of a program in that they may be
used to get groups of boys to work cooperatively affected by it in some way. Stakeholders include
toward superordinte goals. not only the potential recipients of the program
We alsocan identify some interventions as trial but also individuals such as program funders,
interventions. Trial interventions are those that administrators of the organizations responsible
are implemented to determine whether the inter- for delivering the program, program managers,
ventions, as designed, in fact have the intended and frontine staff members (.e., theemployees
positive consequences. These are also known 8 as who actually carry out the program activities). As
program ejiCacy stuctes Crano &Brewer, 2002). you might imagine, difficulties arise when difer
Trial interventions typically are associated with ent stakeholders disagree about whethera prob
programmatic interventions, although theoreti lem exists, how serious a problem is, or which
cally personal interventions can be tried out as problems should be given highest priority
well. There are two basic kinds of trial interven Needs assessment is the term that is com
tions. One is when researchers designastudy to test monly used to refer to the proçess of establishing
out a possible interventionstrategy. The interven whether or not there isa need orproblem (these
ti¡ns of Sherif (19665) and Hodges and colleagues words are usedinterchangeably) to sufficiently
(2000) represent this type of trial intervention. warrant the development of a program. A needs
The second kind of trial intervention is when an assessment may be informal in nature, for exam
organization conducts a pilot prograrm to deter- ple, wher acoach etermines that a team-building
mine its effectiveness before imptementig it on exercise is necessary based on her own experi
a more permanent basis or before implementing ence with her team or when a manager decides
iron a wider scale: For example, as some police that his department needs a workshop on sexual
departments in North America have attempted harassment after overhearing some of his staff
to transition to community policing, which is a members making sexuàlly inappropriate
new model of policing, they frequently test the remarks. In general, we have more confidence in
efficacy of the model by trying out a community the conclusions of a formal needs assessment
policing unit in one or two neighborhoods. Many that relies on systematic research procedures for
interventions considered in this book are trial inter-, collecting data that are relevant to problem
ventions, particularly [Link] involve a research severity and prevalence. Problems may be inves
test of an intervention design: We now turn to a tigated using a variety of qualitative and/or
consideration of the main tasks in intervention quantitative procedures (see Chapter 3), for
instance, by meansof interviewing representa
design and delivery. of various stakeholder groups or adminis
tives
tering questionnaires to them. Also, a formal
needs assessment gauges the availability of exist_k
Key Tasks in Intervention ing programs or sèrvices as well as possible
bar
Design and Delivery riers to or gaps in service,s
The process of intervention design and imple
mentation follows four overarching steps that Step 2: Arriving at a solution. AscertaYming the
existence of a problem or need is one thing; deter
reflect the general problem-solving approach
adopted by many areas of applied psychology and mining how best to address it is quite another. To
are applicable whether the recipient of the interven arrive at a solution, it is important to identify the
tion is one individual or many individuals, These factors responsible for causing the problem.
Steps are (a) identifying a problem, (b) arriving at a When identifying causal factors, one should dis
solution, (c) setting goals and designing the inter tÉnguish between precipitating factors (i.e., those
vention, and (d) implementing the intervention that triggered the problem) and perpetuating
1(Oskamp &Schultz, 1997). factors (1.e., those that sustain the problem and
 64 PART I FOUNDATIONS OF APPLIED SOCIAL PSYCHOLOGY

immunized by first receiving a small challenge to

keep it frombeing solved).Making the distinction
have an opportunity to prepare a
between precipitating factors and perpetuating their beliefs andweak
factors is critical to the design of an intervention attack, they are better able to
rebuttal to this
becausethe tactors or events that lead to aprob- restst asubsequent more powerful attack{ina cre
McAlister,
lem are not always directly involved in its contin ative leap from research toapplication, whether it
wondered
uation. For instance, factory employees may be Perry, and Maccoby (1980) inoculation effect to an
laid off for one reason (the precipitating factor is possible to apply the
helping seventh
may be a slowdown in the economy) but unable actual social problem, namely,
6reten

smoke.
employment for another reason grade students resist inducements to
secure new
Drawing on McGuire's (1964) inoculation theory
(the perpetuating factor may be a lack of skills and his innovative methodology, McAlister and
that are demanded by alternative jobs). In this involving
anel
case, one must identify the perpetuating cause colleagues (1980) developed aprogram
lack of important skills-as the factor to be tar dren röle-playing that was successful in getting chil
to resist peer pressure to smoke/ commu
geted so as to solve the problem (inability to find
new employment).
For clarity of conceptualization and
causal factors have been identified, the nication, solutiöns to problems should be
Once
through a literature expressed as intervention hypotheses (Lodzinski,
next step is to find out (often
terventions that have effec- Intervention hypotheses are "if-then"
2003)
review) wiether summarize the intervention and
tively addressed the samefeeds already exist. statements that McAlister
interventions can be used to guide the devel the expected 'outcomes. In the case of
Such coHeagues' (1980) smoking prevention pro
opment of a solution to the current problem. If and the intervention hypothesis could be
previous interventions cannot be located, then a gram, seventh-grade students are
solution must be developed independeñtly. When stated as follows: "If
opportunity to rehearse rebuttals
possible, solutions shoutd be based on relevant provided the
social psychological theory and research evidence to inducements to smoke in realistic role-playing,
scenarios, then in their daily lives
as wellas theory and evidence from any other peer pressure to resist pressures to
field that may contribute to a solution. Recall that they will be more likely
Chapter 3 ended by noting that Ross, Lepper, and smoke and less likely to begin smoking."/
Ward (2010) identified the "nonobviousness" of Step 3: Setting goals and designing the interven
experimental outcomes as an important theme of
social psychology research. Ross and colleagues tion. Once the need and the proposed solution
also extended this observation to include the non have been determined, it is necessary to develop
obviousness of the outcomes of intervention the program activities, which refer to the speçific
strategies. Some 19 years earlier, Ross and his col components and procedures of the program The
league Nisbett (1991) cautioned against the devel- best place to begin this process is to set the goals
opment of interventions based on conventional and objectives of the intervention. Khowledge of
lay understanding and intuition because "lay goals and objectives serves to guide the selection
predictions are often both wrong and too confi- of program activities. Goals refer to the ultimate
dently made" (p. 18). They added, "The hard won or long-term outcomes that one hopes to accom
lessons of social psychology ... constitute a rep- plish through an intervention. For exarHple, a goal
ertoire of strategies with which to supplement the for a substance abuse program might be to have
guidance of common sense in constructing inter the clients abstain from alcohot and other drugs.
ventions" (p. 245), Once goals have been established, it is important
Consider the following example of an effective uÑ define the program objectives. Objectives refer
smoking prevention intervention that was based to short-term outcomes (e.g., during or immedi
McGuire's (1964) research on the "inocula ately after an intervention) and intermediate
tion effect McGure discovered that just as it is term changes (e.g., one or twO months later) that
possible to be immunized against a disease, such Occur as a result of the intervention and are
as polio, by being inoculated with a vaccine con required for (i.e., support) the attainment of the
taining a weak strain of the virus, so too is it pos program goals. In other words, goals refer to the
hle for people to be immunizea agansT attacks ends, whereas objectives refer to the means or
experi
on their beliefs/In a series of laboratory
participants
steps by which the ends are achieved. For
are if the goal instance,
ments, he demonstrated that if is for clients with
substance abuse
 Chapter 4 Intervention and Evaluation 65

problems to remain abstinent, one objective Fundamental to a program logic model is its
might be for them to understand why they use theoretical basis (Unrau, Gabor, & Grinnell,
drugs in the first place. 2001). That is, a logic model should be based on
Once the goals and objectives have been set, atheoretical rationale that explains the causal
the next step in intervention design is to deter- connections among its various components, for
mine the program activities. When choosing example, why rehearsing rebuttals will induce
activities, one of the most important questions is resistance to peer pressure to smoke. From the
the following:What objective (and ultimate goal) point of view of intervention design, this means
will the proposed activity help to meet? For that one should be able to point to any comp0
example, for clients to learn the reasons for their nent of the intervention and indicate not only
drug use (objective), they might need to have what its contribution is but also why the effect
individual counseling sessions with a certified should occur. The use of a program logic model
counselor (activity). Figure 4.1 illustrates addi ensures a careful theoretically and empirically
tional examples of intervention activities, objec- based articulation of the program and increases
tives, and goals. the likelihood of its succéss. This, of course,
The process of specifying the various com helps to ensure that program resources are used
ponents of a program-goals, objectives, and as effectively as possible. Unfortunately, in
activities_-requires a sound rationale, often applied settings, too often it is the case that pro
referred to as a program logic model. Aprogram grams are designed without formal articulation
logic model is an explanation or a blueprint of of logic [Link] evaluators have begun to
howthe program activities lead to the attainment employ the notion of a theory of change model
of theprogram objectives and, in turn, how the instead of programlogic to underscore the need
objectives logically and operationally contribute to make explicit an intervention's underlying
to the eventual achievement öf the program theory including the steps involved along the
goal(s) (Wholey, 1983). Logic models vary in pathto desired change, the assumptions being
omplexity and detail, but all of them stress a made, and the preconditions that may enable (or
"cause and effect" flow as expressed in theinter- inhibit) the desired change (Mackinnon, Amott,
vention hypothesis, Program logicis the glue that & McGarvey, 2006; Organizational, Researcbd
holds the activitiës, objectives, and goals t9gether SgrYne n e e , d .

Figure 4.1 ExamplesofActivities, Objectives, and Goals for Three Programs

e.
Activities Objectives Goals

Decrease the
Teaching Improve child-parent incidence of child
parenting skills relatlonship abuse

Maintain
Providing nutrition Improve eating
education habits approprlate weight
for a given build

Increase self.
Role-playing a job presentatlon skills
intervieW Obtain employment
in job interviews
66 PART I FOUNDATIONS OF APPLIED SOCIAL PSYCHOLOGY

yStep 4: Implementing the intervention. As the on Psychology as "one particularly hot

term implies, implementation refers to the actual for psychologists"
(DeAngelis,
2008). In fact One growth area
process of enactingthe intervention activities, that
author of this chapter
(Motomura)
ation research as her main vocationaldefines evalu-
is, of delivering themtothe recipients of the inter
vention Apoint worth noting is that there are second author area, and
many practical details that might need to be in (Lodzinski)
as beingregards
design and
evaluation program
at the core
place to impement programs properly. Depending consulting of his
on the complexity of an intervention, determined individuals practice; it is noteworthy that both
received their
by factors such as its size and structure,
details might include securing an practical applied social psychology. graduate degrees in
facility, hiring staff members, ensuringappropriate
adequate
training, and developing things such as operating
budgets,,management
ma structures, job descriptions, Reasons for Evaluating
performance appraisal methods, promotional Interventions
strategies, and cross-agency reterral protocols.
Moreover, an intervention always should be Theeffectiveness of all
designed and implemented in such a way that its tions should be subjected programmatic
to interven
degree of effectiveness can be evaluated. this also applies to personal evaluation. (Actually,
case of individuals who wantinterventions
to
in the
the same mistakes in their avoid repeating
reasons why programmatic lives.) There are several
EVALUATION OF INTERVENTIONS be evaluated (e.g., interventions should
is scientificand stems Lodzinski, 1995). Amajor one
Donald Campbell, aleading authority on research (Cartwright, 1951), from the tradition of Lewin
design in the social sciences, spoke as who viewed all applied work
strongly on behalf of the need for Program eval- often and the being guided by theory and as contributing to
uation. More than development of theory. That is, a major reason
called for a culture 40of years ago, Campbell (1969) for evaluation research is to test the theoretical
evatuation: assumptions ofunderlying the intervention (i.e., the
The United States and other
of
foundation the program logic model or theory
should be ready for an modern nations change). If the intervention is
to social reform, an experimental approach designed, its degree of effectivenessimplemented as
out new programs approachto which we try
designed
in
cure
involved in its design of the validityinforms of the
those
theo
social problems, in specific retical rationale on which the
not these programswhich we learn whether or based. This evaluative intervention is
we retain,
are effective, and in which feedback
respect to attempts to develop theories
is important with
the basis ofimitate, modify, or discard them on
apparent effectiveness. (p. 409) utlity in the world of that- have
application.
Anotherimportant reason for
Campbell'
nition s proposal was driven by the recog ventions is ar ethicat one Royseevaluating and
inter
that no matter how carefully conceived and (2006) listed 14 major social
problems colleagues
(e.g., sub
convincing a program hypothesis is, it is only a stance abuse, poverty,
proposition until the program is actually designed, violence, adolescent pregnancy, unemployment, domestic
mplemented,
ter considers and evaluated. Although this chap- suggested that for each problem illiteracy)
States, in the
and
United
evaluation in a separate section, be
assured that evaluation there are
hundredsoften
is an integral part of
process of intervention design and programs.
the translates into The existence of that manythousands-of
programs
tion and may be implementa- whose lives millions of program recipientS
process In fact, considered
a good
a fifth step in the
various ways
are affected by program activities in
(e-g.,
includes aplan for program intervention design tionally, physically, socially, educa
1995). Testimony to the evaluationof(Lodzinski, Whereas economically) the and to varying
degrees.
interventions importance
is the existence of a separate evaluating improve the
goals of such
programs are to
the social sciences called field in various ways,functioning of their
recipients in
program evaluation research
evaluation). worth noting that pro Those
It is (or the only
appropriate
ensuring that they do so is through means of
gram evaluation recentlywas cited in the Monitor programs have responsible for the evaluation.
an implementation of
 Chapter 4 Intervention and Evaluation 67

determine whether their recipients are receiving designed ones--cannot be assumed. If we can
the intended benefits and whether they are expe- generalize from the literature on program eval
riencing any unintended undesirable conse- uation, the intervention world is replete with
quences The ethical responsibility aiso Extends
ineffective programs. For example, one system
to a readiness to modify or even discontinue
atic review of "Scared Straight" programs that
ineffective programs and to improve on pro involve organized visits to prison facilities by
grams that have been evaluated as effective. Later juvenile delinguents or children at risk of
in this chapter, further ethical issues in the design becoming delinquent found that not only do
and evaluation interventions are considered. the programs not deter criminal behavior, but
The need for financial accountability is another they actually lead to an increase in offending
reason for program evaluaiion. Imagine the bil- behavior (Petrosino, Turpin-Petrosino, &
ions of olars heeed to ind the many thou- Buehler, 2002).
sands of social programs that exist. Most of us Now we will discuss in some detail an early
likely would say that the money isworth it-if the dramatic example of a failed intervention, the
programs in fact do improve people's lives as Cambridge-Somerville project (described in Ross
intended. On the other hand, most of us would & Nisbett, 1991). What is especially noteworthy
be pretty put off if all those tax dollars (our about the intervention, as observed by Ross and
money) were being wasted on ineffective pro Nisbett (1991), is the fact that although the pro
grams. Royse and colleagues (2006) pointed to a gram was the "kind of multifaceted intervention
government-sponsored $2 billion antidrug media that many social scientists would love to see
campaign that was initiated when there were"no implemented today" (p. 214), it turned out to be
well-controlled studies showing that media cam a dismal failure. A major goal of the intervention
paigns are effective in changing behavior" (p. 15). was prevention, that is, to reduce the likelihood of
This astonishing example of fiscal irresponsibility young boys from lower-class backgrounds in a
very clearly highlights the need for rigorous Boston suburb-some of whom were identified
research on the effectiveness of programmatic as delinquency prone-going down a criminal
interventions. For the sake of the funders of pro- path. The program was a model of intervention
grams, there should be accountability in the sense design, with approximately 250 boys randomly
of convincing demonstrations of program effec- assigned to the program and another 250 or so
tiveness(Moreover, even ifa programIs found to randomly assigned to a control group. The boys
be effective in accomplishing its goals, the pro were involved in the program for 5 years and
gram might be too costly. That is, an effective receiveda wide variety of supports (i.e., program
program might cost too much given the available activities). For instance, all boys received twice
resources that might instead be used for other monthly visits to their homes from caseworkers,
purposes, including the implementation of other and substantial percentages received academic
beneficial, yet less expensive, programs. tutoring, psychiatric attention, medical attention,
Finally, program development may be viewed and opportunities for involvement with the Boy
as tie overriding reason for evaluation research. Scouts and summer camps.
Programs are evaluated as part of the process of The long-term effects of the program were
developing the most effective programs possible. evaluated in aseries of studies for (remarkably)
In fact, the three reasons for evaluation men 40 years following the intervention. Two kinds of
tioned previously dovetail into program develop- evaluative data sharply contradicted each othe.
ment by seeking to ensure that programs (a) are On the one hand, the subjective impressions of
based on sound empirically tested theoretical caseworkers and many program participants pre
assumptiOns, (b) are conducted with approprite sented a positive picture of the benefits of the
ethical safeguards, and (c) have a satisfactory . program. On the other hand, the more reliable
ratio of benefits to costs' statistical evidence indicated that the program
participants, as compared with the control group,
Ineffective Interventions had no fewer juvenile and adult offenses and did
not fare better on a number of other indicators
As suggested, programs must besubjected to such as health, mortality, and life satisfaction.
evaluation research because their success Ross and Nisbett (1991)offered several plau
even that of the most carefully rationalized and sible éxplanations for the failure of the
68 PART I FOUNDATIONS OF APPLIED SOCIAL PSYCHOLOGY

Cambridge-Somerville project, including the Athird cause of program failure can be

fact that the situational factors (i.e., program operation of reactance in
program the
activities) that were manipulated as the interven
tion, although impressive in terms of expendi
(Rothman, Haddock, &
Reactance refers to the idea thatSchwarz,
when a
recip200ients
ture of time and human resources, were "trivial" source to
influence threatens people's sense of freedom of
compared with the environmental forces that the think or behave as they see fit, people will act
boys faced on an ongoing basis. They also men against the influence to protect their freedom
tioned the possibility that being identified with (much like Romeo and Juliet did when their
the program mighthave had a stigmatizing effect families opposed their love for each other)
on the boys, such that they and others would (Brehm & Brehm, 1981). Thus, even though an
have viewed them as troubled and delinquency intervention is intended to help people, if the
prone, with such a view becoming aself-fulfilling people feel pressure to change, they might resist
prophecy (see Chapter 9). This explanation raises the social influence attémpt that the program
the unfortunate possibility that the intervention represents.(Thus, esigners of interventions often
actually may have had a harmful effect on the must take steps to minimize the undermining
boys. In fact, on several indicators (e.g., multiple effects of reactance, for instance, by avoiding the
offenses, alcoholism, achievement of professional use of overly strong (i.e.,
reactance-triggering)
status), the program participants were less well persuasive communications and by helping (as
off than the control participants during adult much as possible) to sustain in individuals a
hood. Another possible reason why the program sense of choice or control about being exposed to
may have workedto the detriment of the partici program activities. For instance, one study
pants is that because the boys could be seen as showed that grade school and high school stu
already receiving help from the program, the dents who viewed an antismoking video evalu
usual community sources of help (e.g., clergy, ated the message and its source more favorably if
teachers, social service agencies) might have been the message was implicit as opposed to explicit,
less likely to provide their assistance. The possi that is, if it minimized its persuasive intent and
bility of unanticipated negative consequençes did not impinge on the viewer's freedom of
must always be recognized and assessed in the behavior, for example, by strongly advising
evatuation of a program. ) against smoking (Grandpre, Alvaro, Burgoon,
Our main purpose here is not so much to Miller, & Hall, 2003). The potential undermining
account for the failure of the Cambridge role of reactance speaks to the need for program
Somerville project as to underscore that an designers to understand and anticipate how pro
apparently well-designed and intuitively com gram recipients willdefine the program's objec
pelling intervention can fail. If the program tives, goals, and activities (Ross et al., 2010).
evaluation had not been properly designed A fourth reason for program failure may
(i.e., if it had relied solely on the subjective stem from an incompatibility between program
testimonies of key stakeholders), its failure design and cultural [Link] example of
would not have been detected; in fact, it may major magnitude occurred during the 1990s
very likely have been deemed a success and when the United Nations came to recognize,
regarded as an exemplary model of prevention. after a number of failure experiences, that HIV/
As suggested by the possible explanations for AIDS prevention programsthat are effective in
the failure of the Cambridge-Somerville proj some countries or regions are less effective or
ect, there is a wide variety of factors that can compBetely ineffective in other countries or
contribute to the failure of an intervention to regions. For example, school-based HIV/AIDS
achieve its goal(s). prevention programs are regarded as perhaps
Here are four more possible reasons for pro the most effective way in which to reduce rates
gram failure. One is that the theoretical rationale
of HIVIAIDS (Gallant & Maticka-Tyndale,
the programlogic model-may be inadequate. and 2004). However, their potential in some devel
require revision, A second is that the program oping countries is curbed enormously by the
might not be inplemented as designed (see next fact that the majority of children do not attend
sectioH
onprocess evaluation) despite a sound school and are illiterate and, thus, would be
theoretical and research basis. unable to benefit from information-based
 Chapter 4 Intervention and Evaluation 69

program activities in the first place. Also, the the end of the program. Evaluation of the goal of
goal of condom use, which is a fundamental actual long-term abstinence would occur after a
component of HIV/AIDS prevention programs, specified period of time (e.g., 1year) following
is opposed by very powerful cultural forces in completion of the program.
many countries where having children is an As suggested previously, program evaluation
important symbol of sexual and economic is an integral part of program development. The
potency as well as atremendous source of pride results of a program evaluation often willlead
and status. The paramount importance of one to revisit and revise the program logic
addressing culture in the design of intervention modeland, accordingly, to make changes in the
efforts was formalized in UN policy during the goals, objectives, and [Link] instance, the
late 1990s (e.g., Kondowe & Mulera, 1999), logic model for an alcohol addiction program
might have to be amended to include the impor
Types of Evaluation tant role of family support, thereby necessitating
the inclusion of new program activities (e.g.
In assessing the effectiveness of an interven family counseling) and objectives that empha
tion, there are two main types of program evalu size family involvement.
ation: process and outcome (Posavac & Carey, Evaluators also have recognized the impor
2007). Process evaluation (also known as for tant role evaluation can play in supporting the
mative evaluation) is undertaken to determine process of innovation and learning within an
whether the program has reached its target-audi organization. This type of evaluation has been
ence (as identified in the intervention hypothesis) coined developmental evaluation (Patton,
and whether the program activities (as outlined 1994) and can be used when interventions are
in theprogram's logic model) have been imple- in a stage of early innovation or in situations of
mented în the préscribed manner. Basically, one high complexity, like poverty or homelessness,
wants to änswer the following question: Is the where the causes and solutions to the problem
program being implemented in the way in are unclear and intervention stakeholders are
which it was planned? For instance, if an alcohol not on the same page. Adevelopmental evalua
addiction program's activities include giving tion typically involves testing or experimenting
addicted people five individual counseling ses with new approaches to a problem-perhaps
sions and five group counseling sessions, apro- involving multiple trial interventions--with
cess evaluation would ensure that the clients the intention of developing an innovative solu
were in fact addicted to alcohol and did indeed tion that can be subjected to process and out
receive the prescribed number and types of come evaluation approaches (Gamble, 2008;
counseling sessions. Patton,2011).
Outcome evaluations typically are conducted
after process evaluations. An outcome evalua
tion (also known as summative evaluation) Importance of Research Design
assesses how well a program meets itsobJeCtives in Eyaluating Interventions
(i.e., short-term outcomes as described in the
program logic model), and in a more compre Chapter 3 stressed the advantages of experi
mental designs in the study of causal relation
hensive evaluation, it also assesses how well the
program is achieving its goals (i.e., long-term ships. Heightened confidence in internal validity
outcomes, also part of the logic model). (i.e, the independent variable did in fact cause
Essentially, the overriding purpose of an out changes in the dependent variable) comes with
come evaluation is to determine whether the sound experimentation. Similarly, as Campbell
hypothesized improvement in functioning (1969) and others (e.g., Crano &Brewer, 2002)
occurs amnong the recipients of the program as a' have argued, if we want to know whether an
result of exposure to its activities. For example, intervention does indeed result in its intended
an outcome evaluation of an alcohol treatment consequences, evaluators are in a much stronger
program might assess whether participants pub- position to reach aconfident conclusion with an
icly (e.g, among all other participants) express intervention that is conducted as an experiment.
strong commitment to long-term abstinence by In such a case, potential program participans
70 PART I FOUNDATIONS OF APPLIED SOCIAL PSYCHOLOGY

are randomly assigned to either the intervention Education to provide evidence regarding what
group or a "no intervention" control group. works in education; the Cochrane Collaboration
Assuming that all other environmental factors that provides systematic reviews of research in
are similar for the two groups (this assumption human health care and health policy; and the
becomes less tenable with interventions in the Campbell Collaboration that produces system
field than with those in the laboratory), dffer- atic reviews of interventions in education, crime.
ences between the groups (e.g., gains in students' social justice, and social welfare.
self-esteenm) can be attributed to the interven Some organizations have also introduced
tion with some confidence. evidence-based decision-making tools that
Confidence about program effects will be guide practitioners to critically assess research
lower in the case of interventions conducted as and apply research findings to their
program
quasi-experiments, for example, when random decisions. In this context, Biglan and colleagues
assignment is not used and the outcomes for recommended standards that scientific organi
the intervention group are compared with zations should apply before promoting pro
those for agroup of individuals whom the grams and policies directed toward the
researchers deemed to be similar to the inter prevention of behavior problems in youth
vention group (e.g., the groups may be matched (Biglan, Mrazek, Carnine, &Flay, 2003). Such
on a few characteristics such as income level
and IQ). This is not to say that evaluations of
organizations include, for example, the Center
for the Study and Prevention of Violence and
interventions conducted as quasi-experiments the Center for Substance Abuse Prevention,
are not of value. In fact, when it comes to inter- Biglan and colleagues proposed that before an
ventions in the "real world experimentation organization recommends adoption of a given
frequently is not feasible for avariety of practi- prevention program or policy, its positive out
cal reasons (e.g., an agency cannot afford to run comes for the target
a program as an experiment, random assign- strated by at least two group should be demon
ment cannot occur because the evaluator is trials or three well-designed experimental
faced with an intervention group that is prese series well-designed interrupted time
lected). When experimental procedures are not experiments (one of the more rigorous
possible, quasi-experimental procedures should Chapter 3). According to Biglanas anddescribed in
quasi-experimental designs,
this "standard would mean colleagues,
be employed. However, with such procedures, the use of
conclusions about program effectiveness must tific that scien
be more guarded (Campbell, 1969). organizations would put their resources into
disseminating the programs and policies
that ... would concentrate the limited
resources
of scientific, government, and nonprofit organi
Evidence-Based Interventions zations on the policies and programs that are
In step with increasing demands on most likely to have an impact" (p. 436).
and policies for accountability, whetherprograms As
to a pro Controlanother example, the Centers for Disease
gram funder, the government, or the and Prevention has a panel of scientists
public, is the continually increasing emphasis general assign HIVIAIDS programs to a
the employment of evidence-based intervenon tem that reflects the programs' levels four-tiered sys
tions or "best practices." For instance, in the and scientific rigor. A program's tierof efficacy
of education, the No Child Letft field adjusted up or down in can be
2001 (2002) required that practicesBehind Act of availability of new accordance with the
and policies drawing on such
adopted by schools and school districts evidence (Price, 2008). Now,
be "sci- systems
ing existing programs, for rigorously evaluat
entifically based," a phrase that is
over 100 times in the legislation. Thismentioned fields have a much decision makers in many
trend has greater opportunity
spawned many books and clearinghouses on programs that have scientifically established to select
evidence-based
and social practices across many disciplines
records of effectiveness.
problems. Some notable examples of vention
The following section further clarifies inter
clearinghouses include the What Works design and evaluation by taking you
learinghouse set up by the U.S. Department of ougn each step of the design and evaluation of
an actual program.
 71
Chapter 4 Intervention and Evaluation

AN INTERVENTION EXAMPLE: To deal with problem drinking on campus,

REDUCINGALCOHOL PROBLEMS many colleges, including NIU, have tried "tradi
tional" methods of intervention such as alcohol
ON CAMPUS education and awareness campaigns, which are
based on the idea that increased knowledge of
As noted previously, when designing an interven the negative consequences of alcohol consump
tion, is important to draw on relevant theory tion will reduce drinking levels. However, there is
and research to develop the most effective inter little research evidence that education-oriented
vention possible. This section describes an inter programs reduce alcohol consumption among
vention that involves the direct application of college students (Moskowitz, 1989). Consistent
social psychological theory and research evi with the research, NIU found that these tradi
dence. The goals of the intervention, which was tional methods did not reduce drinking rates or
developed and implemented at Northern Ilinois alcohol-related injury rates among its students. It
University (NIU), were to reduce high-risk was clear that there was a need for a new kind of
drinking anmong students and to reduce the inci intervention to deal with problem drinking
dence of injuries due to alcohol consumption. among NIU students.
The intervention was conducted by the Health
Enhancement Services Office of the University
Health Service at NIU and was initially funded by Developing a Solution: Forming
agrant from the U.S. Department of Education's the Intervention Hypothesis
Fund for the Improvement of Postsecondary
Education (Haines, 1996). To reduce alcohol consumption among NIU
students, an intervention that represented an
ldentifyingthe Prohlwo application of social norm theory was designed
(Berkowitz, 2004; Perkins, 2003). Fundamental to
Alcohol consumption is a very big problem on social norm theory is the central role that norms
many college campuses. The level of alcohol con play in people's lives. Norms refer to shared
sumption among college students has long been beliefs about which behaviors are acceptable and
a concern of school administrators, parents, and which behaviors are not acceptable for members
other community members. The Harvard School of a given group to engage in. Essentially, norms
of Public Health surveyed students at 119 col are prescriptions for how people should act in
leges and universities in the United States particular situations. There is ample research evi
(Wechsler, Davenport, Dowdall, Moeykens, & dence that people do tend to guide their behavior
Castillo, 1994; Wechsler et al., 2002). The results in accordance with what they perceive to be the
showed that, overall, 44% of students had situation-relevant norms (Cialdini & Goldstein,
engaged in "binge drinking" (defined as having 5 2004). Social psychologists have long recognized
or more drinks in a row) during the 2 weeks the influence of social norms on people. However,
prior to the survey. The survey also showed that what is distinctive about social norm theory is its
ahigher percentage of binge drinkers had experi- emphasis on the idea that people often perceive
enced alcohol-related problems since the begin norms incorrectly and use such erroneous per
ning of the school year than had non-binge ceptions to guide their actions. In fact, social
drinkers. Frequent binge drinkers were 7 to 16 norm theory is based on the following three prin
times more likely than non-binge drinkers to ciples. First, individuals tend to conform to what
have reported the following: missed class, fallen they perceive to be the norm for a particular
behind in schoolwork, engaged in unplanned behavior. Second, in some instances, individuals
sexual activity, had sex without protection, found may behaviorally conform to misperceived
themselves in trouble with campus police, dam norms. Third, if misperceptions of norms are cor
aged property, and been physically hurt or rected, individuals willchange their behavior to
injured. A survey of NIU students (details agree with the corrected perceptions.
reported later) confirmed that NIU was not an Research on social cognition has identified
exception with respect to having aserious campus Some thinking errors that can lead people to
wide alcohol problem. misperceive norms. Two are particularly relevant
72 PART I FOUNDATIONS OF APPLIED SOCIAL PSYCHOLOGY

tothe problen of drinking on college campuses. decline. Social I

norm theory
use theand related
One cognitive error is false consensus, which is therefore. led
NIU to
thetendency for peopleto believe that others are tion
like them when in fact they are not (Ross, Green,
& House, 1977). For instance, heavy drinkers
norm
hypothesis for the design of its
NIU students are
for
led to fol owingproingrretesaermavr:echn,.
perceive the f
may believe that most of their peers (e.g., stu
dents at their school) are also heavy
drinking
the levels are lower than levels more accuratelyCampusi
drinkers els among students shouldbeldecrease ieved),, ler.
when in fact they are not. We can see how
kind of thinking would allow heavy drinkers to
this drinking
rationalize their own levels of alcohol consump-
tion as Goal Setting and
thus, helpnormativesocially
to sustain their excessive acceptable and, the Intervention Designing
Another way in which norm drinking.
occur is through pluralistic misperception can The main goals of the
when the majority of ignorance, which is reduce high-risk drinking intervention were to
individuals
assume that others behavior or incorrectly injuries among students at NIU. The and alcohol-related
more different from theirs than itthinking is the intervention was to reduce the objective of
(Toch &Klofas, 1984). For
instance,
actually is of the amount of
drinking on campus. misperception
may believe that others drink students reasoned that reaching this objective would lead
It was
do when in fact they do more than they to the
studies on a variety of college not. In fact, survey
vention.
achievement of the main goals of the
The campaign plan was inter
shown that these two types of campuses have students who drank alcohol. to target all
especially misperception, students Because nearly al
pluralistic
mon. That is, most students ignorance, are very (90%)
com wide intervention was
at NIU drank
alcohol, campus
a
students their campuses drink more
on think that other to guide the planned. Four rules used
actually the case (Graham, Marks & than is sage to the development student body
of the campaign mes
1991; Perkins, Haines, &Rice, Hansen, simple, (b) tell the truth, (c) be to (a) keep it
were
similar results were found. A19882005). NIU, (d) highlight the norm of consistent, and
At
cated that less than half
(43%)
survey indi- Reflecting these rules,
of the
moderate drinking
reported that they drink more than 5students NIU students (55%) drink 5 or was that most message
when they "party" whereas 70% of drinks when they party. fewer drinks
reported that they believed most NIU students Because of the limited resources
drink more than 5drinks when
they
students (one full-time staff person) for this project available
at NIU, the norm party. Thus, the
was that most
drink more than 5 drinks students did not was large number of students at NIU (23,000),andit
solid majority when partying, yet a Woulddecided that the main program
activ1ty
ing the operation misperceived
of
the norm (suggest- that be to use a mass media
most students at NIU campaign. Given
As the second pluralistic ignorance). pus newspaper was their reported that the cam
not onlyprinciple
of social norm primary
suggests, do theory mation about campus activities, it was source of infor
ceive the norm for college students misper- that a print decided
drinking
higher than it really is (e.g., behavior to be largest numbermedia campaign would reach the
it"), but they also "everyone is doing of students at the
behave ways that conform to
in print media campaign included lowest cost. The
this
havemisperception.
determined
As evidence of this,
that if students
per
studies pressadvertisements, a campus campus newspa
newspaper column,
their peers are believe that It wasreleases, ilyers, and posters
selves are drinking,drinking more than they them Sourcealsoof important that students(seeperceived Figure 4.2).
the
(Graham et al., 1991;drinking rates tend to rise the message as
credible. A survey
The Prentice &Miller,
that third principle of social norm 1993).
determinedas that
fessionals more
NIU students rated
health pro
people will conform to theory friends. As a believable than educators and
tions of corrected percep result, a print media
that if normsleads directly the
to was endorsed
of the drinkinghypothesis was deemed to behealth
by
professionals campaign that
misperceptions
corrected, levels of high-risk norm are communicating thethe most effective
on campus
method of
drinking will NIUstudents. intervention message to
 Chapter 4 Intervention and Evaluation 73

Figure 4.2 One Poster Used for Social Norm Media Campaign at Northern Illinois University

Most students drink moderately

2 2 ]Henlth
nancenent

Most men drink 0-5 Most women drink 0-3

drinks when they "party."1 drinks wher1 they "party."
Here are some tips to help drinkers stay safe and have fun when they "party:"
Eat while or before drinking, thereby reducing " Weight makes a difference! If two people drink
the rate of your intoxication the same amount of alcohol, the heavier person
will be less intoxicaled.
" Keep track of how much you drink by " It takes about 1hour per drink to "clear" alcohol
counting stirs, napkins, money, cups, ctc. from the body.
Based on survey data collected by University Health Service (1997) from a representat ive sample of NIU students {N=882)

SOURCE: Courtesy of Northern Ilinois University.

An additional program activity involved a year. This social norm-based intervention has
means of increasing the likelihood that students
been ongoing at NIUsince its inception during
would read and remember the campaign mes the early 1990s.
sage. This entailed rewarding students who
remembered the message and spread the mes
sage to others. For example, groups of students Evaluating the Intervention
were approached at random and asked, "Who An outcome evaluation of the social norm
knows how many drinks most NIU students
drink when they party?" The student with the media campaign at NIUwas conducted to deter
mine whether the intervention was able to reach
correct answer received $1. Students also received
$5 for puting campaign posters on their dorm both its main objective and its two key goals.
room walls. The evaluation sought to answer three ques
tions. First, did the perceived rate of high-risk
drinking (defined as having more than 5 drinks
Implementing the Intervention when partying) among peers decrease to a more
accurate perception? Second, did the rate of
The media campaign was first implemented at actual high-risk drinking decrease? Third, did
the beginning of the fall semester in 1989. The the rate of alcohol-related injuries decrease? To
long-term intervention strategy was to initiate answer these three questions, baseline informa
the media campaign at the beginning of each fall tion was collected from the students in 1988,
semester and to keep it highly visible until spring that is, before the intervention was implemented.
break. The messages would taper off after spring Baseline information refers to data that are col
break and begin again at the start of the follow lected on the target population prior to an inter
ing academic year. Because new students arrive vention (ie., the pretest) and that are compared
every year bringing their own misperceptions with data collected after the intervention has
about drinking levels, it was important to con been implemented (i.e., the posttest). Astudent
duct the intervention at the start of each school survey was used in 1988 to collect three pieces of
74 PART I FOUNDATIONS OF APPLIED SOCIAL PSYCHOLOGY

information about NIUstudents: the perceived From 1989 to 1990, afterfound

the first year of im-l
rate of high-risk drinking among other students, mentation, researchers that
the actual rate of high-risk drinking, and the rate 12% reduction in perceived high-risk there was a
of alcohol-related injuries. This pretest informa
tion was used both to justify the need for the
and an 8% reduction in actual high-risk
but no significant reduction in
drdriinnkikinngg
intervention and tocompare with data collected alcohol
injuries to self or others. Over a span of l0-relveated
after the intervention was implemented (post- experienced
using the social norm media campaign, NI
a 37% reduction in perceived high.
test). To rule out other possible explanations of
the evaluation results (i.e., high-risk drinking is risk drinking, a 30% reduction in actual high-risk
reduced because of reasons beyond the interven drinking, and a 20% reduction in alcohol-related
tion), data on national drinking levels anmong injuries to self or others (Haines, 2003). In addi.
U.S. college students during this same time tion, the overall rate of binge drinking among
period were also recorded. This evaluation study [Link] students (i.e., the nonequivalent com
demonstrates a quasi-experimental design because parison group) remained virtually the samne over
a nonequivalent comparison group of U.S. col the same time period (Johnston, O'Malley, &
lege students was used. Bachman, 1999). These data suggest that the NIU
Data were collected at the end of each aca social norm intervention has been very successful
demic year from 1988 to 1998 (see Figure 4.3). in reaching its goals.

Figure 4.3
Effects of theSocial Norm Media Campaign at NIU on Students Perceptions of and Actual
High-Risk Drinking Behavior

80

70

60

50

40

30

20

10

1994 1995 1998

1988 1989
Baseline 1990 1991 1997

Perceived high-risk drinking rate*

Actual high-risk drinking rate
National college high-risk drinking rate** (Johnston et al., 1999)

SOURCE: Fron Haines (2003).

*High-risk drinking defined as having more than 5 alcoholic drinks when "partying."
**High-risk drinking defined as having5 or more alcoholic drinks at a sitting within the past 2 weeks.
 Lebalnom he disnore lhem
75
TChapter 4 Intervention and Evaluation

Further Applications initiated. One of the most effective interventions,

of Social Norm Theory developed by Stice and his coresearchers, relies on
the application of Festinger's (1957) cognitive
The social norm approach has also been used dissonance theory (discussed in Chapter 2). In
to reduce high-risk drining among students at four published articles, Stice and colleagues
many other colleges and universities across the report separate trial interventions that were
IInitedStates (Berkowitz, 2004). In fact, a survey based on the same programn logic model and
of 118schools in 2002 indicated that nearly half used essentially the same intervention strategy
bad adopted some form of norms-based inter (e.g., Stice, Chase, Stormer, &Appel, 2001; Stice,
vention to combat binge drinking (Wechsler Mazoti, Weibel, &Agras, 2000; Stice, Shaw,
et al., 2003). Social norm theory also has been Burton, & Wade, 2006; Stice, Trost, & Chase,
applied to other social issues where individuals 2003), The thinking behind the intervention is
tend to misperceive the norm, including, for that a woman's internalization of the thin ideal of
example, c1garette smoking, driving while intoxi- beauty (resulting from sociocultural pressures)
cated, seatbelt use, gambling, tax compliance, leads to body dissatisfaction, negative affect (e-g»
risky sexual behavior, and energy conservation. sadness, anxiety), dieting, and eating disorder
In 2006 the National Social Norms Institute symptoms. Therefore, they undermine awoman's
opened its offices onthe campus of the University internalization of the thin ideal, and improve
of Virginia with the mission of supporting ments should follow in the other negative factors.
research, evaluation, and the dissemination of Young women (early teens into the 20s) with
information related to social norms programs body image concerns who did not meet thecrite
with respect to a diversity of health issues and ria for having an eating disorder (e.g., anorexia
populations. The institute's website is an excel nervosa) participated in a 3-hour (spread over
lent source of information concerning current three occasions) dissonance intervention in which
and past work on social norms interventions they voluntarily argued against and critiqued the
htp://[Link]).7 thin ideal (e.g., wrote essays, role-played).
The application of socialnorm theory in pro- According to Festinger's (1957) theory, such coun
grammatic interventions that successfully reduce terattitudinal behavior would generate psycho
a variety of high-risk behaviors aptly demon- logical discomfort (called dissònance) that would
of ben motivate the indiyidual to internalize the thin
strates the potential for the development
eficial interventions based on social psychological ideal less, which in turm would cause decreases in
theory and research. The following section body dissatisfaction, dieting, negative affect, and
describes two other examples of applying social eating disorder symptoms. The key variables were
psychological theory to intervention design and measured by interviews and surveys for each study
also underscores that many social problems can at pretest and posttest and then, depending on the
be very complex and call for complex solutions. study, at follow-ups of 1,3, 6,or 12 months. Across
the four studies and all follow-up intervals, com
pared with control participants who had also
reported eating disorder concerns, the women in
OTHER INTERVENTIONS the dissonance condition showed reductions in
the three risk factors (i.e., thin-ideal internaliza
Examples of Other Interventions tion, body dissatisfaction, and negative affect) and
Among the many excellent examples of how also in the dieting and eating disorder symptoms.
social psychological theory can be applied suc Stice, Marti, Spoor, Presnell, and Shaw (2008) con
cesfully to the amelioration of social problems is ducted a 2- and 3-year follow-up of the interven
the work of Stice and colleagues on eating disor- tion reported in Stice and colleagues (2006) and
ders. In the context of the severe functional notably found that all of the reductions in the key
impairment and health problems, including mor factors had been sustained for either 2 or 3 years,
tality, commonly associated with eating disorders as well as finding a lower risk of the onset of an
and the reluctance of many people suffering from eating disorder at 3years. Extrapolating from the
eating disorders to seek treatment, alarge number latter result, the researchers estimated that out
of eating disorder prevention programs have been of every 100 women who participated in the
76 PART I FOUNDATIONS OF APPLIED SOCIAL PSYCHOLOGY

intervention, 9 fewer would develop eating pathol serious social issues confronted by
ogy. Needless to say, these results are remarkably reflect very complex problems that social scientists
in a highly interconnected cluster are
promising, perhaps especially for a relatively
ple 3-hour intervention. Add to this that
sim
Stice and ables and, thus, are apt to be
to solve. Examples include extremely
embedded
of causal
vari-
colleagues (2008) cite five other research teams
who also have demonstrated the efficacy of the
environmental
ability, raising a child, mental
Highly complex problems tend and
illness, chal enging
sustain-
dissonance intervention. Clearly this is an excel-
lent example of an evidence-based intervention. resource-intensive, multifaceted
to
holistic interven
povert
requirey.
There are also interventions that incorporate tions that are able to take into
consideration the
more than one theory into their designs. For interconnectedness of multiple key variables and
instance, Hansen, Meissler, and Ovens (2000) used address multiple causes. The
aspects of social learning theory and the funda following quotation
from David Shipler (2004) centering on the issue
mental attribution error to create group play- of povertyaptly illustrates what we are suggestin
therapy programs for children with attention
deficit/hyperactivity disorder (ADHD). The major Every problem magnifies the impact of the
tenet of social learning theory is that we learrn others. Andall are so tightly interlocked that
new behaviors by observing and imitating others one reversal can produce a chain reaction with
(e-g., role models) as well as by observing conse results far distant from the original causes. A
quences of behaviors (Bandura & Walters, 1963). rundown apartment can exacerbate a child's
asthma, which leads to a call for an ambulance.
The group play-therapy program allows children which generates a medical billthat cannot be
to learn from others through the process of mod paid, which ruins a credit record, which hikes
eling and the enforcement of positive or negative the interest rate on an auto loan, which forces
consequences for actions. As reported in Chapter the purchase of an unreliable used car, which
1,the fundamental attribution error refers to the jeopardizes a mother's punctuality at work,
notion that people tend to underestimate the role which limits her promotions and earning
of situational determinants of people's behavior capacity, which confines her to poor
and tend tooverestimate the role of dispositional housing.... If problems are interlocking, then
determinants (e.g., personality) (Ross, 1977). For so must solutions be. A job alone is not
children with ADHD, this means that the behay enough. Medical insurance alone is not
iors that they display (e.g., excitement) tend to be enough. Good housing alone is not enough.
attributed (e.g., by significant others) to their Reliable transportation, careful family
budgeting, effective parenting, and effective
condition, whereas situational influences (e.g., fun schooling are not enough when each is
events) are ignored. The group play-therapy inter achieved in isolation from the rest. (p. 11)
vention attemptstoreduce this bias in judgment
by conducting the therapy groups in a commu The following are two intervention efforts
nity-based setting away from stigmatizing mental that have attempted to take a mnore holistic
health facilities and by increasing awareness of the approach: (a) comprehensive community initia
attribution error and its potential consequence for tives, which aim to transform distressed neigh
children with ADHD.
borhoods by engaging a variety of neighborhood
stakeholders (e.g., residents, community agen
Complex Problems Require cies, businesses,government) around a common
Complex Solutions plan for change (Kubisch, Auspos, Brown &
Delwar, 2010); (b), wraparound initiatives, which
Although many of the problems that we target high-risk children and their families by
have considered in this chapter for the pur assembling a diverse team of formal and infor
poses of explaining the design and evaluation of mal support persons and agencies in their com
interventions/programs were presented as rela munity (e.g., community leaders; representatives
tively unidimensional, it is important to recognize of natural or informal community support.
that many social and practical problems that occur agencies/groups such as recreation, faith, busi
in our incredibly multidimensional and complex ness, and service clubs; representatives of_the
physical and social world are anything but unidi -formal child, family, and adult services) (Bruns
mensional. Without a doubt, many of the most & Walker, 2008).
 Chapter 4 Intervention and Evaluation 77

For the Applied Social Psychologist, It Is Not Enough

to Know Only That an Intervention Has Worked
Although this chapter has 'emphasized the importance of developing a program logic model to
explain why an intervention will result in a particular outcome or set of outcomes, the applied
social psychologist involved in designing and evaluating interventions may be the only stake
holder who is particularly concerned about understanding the intervening processes that link
intervention to outcome. Generally speaking, other stakeholders care little, if at all, regarding
how intervention outcomes come about as a result of program activities (cf. Mark & Bryant,
1984). For them, the bottom line is whether the intervention works. An overriding emphasis on
results makes sense for most stakeholders because their investment in the intervention is pre
dominantly practical. Program funders want visible, cost-effective returns on their dollars, pro
gram managers and staff members want to demonstrate their value to the organization through
program accountability, and the only important thing for program recipients is that they experi
ence the purported benefits of the intervention. Needless to say, it is of great importance to the
applied social psychologist that the intervention proves to have its intended beneficial conse
quences. However, unlike other stakeholders, focusing on the practical aspects is not enough. As
a scientist, it is necessary for the applied social psychologist to avoid giving insufficient attention
to understanding why an intervention works and why it does not work. It is vital that the applied
social psycholoist does not compromise the scientific integrity of the intervention-evaluation
process by dweling onlyon the practical "does it work" question (i.e.,the question of cause and
effect) and giving insuffIicient attention tothe why it works" question (i.e., the question of expla
nation). For example, if we were to develop an intervention aimed at improving grade school
students' academic performance by improving their self-concepts, it would not be sufficient fram
asocial scientific perspective only to demonstrate that the students' grads improved after the
intervention; it also would be important to confirm that their self-concepts improved significantly
and that students whose grades increased the' most tended to be those whose self-concepts
improved the most, and those whose grades increased the least had self-concepts that improved
the least. Evidence that addresses the program logic model as a whole is required for applied
social psychology to move toward fulfilling its potential as a social science to contribute to a bet
ter world (Ross et al., 2010).

INFLUENCING SocIAL POLICY animosity of the Arab world against the United
States, education failures, addictions) have psy
This chapter would be remiss if it did not recog- chological causes and/or consequences and that
psychologists need to be heard and to be at the
nize another avenue of application-influencing
the development of social policy. The potential of table of influential leaders and policy makers
psychology and other social sciences to contrib because psychologists have more to say about
ute to public policy has long been recognized. these issues than do members of any other disci
Also long recognized is the existence of a consid pline" (p. 432).
erable gap between the actual amount of social Probably the best-known example of social
scientific knowledge about social issues and the psychological research haying arole in policy
impact of this knowledge on social policies formation occurred in(1954 when the U.S.
(Hennigan, Flay, & Cook, 1980; Miller, 1969). Ten Supreme Court, in its Brown v. Board of Education
years ago two recent American Psychological separate
ruling, struck down the 1892 Plessyv Ferguson
but equal doctrine, thereby making
Assoçiation (APA) presidents, DeLeon (2002)
and Zimbardo (2002), challenged psychologists racial segregation in public schools unconstitu
to become more involved in public poticy devet tional (Benjamin & Crouse, 2002). In making its
opment. Zimbardo noted that many of the most ruling, the court cited seven social science publi
serious problems facing the United States (e.g., cations as having a role in its decision. The main.
 78 PART I FOUNDATIONS OF APPLIED SOCIAL PSYCHOLOGY

thrust of the social science argument was that science disciplines, including social
segregation negatively affected the self-esteem of However, instead of focusing on social poli psychology.
African American children and also fostered influence, this chapter (and this book overaln
0nterracial prejudice. This event is particularly the
concentrates on the role of social psychology in
design of interventions that are viewed as the
noteworthy because it marked the first time that
the Suprenme Court had recognized psychological sine qua non of applied social psychology. The
research in a decision, and the ruling was perhaps path from social psychological knowledge to
the most socially signitcant one made during the infuencing social policy is seldom straightfor.
20th century. In recognizing the 50th anniversary ward. Many factors that preclude the direct use
of Brown v. Board of Educationand the important of knowledge in policy decision making have
inluence of Kenneth [Link] on the court's deci- been identified (e.g., Hennigan et al., 1980),
sion, Tomes (2004) appropriately titled his article including the failure of researchers to adequately
"The Caseand the Research-That Forever communicate research findings to policy mak
Çonnected Psychology and Policy. ers, time presures on decision makers that
Another example of social science research undercut thorough assimilation of research evi
contributing to policy development comes from dence, resistance to change stemming from com
the criminal justice field in North Amnerica, fort with the status quo, pressures from
which during the past 25 years has witnessed a stakeholders whose positions do not accord with
shift in philosophy and operational strategy. The the scientific evidence, and so forth. In the case
change has been from less emphasis on the tradi- of the development of interventions, the route
stional law enforcement model of policing (a from knowledge to practice is more direct in
reactive, response-driven approach) to more that there are likely to be fewer obstacles to
emphasis on the community-oriented model of application than is the case for social policy. In
Apolicing (a proactive, police-community part Chapter 12 further consideration is given to the
nershíp approach). In embracing this transition, topic of influencing social policy.
leaders in policing and relevant government
agencies drew considerably on the research evi
dence that pointed to the limited effectiveness INTERVENTION IssUES
of the core operational procedures used in tra
ditional policing and to the need for more Process Issues
attention to quality-of-ife issues andimproving
police-community relations (Schneider, Pilon, The focus of this chapter has been on program
Horrobin, &Sideris, 2000). matic intervention. We have described the basic
Thus,social psychology can contribute to both steps in the design of interventions, and although
the development of policy and the development of these steps are fundamental to the development of
interventions. The difference between policy and effective programs, it is important to recognize
intervention is that policy refers to a general that they are not carried out in a vacuum. Al
course of action endorsed by an organization, interventions-and, byextension, intervention
whereas an intervention refers to the specificcon designs -have to operate within constraints, and
crete action(s) (ie, program activities) that the it is sometimes very challenging to both work
organization chooses to take to implement the within these constraints and maintain the integ
policy. For instance, on the one hand, the leader rity and effectiveness of the interventions. The
ship of apolice agency may decide on the policy of most prevalent constraint is budgetary. Other
implementing community policing on an organi constraints are subtler. Following are three general
zation-wide basis. On the other hand, interven constraints.
tions pertain to exactly how the agency chooses to First, interventions must be paid for, and
implement the policy, that is, the actual steps taken funding typicatty is rooted in poticy directions
to carry out the policy of community policing that, in turn, are embedded in the funder's
(e.g., involving oficers in courses on community broader political philosophy or ideology (whether
policing principles, setting up a neighborhood conservative or liberal). Thus, the view that inter
mini-station, conducting foot patrols). ventions are driven exclusively by the emergence
Without adoubt, influencing social policy is of empirically validated methodologies is naive
avery important applied function of the social (Mark &Bryant, 1984). To be sure, empirical
 Chapter4 Intervention and Evaluation 79

validation is critical to the successful application parameters, for example, getting more flexible
of social psychological theory and knowledge, intake or referral criteria to give access to more
but interventions based on such insight must, to people and getting agreement from management
varying degrees, conform tofunders' needs and to change the program delivery model in response
wants. Ideas for interventions, no matter how to participant feedback.
well validated, will always have to be vetted for
approval by those who have been approached to Ethical Issues
[Link].
Second, intervention design is a collaborative As this chapter has illustrated, applied social
process in which the responsiblity always is shared psychologists contribute to improving the quality
among stakeholders. For example, intervention of life andthe betterment of society as a whole
design often involves professionals from different through the application of social psychological
disciplines (sometimes together with program theory and research. Recall from Chapter 3that
participants)working as ateam. Hence, designing social psychological research undertaken to
an intervention to assist in relapse prevention in a develop an understanding of social phenomena is
substance abuse clinic is likely to involve--in guided by codes of ethics (American Psychological
addition to the applied social psychologist-- Asociation, 2010: Canadian Psychological
Association, 2000), Just as ethical issues exist in
medical professionals and clinical psychologists as
well as social workers and counselors. Each profes-the conduct of basic research, concerns also can
sional group will contribute a unique experience arise around the ethical use of both research
and perspective on the design of the intervention. methodology and research findings in applied
To be efective, therefore, an intervention's settings. In fact, the focus on effecting change in
designer(s) must be able to take into account a the functioning of individuals, groups, and orga
wide range of stakeholder ideas while at the same nizations raises a host of ethical issues beyond
time ensuring that the key elements of the design those that typically emerge in understanding
remain linked with one another both logically and focused research. Ethical issues can arise in both
operationally to form a cohesive and integrated the design and the evaluation of interventions,
whole. Also note that the challenge is not only and they can challenge a practitioner's profes
intellectual but also interpersonal given that lis- sionalism and integrity.
tening, relationship building, and negotiating Although psychologists are guided by their
national associations' codes of ethics, and often
become part of the toolbox of everyapplied sociat
psychologist involved with intervention design: by specialty-specific [Link] ethics such as those,
(You likely will find that your grounding in basic for clinical and educational psychologists,some
social psychology is extremely helpful in this have [Link].(e.g., ONeill, 1989, 1998) that
regard. For an excellent chapter on the practice ethical standards for much of applied psychology
skills of the applied psychologist, see Fisher, 1982.) focus on issues that typically arise between apsy
Third, constraints that are specified by the chologist and an individual client. These writers
law, and by various organizational policies and point out that these ethical guidelines have been
established procedures, also exist. Organizational designed specifically to protect the individual V
constraints that must be accommodated include rights of clients (e.g., maintenance of confidenti
such, they
restrictions to access, availability of facilities, ality, prohibition of sexual contact). Aswith issues
hours ofoperation, staffing availability, and are not wholly adequate for dealing
that arise with regard to programmatic interven
scheduling restrictions. tions that by definition involve multiple stake
None of these constraints can be ignored, so
all must be addressed in the process of interven holders (e.g., program participants, program
personnel, funders, members of the community
tion design and eventual implementation.
Sometimes it is possible to work around them, at large). Two key questions that arise from this
and sometimes it is not. For example, some con observation were voiced by ONeill (1989). The
stakeholder
straints are nonnegotiable (e.g., maintaining first question is to whom (i.e., which
confidentiality of client records, being required group) is the applied psychologist responsible?
to accept nonvoluntary participants), whereas The second question is for what (i.e., which end
others may be negotiable to varying degrees. product or deliverable) is the applied psycholo
Typically, designers have to advocate for certain gist ultimately responsible?
 80 PART I FOUNDATIONS OF APPLIED SOCIAL PSYCHOLOGY

As a practical matter, the contract) that is management to focus more on the positive than
drawn up between a practitioner (e.g., applied on the negative in the evaluation report, which
social psychologist) and his or her employing would, in effect, distort the overall findings of th
organization can be very helpful by providing evaluation. If the psychologist refuses to mine
some clarity with respect to both of these ques resent the findings, his or her
the agency (and possible futurerelationship with
tions. For example, contracts typically specify the
following: the individual or committee to whomn contracts) may
be jeopardized along with the survival of the
the practitioner will report, the scope of the work
agency. On the other hand, if the
expected, the end product of the work (eg., a does comply with the request, the peoplepsychologist
who
program evaluation report outliningconclusions need the services will not be served as well as they
and recòmmendations), timelines for the com could be, and financial and other
pletion of the work, agreement on who owns the resources pro
vided by the government will not be invested
data colected or copyright to any materials pro optimally.
duced, whether publication of the findings is It is important to note that these are just two
permissible, and so forth. Typically, contracts examples and that dilemmas can stem from
do not spell out generally accepted norms of many factors such as cross-cultural differences
professional conduct (e.g. confidentiality of (e.g., among client groups, funders, and policy
information, objectivity, prohibitions on profit makers) and competing value systems. Some
ing from privileged information) as adherence to argue (e.g., Riger, 1989) that such ethical dilem
these is assumed.
mas are essentially political conflicts arising from
Contracts, however, cannot anticipate every the presence of
many needs and scarce resources
dilemma that may occur. Ethical issues in pro and the
gram intervention can arise when the applied resulting necessity to compete for these
resources. Kimmel (2004) suggested that applied
social psychologist is aware of, or discovers, social psychologists
information that, if divulged, will further the sions with the have frank and open discus
interests of one stakeholder group at the expense on any organization prior to embarking
of another. As an example, consider an appied The purpose program design or evaluation project.
of the initial discussion is to clarify
social psychologist who, is hired to head the the
psychologist's role as well as to articulate the
design of an intervention and finds that the professional and ethical standards and any limits
employing organization (a government agency) they may place on the extent to which the
only supports a politically popular intervention orga
nization can be assisted. In an evaluation context
that, it turns out, has questionable support in like the second example above, it would be
evaluation research. The dilemma is that if the imnportant to discuss up front the purpose of the
applied social psychologist does not reveal the evaluation and the implications of negative find
lack of evidence for the program's hypothesis, it ings; for instance, are there serious funding
is likely that considerable resources will be wasted
on an intervention that is ineffective, and so the implications, or can the findings be used as a
intended program recipients will not be well developmental opportunity for the program?
served. On the other hand, if the lack of evidence Another approach to preventing or resolving
is revealed, not only may the psychologist's job be ethical dilemmas is to involve various stakeholders
at stake, but the program may well be imple (program recipients, front-line staff, managers
board members, funders, etc.) in making key deci
mented anyway. sions regarding the intervention design or evaluaf
To take a parallel example from intervention
tion project.A multistakeholder steering committee
evaluation, consider the case of an applied social would help to address power imbalances, give rec
psychologist who has been hired to conduct an ognition to different stakeholder needs, promote
outcome evaluation of a major program offered transparency, and ensure the sponsoring organiza
by a social service agency. The evaluation deter tion is held accoåntable for its decisions.
mines that although the program benefits the
program participants in incidental ways, it is it isKimmelnecessary
(2004)pointed out that occasionally
for the applied social psychologist
largely ineffective. Such results may lead to the to refuse to participate
closure of the agency because this program rep potential to conflict withinthe projects that have the
ethical standards of
resents the bulk of the funding that the agency the
receives. The practitioner is asked by the agency's gist, profession. For the applied social psycholo
there is rarely an "easy" way out of ethical
 Chapter4 Intervention and Evaluation 81

dilemmas. It must be remembered that one's (b) developing a solution, (c) goal setting and
expertise carries with it the responsibility to designing the intervention, and (d) implement
always act in ways that placethe interests of those ing the intervention. Evaluation of the effective
who are in the greatest need of help above the ness of interventions can be considered as a fifth
self-interests of the professional. This is the step in the process of program
essence of acting ethically. development.
There are two main types of program evaluation:
process evaluations, which focus on whether pro
grams have been implemented as planned, and
SUMMARY outcome evaluations, which focus on whether
programs meet their objectives. It is important to
This chapter outlines some of the key compo evaluate programs because they are not always
nents involved in the development of interven effective and sometimes produce unintended
tion strategies designed to address social and undesi
undesirable consequences.
practical problems. Although we can design per The chapter also presented a detailed example
sonal interventions, the focus is on programmatic of an intervention aimed at reducing problem
interventions or programsa collection of drinking at NIU. This intervention demonstrated
activities designed to address social or practical the value of using theory in the design of pro
problems with the intention of preventing, reduc grams. The chapter then explored how the results
ing, or eliminating their negative consequences. of interventions can be used to inform public
There are four steps involved in the develop- policy and, finally, noted anumber of intervention
ment of programs: (a) identifying the problem, 0ssues,
issues, including
including program
program constraints and ethics.