I. Introduction

A major goal of the National Network of Libraries of Medicine (NN/LM) is to improve health professionals' access to health information, especially in rural, underserved, and minority communities. In order to achieve this goal, NN/LM has requested experts from diverse fields to write about effective outreach theories and evaluation practices. The purpose of this paper is to outline outreach evaluation theories and practices based on the empirical research of the Health Communication field.

Most Health Communication theories are models of individuals' health-related and health information-seeking behaviors. As such, they identify important variables and specify how these variables work together to produce a desired outcome (such as, utilization of medical resources). Through careful laboratory and field research, much knowledge has accumulated on how to effectively motivate desired behaviors. Evaluation efforts measure variables from these models and based on the strength of these variables, can determine whether or not individuals are engaging in the desired behaviors. If evaluations indicate that certain variables are not receiving the attention needed, then new outreach interventions can be developed that address missing or neglected issues in order to achieve optimal success.

It is important to note that the variables and frameworks identified here offer the basic structure for effective outreach efforts. However, how one specifically defines and uses each framework is entirely up to the target population. For example, every car has a basic structure including doors, windows, and an engine, but each individual company varies the color, size, shape, texture, etc., of the car's basic structure. In other words, they fill in the details according to the target audience to whom they're trying to sell the car. In the same manner, the structure offered by the variables and models identified here are useful because they have been proven to work over and over again across many different topics and populations. But, the details, how one defines each concept, and even the appropriate framework must be generated in collaboration with community members or cultural insiders who know what works and appeals to the target audience.

Following is a description of the key variables found in most health behavior change models, a description of how these variables work together in outreach interventions, a review of evaluation methodologies applied to these outreach interventions, and ideas for models of outreach and outreach evaluation.

II. Key Variables

Following are definitions and examples of the important variables found in most health behavior change models in the Health Communication literature. Section III describes how these variables work together in a theoretical sense.

Threat

A danger or harmful event that exists in the environment that people may or may not be aware. For instance, a lack of pertinent information may be a threat, because this lack of information may cause harm to a patient. Comprised of the following two dimensions:

Severity

The magnitude of harm expected from a threat. The significance or seriousness of a threat. The degree of physical, psychological, or economic harm that can occur. For example, is AIDS severe and serious? Is there any danger in not having up-to-date information?

Susceptibility

The likelihood that a threat will occur to a given person or audience. The degree of vulnerability, personal relevance, or risk of experiencing a threat. For example, am I susceptible to getting AIDS? Are you at-risk for falling behind current medical knowledge?

Fear

A high level of emotional arousal caused by perceiving a significant and personally relevant threat. Fear motivates both protective and maladaptive action, depending on the circumstances. For example, sometimes fear motivates one to seek out information. However, sometimes fear causes people to deny they are at-risk for experiencing a threat or causes them to defensively avoid a threat and thereby ignore recommended responses. Fear can be one of the most powerful influences on behavior, if it is channeled in the right direction.

Efficacy

Efficacy refers to the effectiveness, feasibility, and ease with which a recommended response impedes or averts a threat. For example, can your audience easily and effectively find the information needed to avert a threat? Efficacy is comprised of the following two dimensions:

Response Efficacy

The degree to which the recommended response effectively averts the threat from occurring. Sometimes response efficacy is called "Outcome Expectations" (if you perform a certain behavior, what do you expect the outcome to be?). For example, do people believe condoms prevent HIV transmission? Do people believe that having current medical information prevents harm to patients?

Self-Efficacy

The degree to which the audience perceives that they are able to perform the recommended response to avert the threat. Sometimes self-efficacy is called "Efficacy Expectations" (what do you expect will happen if you attempt to perform a certain behavior?). For example, do people believe they are able to use condoms to prevent HIV transmission? Do people believe they are really competent and able to access the information needed to make good medical decisions?

Barriers

Barriers are anything that inhibits one from carrying out a recommended response like cost, time, different language, cultural differences, etc. Some see barriers as the inverse of self-efficacy in that barriers inhibit one's perceived ability to carry out a recommended response. Therefore, one can think of barriers as "barriers to self-efficacy." Examples of barriers for the NN/LM might be a lack of computer skills, lack of electricity in remote areas, articles too full of jargon and too difficult to read, language issues, cultural issues, etc.

Benefits

The rewards or positive consequences that occur from performing a recommended response. Do people see any benefit to performing a certain behavior? Benefits are somewhat similar to response efficacy.

Subjective Norm

One's motivation to comply with what one believes his or her important referents believe. Notice this definition has two parts. First, your subjective norm is based on what you believe your significant others believe (notice, the focus is on what you believe they believe, not what they actually believe). The more you are motivated to comply with a certain significant other or referent, the stronger the subjective norm. For example, if you are motivated to comply with your boss and your boss thinks it's a good idea that you utilize up-to-date information sources, then your subjective norm is strong for you to utilize up-to-date information sources.

Attitudes

An evaluation of an object, recommended response, or a belief. For example, "computer's are good," "too much information is bad," "anything that takes time is undesirable," are all attitudes.

Intentions

Your plans to carry out a recommended response or do a certain thing. For example, "I intend to use the internet daily." Intentions are often used as proxies for actual behavior.

Behaviors

The actual action carried out. For example, did you or did you not use the internet daily last week?

Defensive Avoidance

A motivated resistance to a recommended response, usually occurring at the subconscious level. Characterized by a lack of attention or remembrance of a particular concept, threat, or recommended responses. For example, following are typical defensive avoidant responses: "It's too overwhelming, I'm just going to not think about it" and "I don't want to know anything about X, I'm just going to block out anything I hear about it." Defensive avoidance occurs when one perceives a significant and relevant threat but believes there is nothing he or she can do to effectively avert the threat.

Reactance

Another form of motivated resistance where one becomes angry at an issue or source and "reacts" against a recommended response. Typically characterized by perceptions of manipulation (e.g., "they're just trying to manipulate us into using this stuff, it doesn't really make a difference") or message/issue derogation (e.g., "this stuff is stupid, I know all I need to know"). Reactance occurs when one feels threatened but feels unable to do anything to effectively avert a threat (e.g., "I guess it's important to have up-to-date information, but there's no time and no money for me to get what I need…").

Cues to Action

External (e.g., public service announcements, informational flyers) or internal (e.g., symptoms of an illness, such as an itchy or bleeding mole) pieces of information that trigger decision-making actions.

Stages of Change

A classification scheme that suggests five stages to the performance of a behavior: Precontemplation, Contemplation, Preparation, Action, and Maintenance. For example, people are in different stages of readiness with regard to given behaviors. Different persuasive strategies are called for in the different stages of change. It is important to determine which stage your client or audience is in before developing a campaign. For example, if your audience is completely unaware of certain technological innovations, they have not begun to even consider using such innovations and would be placed in the "precontemplation" stage. An "awareness and knowledge" campaign would be appropriate for this target group to move them from the "precontemplation" stage to the "contemplation" and "preparation" stages. A "motivational" campaign would be needed to move people from the "contemplation" and "preparation" stages to the "action" and "maintenance" stages."

Social Marketing

The use of marketing principles and ideas to "sell" a pro-social idea or belief. This approach promotes the use of market research to discover important demographic and psychographic characteristics of one's target audience. Then, it uses the Four P's of marketing to "sell" the concept: product (the behavior or the product that you want the target audience to adopt), price, promotion, and place. A popular approach used in many outreach campaigns. It is important to note that social marketing is an approach, and not a theory. Thus, it is best used in conjunction with a theory (which would offer guidance on which variables to study and how these variables work together to produce desired outcomes).

Danger Control

When people believe they are at-risk for a serious or significant threat (i.e., high perceived threat), they believe they are able to effectively avert it from occurring (i.e., high perceived efficacy), they are motivated to control the danger or threat. When people are motivated to control the danger, they change their attitudes, intentions, and behaviors. These changes result in the individual's adoption of behavioral recommendations in the message and subsequent enactment of the behavioral recommendations.

Fear Control

When people believe they are at-risk for a serious or significant threat (i.e., high perceieved threat), but they believe they are unable to perform the recommended response or they believe the recommended response to be ineffective (i.e., low perceived efficacy), then they focus on controlling their fear about the threat. When people control their fear they do not control the actual danger. Instead, they control their fear by denying they are at-risk for the threat, defensively avoiding the threat, or reacting angrily toward those trying to help them.

III. Interventions: Putting the Variables Together

Most of the health behavior change models in the health communication field focus on the variables of threat, efficacy, and barriers. The four most common ones are (a) the Health Belief Model, (b) the Extended Parallel Process Model (a fear appeal theory), (c ) the Theory of Reasoned Action, and (d) Social Cognitive Theory. In addition to these four theoretical models are two approaches or schemes often applied to health behavior change interventions - the Stages of Change (or Transtheoretical Model) and Social Marketing. All of these are discussed below. 1

The Health Belief Model

The health belief model (HBM) (Janz & Becker, 1984; Rosenstock, 1974) is one of the most commonly-used models of health behavior change and is probably the most frequently-taught model in outreach intervention courses. Many have used it to guide the development of intervention and evaluation efforts and its influence on health communication research is enormous. It was developed as an overarching framework on how to promote preventive behaviors (such as immunizations) by a group of social psychologists in the early 1950's (Janz & Becker, 1984). Briefly, the HBM suggests that preventive health behavior is influenced by five factors: (a) perceived barriers to performing the recommended response; (b) perceived benefits of performing the recommended response; (c) perceived susceptibility to a health threat; (d) perceived severity of a health threat; and (e) cues to action (see Figure 1).

Figure 1: The Health Belief Model (HBM)

The HBM suggests that individuals weigh the potential benefits of the recommended response against the psychological, physical, and financial costs of the action (the barriers) when deciding to act. For example, a nurse may realize the benefit of having up-to-date information but may lack access, the skills, or even the transportation needed to get to a library. In this case, the barriers would outweigh any benefits and the nurse probably would not seek out up-to-date information. Similarly, the HBM suggests that individuals evaluate whether or not they are really susceptible to a threat and whether or not the threat is truly severe. Rosenstock (1974) has noted that the combination of perceived susceptibility and severity provide the motivation for action, and the comparison of perceived benefits to perceived barriers provides the means or pathway to action. Thus, the stronger the perceptions of severity, susceptibility, and benefits, and the weaker the perception of barriers, the greater the likelihood that health-protective actions would be taken.

Demographics and prior experiences are said to affect the four variables just described (i.e., perceived susceptibility, severity, benefits and barriers) as are "cues to action." For example, a physician skilled in computer use and use of the internet is much more likely to see the benefits of using Grateful Med and see few barriers to its use compared to a physician who still uses a typewriter for all of his correspondence. In terms of cues to action, there can be external cues (such as television shows or mass mailings) and internal cues (such as symptoms) which are suggested to increase perceptions of susceptibility and severity, and in turn trigger the decision-making process whereby perceived barriers and benefits are weighed against each other.

The HBM has been empirically tested as the basis for educational campaigns on a number of health behaviors including bicycle helmet use (Witte, Stokols, Ituarte, Schneider, 1993), vaccination for infectious diseases (Larson, Bergman, 1982), adolescent fertility control (Eisen, Zellman, & McAlister, 1985), and risky sexual practices (Vanlandingham, Suprasert, Grandjean, & Sittirai, 1995). Overall, percevied barriers have been the strongest predictor of whether or not individuals engage in health-protective behaviors, followed by perceived susceptibility (Janz & Becker, 1984). Janz and Becker (1984) found that the perceived severity was the weakest predictor across studies employing the HBM. The HBM may be viewed as the grandmother of most modern health education theories. As such, its variables and principles can be seen in many of the other models to be discussed in this report.

The Extended Parallel Process Model

Most campaigns either intentionally or unintentionally raise anxiety or fear in audiences, because they focus on a health, physical, or social risk. For example, what is one reason people should use Grateful Med to get the latest information on certain medical procedures? Answer: To prevent the threat of malpractice and mistreatment of patients. This area of research is called "fear appeal" research, which is the study of effective risk messages.

Fear appeals are defined as persuasive messages that scare an audience into adopting a recommended response and are used all of the time by politicians, advertisers, and (especially!) parents. The first part of a fear appeal typically focuses on a threat by emphasizing the magnitude of harm of the threat (i.e., severity) and the probability that the threat will occur (i.e., the audience's likelihood of experiencing that threat). The second section of a fear appeal usually focuses on the efficacy (e.g., effectiveness) of the recommended response in terms of (a) response efficacy and (b) self-efficacy. When addressing efficacy issues, the fear appeal usually will describe how the recommended response effectively averts or minimizes the threat (response efficacy) and it will outline specific, easy-to-understand steps on how to avert the threat, in an attempt to increase perceived self-efficacy. The fear appeal may be thought of as the "cue to action" in the health belief model sense.

The research into fear appeals has shown them to be potent persuasive devices, but only in certain conditions. The most recent fear appeal theory, the Extended Parallel Process Model (EPPM) (Witte, 1992a, 1992b, 1994, in press; Witte et al., 1993), is based on Leventhal's danger control/fear control framework and is an expansion of previous fear appeal theoretical approaches (Janis, 1967; Leventhal, 1970; Rogers, 1975, 1983). Readers will note similarities between the Health Belief Model and fear appeal theories. Fear appeal models can be thought of as experimental variants or explanatory (as compared to descriptive) versions of the HBM.

Figure 2: The Extended Parallel Process Model (EPPM)

According to the EPPM (see Figure 2), the evaluation of a threat initiates two appraisals, which result in either danger control or fear control processes. First, persons appraise the threat of the hazard by determining whether they think the threat is serious (e.g., "is lack of information a serious problem that can cause harm to my patients?") and whether they think they are susceptible to the threat (e.g., "is it possible that I don't have up-to-date information on X technique?"). The greater the threat perceived, the more motivated individuals are to begin the second appraisal, which is an evaluation of the efficacy of the recommended response. When people think about the recommended response, they evaluate its level of response efficacy (e.g., "Will I get accurate and useful information off of Grateful Med?") and their level of self-efficacy (e.g., "Am I capable of using Grateful Med? Do I have access to it and the skills needed to use it?"). When the threat is regarded as trivial or irrelevant (perceived as low), there is no motivation to consider the issue further; the efficacy of the recommended response is evaluated superficially--if it is evaluated at all--and no response is made to an outreach message. If people do not feel at-risk for a threat or do not feel the threat to be significant, they simply will ignore information about the threat.

When both perceived threat and perceived efficacy are high, then individuals will be motivated to control the danger and adopt the recommended response. Danger control processes are primarily cognitive processes where individuals (a) believe the threat is serious and that they are at-risk for experiencing negative consequences from the threat (high perceived threat), (b) become fearful of this serious and significant threat and become motivated to protect themselves, (c) believe they are able to effectively deter the threat (high perceived efficacy), and (d) deliberately and cognitively confront the danger (e.g., "I'm going to install a modem and use Grateful Med now"). The cognitions or thoughts occurring in the danger control processes elicit protection motivation, which stimulates adaptive actions such as attitude, intention, or behavior changes that control the danger. Thus, when laypersons perceive themselves to be vulnerable to a serious risk, and they believe they can do something to effectively and easily avert that risk, then they protect themselves against the threat.

However, at some critical point, when persons realize that they cannot prevent a serious threat from occurring, either because they believe the response to be ineffective and/or because they have low self-efficacy and believe they are incapable of performing the recommended response (e.g., "I know I need current information but I don't have a computer and Grateful Med's too complicated to use, things probably haven't changed anyways"), fear control processes will begin to dominate over danger control processes. Fear control processes are primarily emotional processes where people respond to and cope with their fear, not to the danger. Defensive motivation is elicited by heightened fear arousal, which occurs when perceived threat is high and perceived efficacy is low, and produces responses that control one's fear such as defensive avoidance or reactance. Studies have shown that fear appeals with high levels of threat and low levels of efficacy result in message rejection, and occasionally in boomerang effects (people do the opposite of what is advocated). Thus, when laypersons believe themselves to be vulnerable to a significant threat but believe that there's nothing they can do to effectively address the threat, then they deny they are at risk, defensively avoid the issue, or lash out in reactance. In this case, fears about a threat inhibit action and risk messages may backfire.

In short, according to the EPPM threat motivates action -- any action -- and perceived efficacy determines whether the action taken controls the danger (which is protective) or controls the fear (which inhibits protective behavior). Individuals typically weigh their risk of actually experiencing the threat (i.e., malpractice suit, patient not getting good care) against actions they can take that would minimize or avert the threat (i.e., will Grateful Med give me the information I need to save my patients? Can I really use it and is it easy to use?). For a successful campaign, it is critical that high threat messages are accompanied by high efficacy messages. If it is difficult or impossible to promote strong perceptions of efficacy, then one probably should not use fear-arousing messages because they may backfire.

Overall, risk messages using the fear appeal approach have been shown to be effective in a variety of domains including skin cancer prevention (Stephenson, 1993), pregnancy prevention (Witte, in press-b), radon awareness (Witte, Berkowitz, McKeon, Cameron, Lapinski, & Liu, 1996), tractor safety (Witte et al., 1993), nutrition programs (Wunsch, 1996), breast self-examination (Kline, 1995), and so on. With a little bit of pilot testing, it is fairly easy to determine an audience's existing perceptions of threat and efficacy and then target campaign messages to produce high levels of threat and efficacy, which should lead to danger control actions.

The Theory of Reasoned Action

Often messages created for outreach efforts are based on intuitive appeal, rather than sound methodology (Fishbein & Ajzen, 1981). Even if a theory is used to develop messages, campaigners tend to use the variables in the theory as guidelines about what to address without special consideration of the actual content or words in a message. For example, campaign designers might address the severity of a threat and the audience's susceptibility to that threat -- the theoretical variables -- in a message, but the actual words or images used to address these variables are not systematically chosen. Fishbein and Ajzen (1981) go so far as to conclude that "the general neglect of the information contained in a message and its relation to the dependent variable is probably the most serious problem in communication and persuasion research" (p. 359).

Fishbein and Ajzen (1975; 1981) suggest specific message construction and evaluation techniques based on their Theory of Reasoned Action (TRA). In TRA, Fishbein and Ajzen (1975) propose that a person's behavior is predicted by intentions, which in turn, are predicted by attitudes toward the behavior and subjective norm. Attitudes are predicted by behavioral beliefs and evaluations of those beliefs. Subjective norms are predicted by normative beliefs and the motivation to comply with those normative beliefs. Fishbein and Ajzen (1975) state that two sets of beliefs must be altered prior to behavior change: (1) beliefs about the consequences of performing a certain behavior and the evaluation of those consequences (attitude); and (2) beliefs about what other people or referents think about the behavior to be performed and the motivation to comply with those referents (subjective norm). Only when a message targets the salient beliefs of these variables do attitudes and subjective norms, and subsequently, behavioral intentions and behavior, change.

Figure 3: The Theory of Reasoned Action (TRA)

Table 1 illustrates how the TRA may be used to analyze a specific audience's behaviors in terms of utilization of Grateful Med. Table 1 indicates a physician's hypothetical attitude toward Grateful Med use. Recall that attitudes are comprised of one's beliefs toward the attitude object (in this case Grateful Med) multiplied by the evaluation of the individual beliefs (whether they are good or bad). This person believes that Grateful Med is easy to use, time-consuming, contains articles that are difficult to read and full of jargon, and contains so much information as to be overwhelming. These are his salient beliefs about Grateful Med. The strength of these beliefs (from .00 to 1.00) is indicated in the second column and the evaluation (ranging from -3 (unfavorable) to +3 (favorable)) of these attributes is indicated in the third column. These belief strengths and evaluations are multiplied individually and then summed to create the overall attitude toward condom use. For example, this person has a moderately strong belief that Grateful Med is time-consuming (.50) and anything that is time-consuming is evaluated unfavorably (-2). In addition, he does not believe very strongly that Grateful Med is easy to use (.60), but anything that is easy to use is evaluated very highly (+3). The sum of these products indicates that this physician's predominant attitude toward Grateful Med is slightly negative at -0.30.

Table 1a: Hypothetical Attitude Toward "Grateful Med"

Salient Beliefs	Strength of Beliefs	Evaluation of Beliefs	Product
1. "easy to use"	.60	+3	1.80
2. "time-consuming"	.50	-2	-1.00
3. "contains hard-to-read articles full of jargon"	.45	-2	-0.90
4. "contains too much information, is overwhelming"	.10	-2	-0.20
TOTAL ATTIUDE TOWARD GRATEFUL MED (sum)			-0.30

Subjective norm is calculated in a similar manner. Table 1b gives a hypothetical example of a physician's subjective norm toward Grateful Med. His salient referents (those people most important to her) include his chief of staff, patients, wife, and peers. His normative beliefs reflect what he thinks each of these specific referents think about him using Grateful Med (ranging from -3 (should not use) to +3 (should use)). His motivation to comply with each referent is indicated on a scale from 0 (not at all) to 3 (strongly). In each case, the normative belief and the motivation to comply is multiplied to yield a product. The sum of these products is the subjective norm. This person's overall subjective norm toward Grateful Med is slightly positive. The attitude and subjective norms are sometimes weighted according to their importance, and then combined, to influence intentions, which then influence behaviors. Thus, this physician's scores suggest that he would not use Grateful Med because his attitude was slightly negative and his subjective norm was also slightly negative, resulting in a neutral stance toward Grateful Med. A campaign directly targeting his negative salient beliefs regarding use of Grateful Med would be appropriate here.

Table 1b: Hypothetical Subjective Norm Toward "Grateful Med"

Salient Referents	Normative Belief	Motivation to Comply	Product
1. Chief of Staff	+1	.2	.20
2. Patients	+1	.3	.30
3. Wife	+1	.1	.10
4. Peers	-1	.7	- .70
TOTAL SUBJECTIVE NORM TOWARD GRATEFUL MED (sum)			-0.10

Overall, TRA is one of the few theories to offer a systematic approach to the construction of the content of a health education message. It has been applied to a number of health-related behaviors including the impact of health risk messages about tap water (Griffin, Neuwirth, & Dunwoody 1995), sexual practices and AIDS related-behaviors (Fishbein & Middlestadt, 1989; Fishbein, Middlestadt, & Hitchcock, 1991; Vanlandingham, Suprasert, Grandjean, & Sittirai, 1995), childbearing intentions (Crawford & Boyer, 1985), testicular cancer prevention (Brubaker & Wickersham, 1990), exercise in schoolchildren (Ferguson, Yesalis, Pomrehn, & Kirkpatrick), alcoholism (Fishbein, Ajzen, & McArdle, 1980), cigarette smoking (Norman & Tedeschi, 1989), and many others.

Social-Cognitive Theory

Albert Bandura's Social Cognitive Theory (sometimes called Social Learning Theory) has been used in a wide variety of interventions and evaluation efforts. It was the theory used in the Stanford 5-Cities project to prevent heart disease and more recently has been used in several AIDS-prevention projects. The focal point of the theory is on perceived self-efficacy. Self-efficacy is defined as "people's beliefs that they can exert control over their motivation and behavior and over their social environment" (Bandura, 1989, p. 128). In other words, perceived self-efficacy is what you believe about your capability to perform a certain action (your perceived self-effectiveness). Bandura (1977) views self-efficacy as the driving force of human behavior. "Efficacy expectations are a major determinant of people's choice of activities, how much effort they will expend, and of how long they will sustain effort in dealing with stressful situations" (Bandura, 1977, p. 194).

Another important construct in Bandura's theory is outcome expectations. Outcome expectations (called response efficacy in other models) refer to an individual's belief that a certain behavior will lead to a certain outcome. For example, "I believe that if I use Grateful Med I will get the information needed to effectively treat a patient" is an outcome expectation. It's what you think will happen if you take a certain action. Outcome expectations are different from efficacy expectations in that the latter is a person's belief on whether he or she is able to "successfully execute the behavior required to produce the outcomes" (Bandura, 1977, p. 193). For example, even if outcome expectations are high, efficacy expectations may be low (e.g., "but I'm not capable of using Grateful Med because I don't know how to use the system"). In short, according to social cognitive theory, a person can believe that certain actions lead to a particular outcome (outcome expectations), but this individual may doubt his or her ability to perform the action (efficacy expectations). According to Bandura (1977), only when efficacy expectations are high will people perform certain behaviors. Efficacy expectations can vary on dimensions of magnitude (level of difficulty of task; people may have different efficacy expectations for simple tasks than for difficult tasks), generality (specific to general), and strength (weak to strong) (Bandura, 1977).

Bandura (1977) states that an individual's self-efficacy perceptions are developed from four sources of information: performance accomplishments, physiological states, verbal persuasion, and vicarious experience. Participant modeling or performance desensitization are examples of performance accomplishments. Suggestion, exhortation, and self-instruction are what constitute verbal persuasion. Vicarious experience stems from live or symbolic modeling. Finally, physiological or emotional arousal comes from relaxation, biofeedback, and symbolic desensitization or exposure. In terms of NN/LM, actual training in how to use informational services would increase perceived self-efficacy the fastest.

Stages of Change Model

One of a number of stage models of behavior change, the transtheoretical model allows educators to determine which stage the majority of their target audience members are in, along a continuum of no action to consistent action (DiClemente & Prochaska, 1985). The model, also referred to as the stages of change model (SOC), suggests there are five stages to the performance of a behavior: Precontemplation, Contemplation, Preparation, Action, and Maintenance. In the Precontemplative stage, individuals do not intend to change their behavior because they are completely unaware of the behavioral options available to them. They may not realize they are engaging in a risky behavior or they may deny that their behavior puts them at risk for harm. In the second stage, however, this risk becomes apparent to the individual. Contemplation is the stage in which individuals begin to think about the behavior that is putting them at risk and to contemplate the need for change. In this stage, for example, an individual recognizes the need for more information. In the third stage, Preparation, individuals make a commitment to change and take some action to prepare for the behavioral change, like take a class on how to use computers. It is in the Actionstage that individuals perform the new behavior consistently. In this stage, for example, the individual may regularly use Grateful Med in his or her practice. In the Maintenance stage, the final stage of the SOC model, the new behavior is continued and steps are taken to avoid lapsing into the formerly risky behaviors. For example, a person might continue to improve his or her computer skills or sign up for additional informational resources.

The SOC model is useful to campaign designers for several reasons. First, individuals in different stages exhibit distinct behavioral characteristics (Weinstein, 1988). Thus, researchers can effectively analyze and segment a target audience according to their different stages of change. Then, practitioners can strategically design messages to move individuals through the stages (Maibach & Cotton, 1995). For example, if campaigners wish to design a campaign to promote a new service and they determine that the majority of the members of the target population are in the Contemplation stage, they can design messages to systematically move audience members through the Preparation, Action, and Maintenance stages. Similarly, if the majority of the target audience is in the Maintenance stage, educators can provide messages which reinforce and support the desired behavior.

This model has been empirically tested with a number of health topics including smoking cessation, sunscreen use, addictive behaviors, pregnancy prevention, and risky sexual behaviors (e.g., Grimley, Riley, Bellis, & Prochaska, 1993; Prochaska, DiClemente, & Norcross, 1992).

Social Marketing

One approach to campaigns that has been widely used by health educators in both the public and private sector is that of social marketing. Social marketing involves the design, implementation, and control of campaigns aimed at altering the level of acceptability of the social ideas or behaviors of a specific target group or groups (Kotler, 1984; Kotler & Roberto, 1989). It is the application of for-profit management and marketing technologies to pro-social, non-profit programs (Meyer & Dearing, 1996). Wallack (1989) suggests that one of the keys to this approach is the reduction of psychological, social, economic and practical distance between the target of the campaign and the behavior.

Kotler and Roberto (1989) outline five basic steps in the social marketing management process. The first step is an analysis of the social marketing environment immediately surrounding the particular campaign. Next, the social marketer must research the target-adopter population and segment the audience into groups with common characteristics. The third step involves the careful design of the campaigns objectives and strategies. It is in this step that the social marketer must consider four concerns basic to every campaign -- the four 'P's.

The four "P's" are product, price, promotion, and place or what has been termed the 'marketing mix.' The product is the behavior or the product that the target audience must change or adopt. Campaigns have targeted a number of health behaviors as "products" including condom use, contraception, and alcohol and drug-related behaviors. For example, in the Stanford Heart Disease Prevention Program (SHDPP), the "products" promoted were regular exercise, smoking cessation, dietary changes, and stress reduction, in order to preven heart disease. The second "P" in the marketing mix refers to price and includes any physical, social, or psychological cost related to compliance with a campaign. In the case of the SHDPP's Smokers' Challenge, the costs of joining the challenge included the money and energy expended in accepting the challenge, as well as the psychological costs of giving up smoking. The third component of the marketing mix, promotion, deals with how the product can be represented or packaged to compensate for the costs of adopting the recommended response. The Smokers' Challenge attempted to promote the contest by removing or reducing the financial cost of the program to make it more appealing to target audiences (Lefebvre & Flora, 1988). Place is the final component in the marketing mix and involves the availability of the recommended response. The designers of Smokers' Challenge attempted to make access to information about the program easily accessible. They even mailed information on how to quit smoking to households participating in the study. Social marketing has achieved widespread adoption and allows campaigns to carefully target their persuasive materials.

IV. Evaluation Methodologies

The typical evaluation methodology used in the health communication literature is the simple but powerful experimental comparison between individuals or groups randomly assigned to an intervention group or a control group. This type of experimental design can include a pretest and a posttest or a posttest only, and is depicted as follows with O = observations, X = intervention, and R = random assignment.

Evaluation Method I - Pretest/Posttest with random assignment

R	Intervention Group:	O1	X	O2
	Control Group:	O1		O2

Evaluation Method II - Posttest Only with random assignment

R	Intervention Group:		X	O1
	Control Group:			O1

The advantage of a pre-test/posttest design is that one can determine how much change there was before and after the intervention (presumably the control group should exhibit no change). However, some prefer to use a posttest only design because they are afraid that the pretest sensitizes individuals to respond in a certain way and may result in increased socially desirable responses where people indicate change because "they're supposed to." There is one additional design that can be used that combines the above and assesses for sensitization effects, called the Solomon 4 design:

Evaluation Method III - Solomon 4 design with random assignment

R	Intervention Group:	O1	X	O2
	Control Group:	O1		O2
	Intervention Group:		X	O2
	Control Group:			O2

In the Solomon 4 design, one can compare both intervention groups to each other and both control groups to each other to see if there are any significant differences between them. If there are, it is likely that the pretest and a significant effect on the second observation because the only difference between these groups is whether or not they received a pretest. It is important to note that one can have multiple intervention groups and compare them with each other and the single control group.

Random assignment is a key feature of these designs and is absolutely essential to ensure a valid and accurate comparison. In random assignment, it is presumed that any pre-existing differences in your subjects (such as skill level, intelligence, race, etc.) will be evenly distributed between the intervention or control group. Random assignment avoids "selection bias" issues where individuals or groups may self-select themselves into either the intervention or control group based on pre-existing characteristics such as familiarity with computers. Random assignment can occur at the individual level (i.e., each person may or may not receive the intervention) or at the group level (i.e., different groups may or may not receive an intervention). If there is a concern that members of a group will talk to each other about an intervention, then it is best to randomly assign by the group instead of by the individual (because you will not get a clear picture of how the intervention worked if those in the control group were exposed to the intervention through friends or colleagues). Typically, each subject or group is given a number from 1 on up and then a random numbers table (found in the back of any basic statistics text) is consulted to place subjects in either the intervention or control group. An arbitrary decision is made before hand about which numbers in the random numbers table will be the control group and which numbers will be the intervention group (e.g., the odd entries will be the intervention group and the even entries will be the control group). Alternatively, one can simply place each person or group's name on a piece of paper, throw the names into a hat, and designate the first 20 draws as the intervention and the next 20 draws as the control group, and so on.

Random assignment is very important because in any evaluation effort there are "threats" to the validity or accuracy of your results. For example, how do you know that your intervention alone caused increased usage of Grateful Med? Perhaps a new promotion by American Online featuring free internet access caused the increase in usage and not your persuasive message. Similarly, groups or people mature or develop at different rates. Without random assignment, you cannot tease out whether or not changes are due to your intervention, or due to something special about a group that chose to be in an intervention (maybe they're fascinated by computers and are more motivated to learn than computer phobics who choose to be in the control group).

If random assignment is not feasible, then one can utilize a "quasi-experimental" design where individuals or groups are matched on various characteristics (like demographics, psychographics, etc.) to other similar individuals or groups and then compare results. However, if variations exist between the groups in a quasi-experiment, it may be because of the intervention (you hope) or it may be because of other unique, idiosyncratic factors between the groups (e.g., one group of individuals live in a university town and have free access to internet while another group of individuals in another town has to pay for internet). There are ways to statistically control for known covariates (or influences on behavior) but it is best to randomly assign groups or individuals to either the intervention or control group.

The size of the intervention and control groups is determined according to "power" estimates. Specifically, you want enough people per group to detect significant differences between the group if there are in fact significant differences. Usually a minimum of 20 per group can provide adequate degree of power, however, it is best to consult power tables when determining how many individuals or groups you need per group.

To analyze your results, you assess the effects of your "independent variable" (the intervention) on your "dependent variables" (outcome measures). Typically, the dependent variables will be measured on your posttest survey and will include things like attitudes, intentions to act a certain way, reports of behaviors, etc. You would assess the average attitude score by group and then compare attitudes between groups for significant differences. Other dependent variables can be assessed without input from the subject. For example, you can tally how many log-ins or how much time individuals or groups spent on the computer. Then, you would average the number of log-ins or the number of minutes spent on the computer by group. Finally, you will compare these numbers according to a t-test (for two groups) or an F-test (for three or more groups). Easy statistical packages are available for calculating differences between groups.

V. Models of Outreach and Outreach Evaluation

Based on the discussion of variables and theoretical approaches in sections II and III, several models of outreach and outreach evaluation can be generated. Before discussing these different types of evaluation, it's important to first discuss typical outcome measures.

Outcome Measures

Anything that will be influenced by the intervention can be thought of as an outcome measure. At a minimum, outcome measures include perceptual variables, nonreactive measures, and behavioral measures. Perceptual variables are assessed directly from the respondent, usually in a survey or interview. Nonreactive measures are gleaned from the environment, usually without direct subject knowledge or awareness (they are non-reactive to the subject). Behavioral measures are the assessment of the actual behaviors of interest. Examples of each are given as follows:

Perceptual Variables - attitudes, beliefs, intentions, perceived severity of threat, perceived susceptibility to threat, perceived response efficacy, perceived self-efficacy, subjective norm, perceived barriers, fear, defensive avoidance, reactance, etc.

Nonreactive Measures -- wear and tear of books (to determine amount of usage, the more wear presumably the more usage), scuff marks on the floor (to determine usage of different sections of a library), wear and tear of computers, number of printer ribbons used (assumes that more usage equals more people printing out information), etc.

Behavioral Measures -- number of people who accessed an internet line, number of pamphlets picked up at a library, number of requests received for materials, amount of time spent on the internet, amount of time spent talking with colleagues about Grateful Med, etc.

These categories and examples do not exhaust all possible outcome measures, but give the most common outcome measures used in health communication.

Evaluation Strategies

According to many experts, the key to outreach intervention and evaluation success is the use of a theory to guide such efforts. Using a theory in evaluation efforts cuts the guesswork and allows one to understand the mechanisms underlying an outreach intervention's success or failure. For example, it is much easier to evaluate a theory-based intervention because the important variables have already been identified and one knows which direction these variables should be going (either high or low, e.g., barriers should be low while benefits should be high).

Prior to any intervention or evaluation effort, it is useful to answer the following three questions:

Questions for Researcher to Answer:

1. What is the threat or negative consequence you are trying to prevent?
2. What is the recommended response to avert the threat or negative consequence (the specific objective of the campaign)?
3. Who is the target audience (be specific)? [Define target population according to demographic variables (e.g., socio-economic status, literacy level, age, employment, residence, primary language, etc.), psychographic variables (e.g., beliefs, values, worldview), and culture (e.g., customs, norms).]

Question #1 is important because the awareness of threats or negative consequences tends to motivate individuals into action. Although not immediately apparent, in most intervention efforts researchers are trying to prevent some type of threat or negative consequence from occurring. For example, the NN/LM is promoting use of their materials because (presumably) they are trying to prevent misdiagnosis and perhaps harm to patients by making sure physicians know the current medical literature. It also is useful to determine what consequences, if any, your target audience perceives from not performing the recommended response. They may see different threats or negative consequences from not performing the recommended response than the researchers do. For example, in a study promoting birth control (the recommended response) to prevent teen pregnancy, the researchers defined the "threat" from not using birth control as "getting pregnant and having a baby." However, a focus group prior to the development of any messages revealed that most of these inner-city teens did not view having a baby as a negative consequence; in fact, they viewed it as a plus! Therefore, it is critical that some pre-intervention research is done prior to the development of any outreach effort, because the outreach developers' perceptions may be very different than the target audience's perceptions.

Question #2 should represent the specific objective of your campaign and should be as quantifiable as possible. For example, "increasing the use of Grateful Med" is a measurable objective, but "increasing (a) the amount of time and (b) the number of times health care professionals use Grateful Med each month" is a better, more specific objective. Question #3 allows one to develop a profile of your audience so that the message can be targeted toward this specific audience. Answers to question #3 are especially important in any outreach effort targeted toward minority or other unique audiences. In any advertisement or promotional materials used in an outreach effort, it is useful to use characters just like the one developed for the audience profile.

Three types of evaluation will be discussed here: formative evaluation, process evaluation, and summative evaluation. Following is a brief discussion of each.

Formative Evaluation

Formative evaluation occurs prior to the development of a campaign and seeks to inform the development of an outreach effort. Social marketing techniques are commonly used, where the key characteristics of one's target audience are identified. Some social marketers have begun to examine "psychographics," or personality/lifestyle characteristics of a target population, in order to more fully develop a profile of the target audience. Another common formative evaluation technique is to determine which stage of readiness one's target audience is in, with the idea that a researcher would develop a campaign to move the audience from one stage to the next. Finally, many researchers survey or interview a subsample of their target audience to determine their salient beliefs on the theoretical variables of interest.

For example, if a campaign expert chose to use the Health Belief Model as the guiding theory for the intervention, s/he might conduct a formative evaluation where 20 members of the target audience were surveyed for their salient beliefs on (a) their susceptibility to the defined threat and (b) their perceived severity of that threat, as well as (c ) the perceived benefits and (d) barriers of performing the recommended response. (Remember, it's also important to find out if target audience members see the main negative consequence or threat the same as you defined it.)

Following are examples of questions to be asked in a formative evaluation for each of the important variables identified. Note that in a formative evaluation, the questions tend to be open-ended because researchers want to find out beliefs and opinions without imposing any kind of preconceptions on audience's beliefs.

---

ATTITUDES/BELIEFS TOWARD THREAT AND RECOMMENDED RESPONSE

1. What are your beliefs about [[PERFORMING RECOMMENDED RESPONSE]]? (attitude toward recommended response)
2. What negative consequences, if any, come from not [[PERFORMING RECOMMENDED RESPONSE]]? (determines audience perceptions of the threat)
3. What is the best way to prevent experiencing the negative consequences just identified? (determines audience perceptions of the "best" recommended response)

STAGES OF CHANGE (determines which stage of readiness the audience is in)

Choose the statement that best represents your thoughts and actions:

1. Yes No I have yet to think about using Grateful Med. (precontemplation)
2. Yes No I have thought about using Grateful Med but have not taken any steps to use it yet. (contemplation)
3. Yes No I have not yet used Grateful Med but have taken steps so that I will be able to use it soon (e.g., hooked up to internet, signed up for training, sent away for information). (preparation - never used)
4. Yes No I have used Grateful Med. (action)
5. Yes No I regularly use Grateful Med. (maintenance)
6. Yes No I have used Grateful Med before but currently do not use it. (relapse -> go to either preparation or contemplation stage)

---

---

THEORETICAL VARIABLES (open-ended to determine salient beliefs)

1. How serious is [[NEGATIVE CONSEQUENCE/THREAT]]? Please explain. (perceived severity of the threat)
2. How likely is it that you will experience the [[NEGATIVE CONSEQUENCES/THREAT]] from not [[PERFORMING THE RECOMMENDED RESPONSE]]? Please explain. (perceived susceptibility to the threat)
3. [[RECOMMENDED RESPONSE]] will keep me from experiencing the [[NEGATIVE CONSEQUENCES/THREAT]]. Why or why not? (perceived response efficacy)
4. I am easily able to do the [[RECOMMENDED RESPONSE]] to prevent my experiencing the [[NEGATIVE CONSEQUENCES/THREAT]]. Why or why not? (perceived self-efficacy)
5. What keeps you from performing the [[RECOMMENDED RESPONSE]]? Please explain. (perceived barriers)
6. What benefits do you see from performing the [[RECOMMENDED RESPONSE]]? Please explain. (perceived benefits)
7. Please list the people or groups who have an important influence on your [[USING RECOMMENDED RESPONSE]]? (salient referents for subjective norm)
8. Please rank order which way you prefer to learn about [[RECOMMENDED RESPONSE]].

   ___ television ___ magazine ___ letter from professional organization

   ___ radio ___ colleague ___ my boss [[AND SO ON)

   ___ other:_______________________________________________________________

   (preferred channel)
   
   

9. Who would you believe most regarding a message on the [[RECOMMENDED RESPONSE]]?

   ___ health care professional ___ library expert ___ Surgeon General ___ the AMA

   ___ my boss ___ my colleagues [[AND SO ON…]] (preferred source)

---

After answers to these questions have been solicited in either a survey or interview, then one can categorize the responses in a chart (as follows). (Focus groups are not good to use to gather salient beliefs because people's perceptions will be colored by what other members of the group say. It's best to have members of the target audience answer these questions individually.)

This chart will facilitate the development of the intervention because it identifies the important beliefs to focus on. Your theory identifies which way to emphasize the beliefs (i.e., either strengthen or weaken).

BELIEFS TO CHANGE	BELIEFS TO REINFORCE	BELIEFS TO INTRODUCE	AUDIENCE PROFILE
Grateful Med is hard to use It takes too long to find the info I need on Grateful Med I don't have enough time to use Grateful Med etc.... (FILL IN MORE)	I need up-to-date info to effectively treat my patients Grateful Med contains the info I need etc.... (FILL IN MORE)	Grateful Med has new services that help one to search quickly and efficiently for info Within 10 minutes you can be up & running on Grateful Med because of a new, easy-to-use tutorial etc.... (FILL IN MORE)	high literacy level over age 50 tend to be computer illiterate; uses typewriters extremely stressed, overwhelmed by daily commitments lives in rural area; few colleagues use internet etc.... (FILL IN MORE)

Now outreach messages can be developed. Based on the theories reviewed in this paper, following is a summary of how to generate an effective persuasive message:

1. Develop persuasive arguments that make the audience feel at-risk and susceptible to a significant and serious threat. Check beliefs chart to see if audience already perceives threat in this manner or not. Develop messages that refute, reinforce, or introduce beliefs.
2. Develop persuasive arguments that make the audience believe the recommended response effectively, easily, and feasibly averts the threat. This section should make the audience believe they are capable and easily able to perform the recommended response. Any perceived barriers should be fully addressed and refuted.
3. Visual or audio elements of the message (non-content aspects of the message) should "fit" the target audience's cultural values, demographic characteristics, specific needs, and source/channel preferences.
4. Initial campaign messages should be evaluated by members of the target audience (in focus groups) to make sure that they produce the beliefs desired. Make sure beliefs are in the proper direction according to your theory (e.g., high threat/high efficacy messages if the EPPM is used). Modify as needed. Repeat pilot evaluation if necessary.

Process Evaluation

Process evaluation refers to evaluation efforts occurring during an intervention. They can be thought of as "spot-checks" on the effectiveness of an intervention. If one is conducting a theoretically-based evaluation, then the important variables have already been identified. Then, in a process evaluation, one can conduct a brief survey (for example) and assess the levels of these different variables in order to determine whether or not the intervention is operating effectively. For example, say that the Extended Parallel Process Model was used to develop the intervention and evaluation. In a process evaluation, researchers would measure perceptions of severity, susceptibility, response efficacy, and self-efficacy to determine whether the intervention was promoting danger control actions (i.e., adoption of the recommended response) or fear control actions (i.e., defensive avoidance, reactance against the recommended response). If the results of a survey indicated high threat and low efficacy, then according to the theory, one would know that the intervention was failing, because it was promoting fear control responses. However, if the results of the survey indicated high threat and high efficacy, then one could be fairly confident that the intervention was producing the actions desired (danger control responses). Following are examples of questions for each of these constructs:

PERCEIVED THREAT

Perceived Susceptibility
1. I am at-risk for falling behind current medical knowledge.

1 Strongly Disagree

2

3

4

5

6

7 Strongly Agree

Perceived Severity
2. It is dangerous to fall behind current medical knowledge.

1 Strongly Disagree

2

3

4

5

6

7 Strongly Agree

PERCEIVED EFFICACY

Perceived Response Efficacy
3. Using Grateful Med prevents me from falling behind current medical knowledge.

1 Strongly Disagree

2

3

4

5

6

7 Strongly Agree

Perceived Self-Efficacy
4. I am easily able to use Grateful Med to avoid falling behind current medical knowledge.

1 Strongly Disagree

2

3

4

5

6

7 Strongly Agree

Suppose that one week into a month-long intervention, one decided to conduct a process evaluation to make sure the intervention was having the intended effect. Also, suppose that the EPPM was used to theoretically guide the intervention and evaluation. If the average scores of one's intervention group on the above four measures was #1 = 5.6, #2 = 6.1, #3 = 6.9, #4 = 6.2, then one can see that the intervention is promoting high levels of threat (i.e., 5.6 and above) and extremely high levels of efficacy (i.e., 6.2 and above). With these scores one could be confident that the intervention was working well because according to the guiding theory, high threat/high efficacy interventions promote adoption of the recommended response. On the other hand, suppose the average scores on the four measures as #1 = 6.2, #2 = 6.7, #3 = 2.1, #4 = 3.0. These scores indicate that the intervention is promoting very high threat perceptions and low efficacy perceptions. According to the guiding theory, an intervention producing these type of responses would fail because it would be promoting fear control responses (such as defensive avoidance and reactance) resulting in the failure of the intervention to produce behavioral changes. As these examples show, process evaluation is extremely useful in ensuring that the intervention is having the desired effect.

Summative Evaluation

The techniques for summative evaluation were discussed in the prior section (Section IV). First, one has to determine a specific goal or behavioral response that one's campaign is promoting. Are you promoting increased use of library services? Increased use of the internet? The specific objective must be clearly and concisely defined. Second, one has to determine what the appropriate outcome measures are. Are you interested in promoting attitude changes or are you only interested in behavioral changes? The most common outcome measures are (a) knowledge - the degree to which participants correctly answer questions, (b) attitudes - the degree to which participants positively or negatively evaluate the recommended response, (c ) intentions - the degree to which participants intend to do the recommended response, and (d) behaviors - the degree to which the recommended response was actually done.

It is critical that one has a clear understanding of how to measure these constructs. Further, double-barreled questions (i.e., questions that ask two instead of one questions) should be avoided. For example, "Using Grateful Med is easy and fun" - Strongly disagree to Strongly agree - is a double-barreled question because it assesses (1) if Grateful Med is easy and (2) if Grateful Med is fun. What happens if the respondent thinks Grateful Med is fun but not easy? S/he is not able to accurately answer the question. Many excellent texts exist on writing questions (in particular, see Dillman, 1978). The key is to be as specific as possible with regard to the recommended response. Following are examples of items measuring each of the above-identified constructs with "increased use of Grateful Med" designated as the recommended response.

Knowledge

1. To log onto Grateful Med, I need special software.

True

False

2. To use Grateful Med, I must be connected with a university.

True

False

3. Grateful Med is only for health care professionals.

True

False

Attitudes

1. Grateful Med is:

1 Undesirable

2

3

4

5

6

7 Desirable

2. Grateful Med is:

1 Not Beneficial

2

3

4

5

6

7 Beneficial

3. Grateful Med is:

1 Bad

2

3

4

5

6

7 Good

Intentions

1. I intend to use Grateful Med daily.

1 Strongly Disagree

2

3

4

5

6

7 Strongly Agree

2. I plan to use Grateful Med daily.

1 Strongly Disagree

2

3

4

5

6

7 Strongly Agree

Behaviors

1. I use Grateful Med daily.

1 Strongly Disagree

2

3

4

5

6

7 Strongly Agree

2. I use Grateful Med whenever I have a medical question.

1 Strongly Disagree

2

3

4

5

6

7 Strongly Agree

Once the behavioral objective and outcome measures have been determined, then one checks for significant differences between the intervention and control group on these measures to determine how effective the intervention was (as described in Section IV). Typically, scores are averaged across groups and then these average scores are compared between groups. These simple comparisons between groups offer the strongest evidence for the success of one's intervention. If the intervention has not resulted in the types of outcomes desired, one may examine other theoretical variables measured to try and explain why the results came out the way they did. For example, if the EPPM was used and it was discovered in the final analysis that subjects did the opposite of what was advocated, then one can examine perceptions of severity, susceptibility, response efficacy and self-efficacy to try and determine the underlying causes for these results. As one can see, using a theory to guide evaluation efforts is extremely useful because one can examine the reasons for why the intervention did or did not work.

VI. Conclusion

This paper first identified important theoretical variables useful for developing and evaluating outreach interventions, and then described how these variables worked together to affect behavior. Next, this paper outlined typical evaluation methods in the health communication literature and then offered suggestions for different models of evaluation. Several specific examples and questions were given as models of how to develop and evaluate outreach efforts.

NN/LM has asked us to summarize the "central or key points found most relevant to planning and evaluating outreach…" In my opinion, there are three:

1. Use a theory to guide intervention and evaluation efforts. Theories cut the guesswork, increase efficiency, and allow one to isolate both how and why an intervention is or is not working.
2. Conduct formative research to fill in the details for your chosen theory. That is, find out how your target audience defines your theoretical variables. Then, use this information to design your outreach effort.
3. Engage opinion leaders or other community activists as full collaborative partners in the outreach effort. They will be able to identify the most appropriate theory for their community as well as offer guidance on proper definitions for theoretical variables.

A theory gives ready-made guidance, based on empirical laboratory and field research, on how to quickly, efficiently, and effectively develop and evaluate outreach efforts.

Dr. Michael DeBakey, chair of NLM's Board of Regents noted, "information is often the critical link in reaching the correct diagnosis, resulting in lives saved, unnecessary treatment avoided, and hospitilization reduced. Even with all of our modern advances in health care, I still consider good information to be the best medicine" (NLM Press Release, 1997). It is hoped that the theories and practices outlined in this paper will assist in the development of effective outreach intervention and evaluation practices for NN/LM.

VII. References

Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ: Prentice Hall.

Bandura, A. (1989). Perceived self-efficacy in the exercise of control over AIDS infection. In V.M. Mays, G.W. Albee, & S.S. Schneider (Eds.), Primary prevention of AIDS: Psychological approaches (pp. 128 - 141). Newbury Park, CA: Sage.

Brubaker, R.G., & Wickersham, D. (1990). Encouraging the practice of testicular self-examination: A field application of the theory of reasoned action. Health psychology, 9, 154-163.

Crawford, T.J., & Boyer, R. (1985). Salient consequences, cultural values, and childbearing intentions. Journal of applied social psychology, 15, 16-30.

DiClemente, C. C., & Prochaska, J. O. (1985). Processes and stages of change: Coping and competence in smoking behavior change. In S. Shiffman & T. A. Willis (Eds.), Coping and substance abuse (pp. 319-334). San Diego, CA: Academic Press.

Dillman, D.A. (1978). Mail and telephone surveys: The total design method. NY: Wiley & sons.

Eisen, M., Zellman, G.I., & McAllister, A.L. (1985). A health belief model approach to adolescents' fertility control: Some pilot program findings. Health Education Quarterly, 12, 185-210.

Ferguson, K.J., Yesalis, C.E., Pomrehn, P.R., & Kirkpatrick, M.B. (1989). Attitudes, knowledge, and beliefs as predictors of exercise intent and behavior in schoolchildren. Journal of school health, 59, 112-115.

Fishbein, M., & Ajzen, I. (1975). Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research. Reading, MA: Addison-Wesley.

Fishbein, M., & Ajzen, I. (1981). Acceptance, yielding and impact: Cognitive processes in persuasion. In R.E. Petty, T. M. Ostrom, & T. C. Brock (Eds.), Cognitive responses in persuasion (pp. 339-359). Hillsdale, NJ: Erlbaum.

Fishbein, M., Ajzen, I., & McArdle, J. (1980). Changing the behavior of alcoholics: Effects of persuasive communication. In I. Ajzen & M. Fishbein (Eds.), Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ: Prentice-Hall.

Fishbein, M., & Middlestadt, S. E. (1989). Using the theory of reasoned action as a framework for understanding and changing AIDS-related behaviors. In V. M. Mays, G. W. Albee, & S. S. Schneider (Eds.), Primary prevention of AIDS: Psychological approaches (pp. 93-110). Newbury Park, CA: Sage.

Fishbein, M., Middlestadt, S. E., & Hitchcock, P. J. (1991). Using information to change sexually transmitted disease related behaviors: An analysis based on the theory of reasoned action. In J.N. Wasserheit, S.O. Aral, and K. K. Holmes (Eds.), Research issues in human behavior change and sexually transmitted diseases in the AIDS era (pp. 243-257). Washington DC: American Society for Microbiology.

Griffin, R. J., Neuwirth, K., & Dunwoody, S. (1995). Using the theory of reasoned action to examine the health impact of risk messages. Communication Yearbook 18, 201-228.

Grimley, D. M., Riley, G. E., Bellis, J. M., Prochaska, J. O. (1993). Assessing the stages of change and decision-making for contraceptive use for the prevention of pregnancy, sexually transmitted diseases, and aquired immunodeficiency syndrome. Health Education Quarterly, 20, 455-470.

Janis, I.L. (1967). Effects of fear arousal on attitude change: Recent developments in theory and experimental research. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 3, pp. 166-225). New York: Academic Press.

Janz, N. K., & Becker, M. H. (1984). The health belief model: A decade later. Health Education Quarterly, 11, 1-47.

Kline, K. (1995). Applying Witte's Extended Parallel Process Model to pamphlets urging women to engage in BSE: Where are the efficacy messages? Paper presented at the annual meeting of the Speech Communication Association, San Antonio, Texas.

Kotler, P. (1984). Social marketing of health behavior. In L. W. Frederikson, L. J. Solomon, & K. A. Brehony (Eds.), Marketing health behavior (pp. 23-39). New York, NY: Plenum Press.

Kotler, P., & Roberto, E. (1989). Social marketing. New York, NY: Free Press.

Lapinski, M.K., & Witte, K. (in press). Health education campaigns. In B.K. Duffy & L.D. Jackson (Eds.), Handbook of Health Communication Research.

Lefebvre, R. C., & Flora, J. A. (1988). Social marketing and public health intervention. Health Education Quarterly, 15, 299-315.

Leventhal, H. (1970). Findings and theory in the study of fear communications. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 5, pp. 119-186). New York: Academic Press.

Maibach, E. W. & Cotton, D. (1995) Moving people to behavior change: A staged social cognitive approach to message design. In E. W. Maibach & R. L. Parrott (Eds.), Designing health messages: Approaches from communication theory and public health practice (pp. 41-64). Newbury Park, CA: Sage.

Meyer, G., & Dearing, J. W. (1996). Respecifying the social marketing model for unique populations. Social Marketing Quarterly, Winter, 1996.

Meyer, G., & Witte, K. (1995). Intrapersonal Health Communication. In J. Aitken & L.J. Shedletsky (Eds.), Intrapersonal Communication Processes (pp. 364-375). Plymouth, MI: Midnight Oil & Speech Communication Association.

Norman, N.M., & Tedeschi, J.T. (1989). Self-presentation, reasoned action, and adolescents' decisions to smoke cigarettes. Journal of applied social psychology, 19, 543-558.

Prochaska, J. O., DiClemente, C. C., Norcross, J. C. (1992). In search of how people change. American Psychologist, 47, 1102-1114.

Rogers, R.W. (1975). A Protection Motivation Theory of fear appeals and attitude change. Journal of Psychology, 91, 93-114.

Rogers, R.W. (1983). Cognitive and physiological processes in fear appeals and attitude change: A revised theory of Protection Motivation. In Cacioppo, J., & Petty, R. (Eds.), Social Psychophysiology (pp. 153-176). New York: Guilford.

Rosenstock, I.M. (1974). Historical origins of the health belief model. Health Education Monographs,2, 328-335.

Stephenson, M.T. (1993). A subliminal manipulation of the extended parallel process model. Unpublished master's thesis, Texas A&M University.

Vanlandingham, M. J., Suprasert, S., Grandjean, N., Sittitrai, W. (1995). Two views of risky sexual practices among northern Thai males: The health belief model and the theory of reasoned action. Journal of Health and Social Behavior, 36, 195-212.

Wallack, L. (1989). Mass communication and health promotion: A critical perspective. In R. E. Rice & C. K. Atkin (Eds.), Pulic Communication Campaigns (2nd. Ed., pp. 353-367). Newbury Park, CA: Sage.

Wallingford, K.T., Ruffin, A.B., Ginter, K.A., Spann, M.L., Johnson, F.E., Dutcher, G.A., Mehnert, R., Nash, D.L., Bridgers, J.W., Lyon, B.J., Siegel, E.R., National Library of Medicine, Roderer, N.K., & Yale University (1996). Outreach activities of the National Library of Medicine: A five-year review. Bulletin of the Medical Library Association, 84(2), special supplement.

Weinstein, N.D. (1988). The precaution adoption process. Health psychology, 7, 355-386.

Witte, K. (1992a). Putting the fear back into fear appeals: The extended parallel process model. Communication Monographs, 59, 329-349.

Witte, K. (1992b). Preventing AIDS through persuasive communications: A framework for constructing effective, culturally-specific, preventive health messages. International and Intercultural Communication Annual, 16, 67-86.

Witte, K. (1994). Generating effective risk messages: How scary should your risk communication be? Communication Yearbook, 18, 229-254.

Witte, K. (in press-a). Fear as Motivator, Fear as Inhibitor: Using the Extended Parallel Process Model to Explain Fear Appeal Successes and Failures. In P. A. Andersen and L. K. Guerrero (Eds.), Communication and Emotion: Theory, Research, and Applications. Academic Press.

Witte, K. (in press-b). Preventing teen pregnancy through persuasive communications: Realities, myths, and the hard-fact truths. Manuscript under review.

Witte, K., Berkowitz, J., McKeon, J., Cameron, K., Lapinski, M.K., & Liu, W.Y. (1996, November). Radon Awareness and Reduction Campaigns for African-Americans: A Theoretically-Based Formative and Summative Evaluation. Paper presented at the annual meeting of the Speech Communication Association, San Diego, California.

Witte, K., Peterson, T.R., Vallabhan, S., Stephenson, M.T., Plugge, C.D., Givens, V.K., Todd, J.D., Becktold, M.G., Hyde, M.K., Jarrett, R. (1993). Preventing tractor-related injuries and deaths in rural populations: Using a Persuasive Health Message (PHM) framework in formative evaluation research. International Quarterly of Community Health Education, 13, 219-251.

Witte, K., Stokols, D., Ituarte, P., & Schneider, M. (1993). Testing the health belief model in a field study to promote bicycle safety helmets. Communication Research, 20, 564-586.

Wunsch, J.M. (1996). Perceived threat of cancer and perceived efficacy of fruit and vegetable intake in reducing cancer risk by participants in a WIC program. Manuscript under review.

---

Footnotes

1 Portions of this section are drawn from Lapinski and Witte's (in press) chapter on health education and Meyer and Witte's chapter on intrapersonal health communication processes. See this chapters for a full explanation of the theories outlined here.

---

Evaluation Project, Index of Contents