Words: 10017
6 FROM INTENTIONS TO ACTIONS [From: I. Ajzen (1988), Attitudes, personality, & Behavior. Chicago: Dorsey Press] It*s a long step from saying to doing. Cervantes In the previous chapter we began to unravel the mystery surrounding prediction and explanation of specific action tendencies by turning our attention to behavioral dispositions that correspond precisely to the particular action tendency of interest.
Based on this principle of compatibility, the present chapter introduces a conceptual framework for the prediction of specific action tendencies, a framework that deals with a limited set of dispositional antecedents assumed to guide specific action tendencies, with the origins of these dispositions, and with the relations among them. Incorporated into this conceptual framework are the two behavior-specific dispositions discussed in Chapter — perceived behavioral control and attitude toward the behavior — as well as a few additional concepts required for a more complete account of the determinants of specific action tendencies.
The case of willful behavior Many behaviors in everyday life, which are often the behaviors of greatest interest to personality and social psychologists, can be thought of as being largely under volitional control. That is to say, people can easily perform these behaviors if they are so inclined, or refrain from performing them if they decide against it. In Western countries most people can, if they so desire, vote in political elections, watch the evening news on television, buy toothpaste at a drugstore, pray at a nearby church or synagogue, or donate blood to their local hospitals.
If they wish, they may also decide against engaging in any of these activities. The important point about willful behaviors of this kind is that their occurrence is a direct result of deliberate attempts made by an individual. The process involved can be described as follows. In accordance with deliberations to be described below,, a person forms an intention to engage in a certain behavior. Intentions are assumed to capture the motivational factors that have an impact on a behavior; they are indications of how hard people are willing to try, of how much of an effort they are planning to exert, in order to perform the behavior.
These intentions remain behavioral dispositions until, at the appropriate time and opportunity, an attempt is made to translate the intention into action. Assuming that the behavior is in fact under volitional control, the attempt will produce the desired act. This implies that the disposition most closely linked to a specific action tendency is the intention to perform the action under consideration. In other words, when dealing with volitional behavior people can be expected to do what they intend to do. Expressions of behavioral intention should thus permit a highly accurate prediction of corresponding volitional action.
Predicting behavior from intention The literature contains many examples of intentions that are highly correlated with volitional behavior. Table 6. 1 shows a few representative findings. It can be seen that intentions have been found to predict a variety of corresponding action tendencies, ranging from simple strategy choices in laboratory games to actions of appreciable personal or social significance, such as having an abortion, smoking marijuana, and choosing among candidates in an election. It is worth noting that the intentions assessed in these ATTITUDES, PERSONALITY, AND BEHAVIOR studies were highly compatible with the behaviors in terms of the target, action, context, and time elements. Thus, in the study reported by King (1975), the behavior of interest was whether or not college students would attend church services in the course of a 2-week vacation. This behavior could be predicted with a high degree of accuracy by asking the students, prior to the recess, how likely it was that they would attend church services at least every 2 weeks. Available evidence also supports the idea that intentions are close antecedents of overt actions.
If intentions are indeed the immediate determinants of volitional behavior then they should correlate more strongly with the behavior than do other kinds of antecedent factors. Consistent with this argument, the predictive validity of intentions is typically found to be significantly greater than that of attitudes toward the behavior. Consider, for example, the study by Manstead et at. (1983) on the prediction of breastfeeding versus bottle-feeding of newborn infants. As we saw in Chapter 5, mothers* attitudes toward these alternative feeding practices had a correlation of 0. 7 with the feeding method they actually employed. By way of comparison, inspection of Table 6. 1 shows that the intention—behavior correlation in this study was 0. 82. Very similar results were obtained with respect to cooperation in Prisoner*s Dilemma games (Ajzen, 1971; Ajzen and Fishbein, 1970). In Chapter 5, the correlations between attitudes toward choosing the cooperative alternative and actual game behavior were reported to have ranged from 0. 63 to 0. 70. When predicted from intentions, correlations with game behavior were found to be in the 0. 82– 0. 85 range.
Another example is contained in a study by Ajzen et at. (1982). The use of marijuana by college students served as one of the behavioral criteria in this study. The students evaluated “my smoking marijuana in the next 3 or 4 weeks” on a set of semantic differential scales and also indicated, on a 7point scale, the likelihood that they would perform this behavior. About 4 weeks later they were contacted by telephone and asked to indicate whether or not they had smoked marijuana during the time that had passed. In Table 6. 1 it can be seen that this self-report of marijuana use correlated 0. 2. with intentions; its correlation with attitude toward smoking marijuana was, at 0. 53 significantly lower. FROM INTENTIONS TO ACTIONS 3 Stability of intentions Intentions are thus closely linked to volitional actions and can predict them with a high degree of accuracy. This is not to say, however, that a measure of intention will always correlate strongly with the corresponding behavior. Clearly, intentions can change over time; the longer the time interval, the greater the likelihood that unforeseen events will produce changes in intentions.
A measure of intention obtained before the changes took place cannot be expected to predict behavior accurately. It follows that accuracy of prediction will usually decline with the amount of time that intervenes between measurement of intention and observation of behavior. Imagine, for example, a woman who intends to vote for the Democratic candidate in a race for the United States Senate. After her intention is assessed, she learns — by watching a television interview with the candidate a few days before the election — that he opposes abortion and equal rights for women.
As a result, she “changes her mind,” decides to vote for the Republican candidate instead, and actually does so in the election. Her actual voting choice corresponds to her most recent intention, but it could not have been predicted from the measure of intention obtained at the earlier point in time. Several studies have demonstrated the disruptive effects of unforeseen events. For instance, SongerNocks (1976a, 1976b) assessed intentions to choose the cooperative alternative at the beginning of a 20trial, two-person experimental game.
One-half of the pairs of players were given feedback after each trial which informed them about the choices made by their partners and of the pay-offs to each player. The other pairs were given no such information. Feedback concerning the partner*s competitive or cooperative behavior may, of course, influence a player*s own intentions regarding future moves in the game. Consistent with this argument, Songer-Nocks reported that providing feedback significantly reduced the accuracy with which initial intentions predicted actual game behavior.
More indirect evidence regarding the disruptive effects of unanticipated events is available from studies that have varied the amount of time between the assessment of intentions and observation of behavior. Since the likelihood of unforeseen events will tend to increase as time passes, we would expect to find stronger intention—behavior correlations with short rather than long periods of delay. Fishbein and Coombs (1974) reported findings in support of this expectation. In this study, intentions to vote for Goldwater in the 1964 United States presidential election correlated o. o with self-reported voting choice when the intention was measured 1 month prior to the election and 0. 89 when it was measured during the week preceding the election. Sejwacz et al. (1980) also obtained support for the disruptive potential of temporal delay in a study of weight loss. A sample of college women indicated their intentions to perform eight weight-reducing behaviors (avoid snacking between meals, participate in sports on a regular basis, etc. ) at the beginning of a 2-month period and again 1 month later.
Correlations were computed between initial intentions and reported behavior over the 2-month period, and between subsequent intentions and reported behavior during the final month. As expected, intention—behavior correlations were stronger for the 1-month period than for the 2-month period. For example, the correlation between intention to avoid long periods of inactivity and performance of this behavior (as recorded by the women in weekly logs) was higher when the time period was 1 month (r = 0. 72) than when it was 2 months (r = 0. 47). Considering all eight behaviors, the average correlation increased from o. 1 for the 2-month period to 0. 67 for the 1-month period. Explaining volitional behavior: a theory of reasoned action The finding that intentions often predict behavior quite accurately does not in itself provide much information about the reasons for the behavior. Beyond confirming that the behavior in question is under volitional control, it is not very illuminating to discover that people do what they intend to do. Since we are interested in understanding human behavior, not merely in predicting it, we must try to identify the determinants of behavioral intentions.
Ajzen and Fishbein*s (1980; Fishbein and Ajzen, 1975) theory of reasoned action, mentioned in Chapter z, is designed to accomplish precisely this goal; that is, the theory is concerned with the causal antecedents of volitional behavior. As its name implies, the theory of reasoned action is based on the assumption that human beings usually behave in a sensible manner; that they take account of available information and implicitly or explicitly consider the implications of their actions. Consistent with its focus on volitional behavior, and ATTITUDES, PERSONALITY, AND BEHAVIOR 4 n line with the findings reported earlier, the theory postulates that a person’s intention to perform (or not to perform) a behavior is the immediate determinant of that action. Barring unforeseen events, people are expected to act in accordance with their intentions. Attitudes and subjective norms According to the theory of reasoned action, intentions are a function of two basic determinants, one personal in nature and the other reflecting social influence. The personal factor is the individual*s attitude toward the behavior, first encountered in Chapter and again earlier in this chapter.
Unlike general attitudes toward institutions, people, or objects that have traditionally been studied by social psychologists, this attitude is the individual*s positive or negative evaluation of performing the particular behavior of interest. The second determinant of intention is the person*s perception of social pressure to perform or not to perform the behavior under consideration. Since it deals with perceived normative prescriptions, this factor is termed subjective norm. Generally speaking, people intend to perform a behavior when they evaluate it positively and when they believe that important others think they should perform it.
The theory assumes that the relative importance of attitude toward the behavior and subjective norm depends in part on the intention under investigation. For some intentions attitudinal considerations are more important than normative considerations, while for other intentions normative considerations predominate. Frequently, both factors are important determinants of the intention. In addition, the relative weights of the attitudinal and normative factors may vary from one person to another. Figure 6. is a graphic representation of the theory of reasoned action as described up to this point. Many studies have provided strong support for the hypothesized links between intention as the dependent variable and attitude toward the behavior and subjective norm as the independent variables. Most studies have used multiple linear regression procedures to estimate, in terms of a multiple correlation (R), the simultaneous predictive power of attitudes and subjective norms, as well as the relative contributions of the two predictors in terms of standardized regression coefficients.
Table 6. 2 shows the results obtained in the studies discussed earlier (see Table 6. 1) as well as a few additional examples. It can be seen that, with respect to a variety of different intentions, consideration of attitudes and subjective norms permitted highly accurate prediction. The multiple correlations in the studies listed ranged from 0. 73 to 0. 89. The relative importance of the two predictors is revealed by inspecting FROM INTENTIONS TO ACTIONS 5 columns 3 and 4.
In all cases, attitudes and subjective norms both made significant contributions to the prediction of intentions, although in eight of the ten studies, the relative contribution of attitudes exceeded that of sub jective norms. Perhaps not surprisingly, however, women*s decisions to have an abortion, and a couple*s decision to have another child, were more strongly affected by perceived social pressure than by personal attitudes. ATTITUDES, PERSONALITY, AND BEHAVIOR 6 For many practical purposes this level of explanation may be sufficient.
We can to some extent account for the intentions people form by examining their attitudes toward the behavior, their subjective norms, and the relative importance of these two factors. However, for a more complete understanding of intentions it is necessary to explore why people hold certain attitudes and subjective norms. Antecedents of attitudes toward a behavior. In Chapter 2 we discussed, in general terms, the formation of attitudes within the framework of the theory of reasoned action. There we showed how evaluations of any object follow reasonably from the beliefs we hold about the object.
We can now apply these ideas to the formation of attitudes toward a behavior. According to the theory of reasoned action, attitude toward a behavior is determined by salient beliefs about that behavior, termed behavioral beliefs. Each behavioral belief links the behavior to a certain outcome, or to some other attribute such as the cost incurred by performing the behavior. For example, a person may believe that “going on a low sodium diet” (the behavior) “reduces blood pressure,” “leads to a change in life style,” “severely restricts the range of approved foods,” and so forth (outcomes).
The attitude toward the behavior is determined by the person*s evaluation of the outcomes associated with the behavior and by the strength of these associations. As we see in Chapter 2 the evaluation of each salient outcome contributes to the attitude in proportion to the person*s subjective probability that the behavior will produce the outcome in question. By multiplying belief strength and outcome evaluation, and summing the resulting products, we obtain an estimate of the attitude toward the behavior, an estimate based on the person*s salient beliefs about the behavior.
This expectancy-value model is described symbolically in Equation 6. 1, where AB stands for attitude toward behavior B; bi is the belief (subjective probability) that performing behavior B will lead to outcome i; ei is the evaluation of outcome i; and the sum is over the n salient beliefs. It can be seen that, generally speaking, a person who believes that performing a given behavior will lead to mostly positive outcomes will hold a favorable attitude toward performing the behavior, whereas a person who believes that performing the behavior will lead to mostly negative outcomes will hold an unfavorable attitude.
AB = 3biei (6. 1) Several of the studies cited earlier have reported data that confirm the expectancy-value model of attitude described in Equation 6. 1. For example, King (1975) assessed behavioral beliefs concerning the advantages and disadvantages of attending church services at least every 2 weeks as well as evaluations of these outcomes. Responses were used to compute an estimate of attitude toward attending church services in accordance with Equation 6. 1.
In addition, King used an evaluative semantic differential to obtain a relatively direct measure of the same attitude. The correlation between the direct evaluation of the behavior and the belief-based measure was found to be 0. 69. High correlations between direct and belief-based measures of attitude have also been reported with respect to such behaviors as voting choice in a United States presidential election (r = 0. 79), using birth control pills (r = 0. 79), and choice of a career orientation (r = 0. 81) (see Ajzen and Fishbein, 1980). Antecedents of subjective norms.
Subjective norms, the second major determinant of intentions in the theory of reasoned action, are also assumed to be a function of beliefs, but beliefs of a different kind, namely the person*s beliefs that specific individuals or groups approve or disapprove of performing the behavior. Serving as a point of reference to guide behavior, these individuals and groups are known as referents. For many behaviors, the important referents include a person*s parents, spouse, close friends, coworkers, and, depending on the behavior involved, perhaps such experts as physicians or tax accountants.
The beliefs that underlie subjective norms are termed normative beliefs. Generally speaking, people who believe that most referents with whom they are motivated to comply think they should perform the behavior will perceive social pressure to do so. Conversely, people who believe that most referents with whom they are motivated to comply would disapprove of their performing the behavior will have a subjective norm that puts pressure on them to avoid performing the behavior. The relation between normative beliefs and subjective norm is expressed symbolically in Equation 6. . Here, SN is the subjective norm; bj is the normative belief concerning referent j; mj is the person*s motivation FROM INTENTIONS TO ACTIONS 7 to comply with referent j; and n is the number of salient normative beliefs. SN % 3bjmj (6. 2) Subjective norms can be assessed in a relatively direct manner by asking respondents to judge how likely it is that most people who are important to them would approve of their performing a given behavior. Such direct measures have been compared with belief-based estimates of subjective norms, computed in accordance with Equation 6. . Correlations between the two types of measures are generally quite high, ranging from 0. 60 to 0. 80 (see Ajzen and Fishbein, 1980). The discussion up to this point shows how volitional behavior can be explained in terms of a limited number of concepts. Through a series of intervening steps the theory of reasoned action traces the causes of behavior to the person*s salient beliefs. Each successive step in this sequence from behavior to beliefs provides a more comprehensive account of the factors that determine the behavior.
At the initial level behavior is assumed to be determined by intention. At the next level these intentions are themselves explained in terms of attitudes toward the behavior and subjective norms. The third level accounts for attitudes and subjective norms in terms of beliefs about the consequences of performing the behavior and about the normative expectations of relevant referents. In the final analysis, then, a person*s behavior is explained by considering his or her beliefs.
Since people*s beliefs represent the information (be it correct or incorrect) they have about themselves and about the world around them, it follows that their behavior is ultimately determined by this information. 1 The informational foundation of behavior A concrete example may help clarify the role of beliefs in determining the performance of a specific behavior. Manstead et a!. (1983) compared the beliefs of mothers who breast-fed their babies with mothers who used the bottle-feeding method.
Based on prior research in the field, the investigators selected the six reasons women cite most frequently for breast-feeding their babies and the six reasons they cite most frequently for bottle-feeding their babies. With respect to each of these 12 salient behavioral beliefs, women about to give birth were asked to provide two measures: their subjective probabilities that a given feeding method is associated with the cited consequence, and their evaluations of that consequence. The following are examples for each feeding method.
Behavioral beliefs Breast-feeding protects a baby against infection likely :___:___:___:___:___:___:___: unlikely Bottle-feeding provides incomplete nourishment for a baby likely :___:___:___:___:___:___:___: unlikely Outcome evaluations Using a feeding method that protects a baby against infection is very important :___:___:___:___:___:___:___: completely unto me important to me Using a feeding method that provides complete nourishment for my baby is very important :___:___:___:___:___:___:___: completely unto me important to me ATTITUDES, PERSONALITY, AND BEHAVIOR 8 Table 6. shows the average likelihood rating (7 = likely, 1 = unlikely) provided by mothers who breast-fed their babies and mothers who bottle-fed their babies. Statistical significance between the two groups is indicated. As can be seen, the two groups of mothers differed significantly on all six of the behavioral beliefs about breast feeding. Examination of these differences reveals some of the reasons for choosing one or the other feeding method. Although all women tended to agree that breast-feeding establishes a close bond between mother and baby, the women who held this belief more strongly were more likely to choose the breast-feeding method.
In a similar vein, the choice of breast-feeding increased with the perceived likelihood that this method is good for the mother*s figure, provides the best nourishment for a baby, and protects a baby against infection. On the other hand, the more a woman believed that breast-feeding is embarrassing for the mother or limits her social life, the less likely she was to use this method. With respect to the bottle-feeding method, the two groups of mothers differed significantly on only three of the six behavioral beliefs.
An examination of the significant differences shows that perceived outcomes of bottle-feeding which best explained the choice of this method were the beliefs that it is a very convenient method, that it enables the father to be involved in feeding, and that it is a trouble-free feeding method. It is possible, in a similar fashion, to compare the outcome evaluations of mothers who breast-fed their babies with those of mothers who chose the bottle-feeding method. Such a comparison provides additional information about the reasons for choosing one method over the other. Table 6. presents the average outcome evaluations for the two groups 1 = completely unimportant, 7 = very important). Examining the six evaluations that distinguished significantly between the two groups, it can be seen that mothers tended to choose the breast-feeding method if, in comparison to mothers who chose the bottle-feeding method, they judged as relatively important the following outcomes: having a good figure, establishing a close bond with their babies, providing complete nourishment for their babies, and FROM INTENTIONS TO ACTIONS 9 protecting their babies against infection.
In addition, these mothers also rated as relatively unimportant the outcomes of feeling embarrassed, allowing the baby*s father to be involved in the feeding, and being able to see exactly how much milk baby has had. The study by Manstead et al. (1983) also reported interesting data concerning the effects of normative beliefs on the choice of breast-versus bottle-feeding. The salient normative referents identified in this context were the baby*s father, the mother*s own mother, her closest female friend, and her medical adviser (usually a gynecologist).
With respect to each referent, normative beliefs about breastfeeding and about bottle-feeding were assessed, as was motivation to comply with each referent. The following scales illustrate the procedures used. Normative beliefs The baby*s father thinks that I definitely should :___:___:___:___:___:___:___: definitely should breast-feed not breast-feed Motivation to comply In general, how much do you care what the baby*s father thinks you should do? Do not care at all :___:___:___:___:___:___:___: Care very much Table 6. 5 shows the average normative beliefs for the two groups of mothers.
The differences between mothers who breast-fed their babies and mothers who used the bottle are statistically significant for each normative belief. Inspection of the normative beliefs for mothers who used the breast-feeding method reveals that, in their opinions, important referents strongly preferred this method over the alternative bottle-feeding method. In contrast, women who believed that their referents had no strong preferences for either method were more likely to feed their babies by means of a bottle. ATTITUDES, PERSONALITY, AND BEHAVIOR 10
Finally, the mothers* average motivations to comply with each of the four salient referent individuals are presented in Table 6. 6. Both groups of mothers were highly motivated to comply with the baby*s father, and they had moderately strong motivations to comply with their own mothers and closest female friends. The only significant difference emerged with respect to the women*s medical advisers. Mothers who eventually decided to breast-feed their babies were more highly motivated to comply with their medical advisers than were mothers who eventually decided to use The bottle.
This is consistent with the finding that the former mothers perceived their medical advisers to be strong advocates of the breastfeeding method. (see Table 6. 5). To summarize briefly, research on the theory of reasoned action describes how people tend to proceed on a course of action in quite a deliberate manner. The initial considerations deal with the likely consequences of performing a certain behavior and expectations of important referent individuals or groups.
Depending on the evaluation of the behavior*s likely consequences and motivation to comply with referent sources, attitudes and subjective norms emerge that guide the formation of behavioral FROM INTENTIONS TO ACTIONS 11 intentions. Barring unforeseen events that might change the intentions, and contingent on the behavior being under volitional control, the intentions are carried out under appropriate circumstances. The case of incomplete volitional control The theory of reasoned action was developed explicitly to deal with purely volitional behaviors.
In this context it has proved quite successful. Complications are encountered, however, when we try to apply the theory to behaviors that are not fully under volitional control. A well-known example is that many smokers intend to quit but, when they try, fail to attain their goal. In the theory of reasoned action, intentions are the prime motivating force and they mediate the effects of other factors, i. e. of attitude toward the behavior and of subjective norm.
The stronger are people*s intentions to engage in a behavior or to achieve their behavioral goals, the more successful they are expected to be. However, the degree of success will depend not only on one*s desire or intention, but also on such partly nonmotivational factors as availability of requisite opportunities and resources. To the extent that people have the required opportunities and resources, and intend to perform the behavior, they should succeed in doing so. At first glance, the problem of behavioral control may appear to apply to a limited range of actions only.
Closer scrutiny reveals, however, that even very mundane activities, which can usually be executed (or not executed) at will, are sometimes subject to the influence of factors beyond one*s control. Such a simple behavior as driving to the supermarket may be thwarted by mechanical trouble with the car. Control over behavior can thus best be viewed as a continuum. On one extreme are behaviors that encounter few if any problems of control. A good case in point is voting choice: once the voter has entered the voting booth, selection among the candidates can be done at will.
At the other extreme are events, such as sneezing or lowering one*s blood pressure, over which we have very little or no control. Most behaviors, of course, fall somewhere in between these extremes. People usually encounter few problems of control when trying to attend lectures or read a book, but problems of control are more readily apparent when they try to overcome such powerful habits as smoking or drinking or when they set their sights on such difficult-to-attain goals as becoming a movie star.
Viewed in this light it becomes clear that, strictly speaking, most intended behaviors are best considered goals whose attainment is subject to some degree of uncertainty. We can thus speak of behavior-goal units, and of intentions as plans of action in pursuit of behavioral goals (Ajzen, 1985). Control factors Many investigators have in recent years turned their attention to the question of volitional control (e. g. KuhI, 1985; Liska, 1984; Sarver, 1983; Triandis, 1977).
On the following pages we review some of the factors that can influence the degree of control a person has over a given behavior. Internal factors Various factors internal to an individual can influence successful performance of an intended action. Some of these factors are readily modified by training and experience while others are more resistant to change. Information, skills, and abilities. A person who intends to perform a behavior may, upon trying to do so, discover that he or she lacks the needed information, skills, or abilities. Everyday life is replete with examples.
We may intend to convert another person to our own political views, to help a boy with his mathematics, or to repair a malfunctioning record player, but fail in our attempts because we lack the required verbal and social skills, knowledge of mathematics, or mechanical aptitudes. To be sure, with experience we tend to acquire some appreciation of our abilities; yet new situations arise frequently, and failure to achieve our goals due to the lack of requisite skills is the order of the day. The lack of ability in an unusual sense is illustrated in a study by
Vinokur-Kaplan (1978) who assessed a couple*s intention to have another child the following year. When interviewed 12 months ATTITUDES, PERSONALITY, AND BEHAVIOR 12 later, actually having given birth to a child correlated o. 55 with intentions, a correlation which, although significant, is lower than the intention—behavior correlation observed in many other contexts. Having another child is, of course, only partially under volitional control, since fecundity, miscarriage, and other factors also mediate attainment of this goal.
Finally, forgetting is an interesting type of internal factor frequently cited as a reason for failure to carry out an intention (see Kuhl, 1985). A planned appointment or a deadline intended to be met can “slip a person*s mind”. In their study on blood donation, Pomazal and Jaccard (1976) interviewed people who had indicated an intention to donate but whose names did not appear on the official donor list. Among the reasons frequently mentioned was that they had simply forgotten all about it. Emotions and compulsions.
Skills, abilities, and information may present problems of behavioral control, but it is usually assumed that, at least in principle, these problems can be overcome. In contrast, some types of behavior are subject to forces that seem to be largely beyond our control. People sometimes appear unable to cease thinking or dreaming about certain events, to stop stuttering, or to hold a tic in check. These compulsive behaviors are performed despite intentions and concerted efforts to the contrary. Emotional behaviors seem to share some of the same characteristics.
Individuals are often not held responsible for behaviors performed under stress or in the presence of strong emotions. We usually attribute little behavioral control to a person who is “overcome by emotion. ” Violent acts and poor performance are expected under such conditions, and there seems to be little we can do about it. In sum, as we move beyond purely volitional acts, various internal factors may influence the successful performance of intended behavior. It may be fairly easy to gain control over some of these factors, as when we acquire the information r skills needed to perform a behavior. Other factors, such as intense emotions, stress or compulsions, are more difficult to neutralize. External factors Also impinging on a person*s control over attainment of behavioral goals are situational or environmental factors external to the individual. These factors determine the extent to which circumstances facilitate or interfere with the performance of the behavior. Opportunity. It takes little imagination to appreciate the importance of incidental factors or opportunities for the successful execution of an intended action.
An intention to see a play cannot be carried through if tickets are sold out on a particular night or if the person is involved in a serious accident on the way to the theater. The Pomazal and Jaccard (1976) study of blood donation again provides relevant examples. When students who had failed to carry out their intentions to donate blood were interviewed, they often mentioned that such unforeseen obligations or events as exams, job interviews, and coming down with a cold had prevented them from participating in the blood drive.
Given the presence of many disruptive factors, it is hardly surprising that the correlation between intention and behavior was found to be of only moderate magnitude (r = 0. 52). In some instances, students came to give blood but were turned away because of overcrowding. When these individuals were considered to have performed the behavior, the intention—behavior correlation increased to 0. 59. At first glance, lack of opportunity may appear equivalent to occurrence of unanticipated events that bring about changes in intentions, as discussed previously.
While it is true that in the absence of appropriate opportunities people may come to change their intentions, there is an important difference between the two cases. When new information becomes available after intentions have been stated, the new information may affect salient beliefs about the behavior and thus lead to changes in attitudes, subjective norms, and intentions; at the end of this process the person is no longer interested in carrying out the original intention. By way of contrast, lack of opportunity disrupts an attempted behavior.
Here, the person tries to carry out the intention but fails because circumstances prevent performance of the behavior. Although the immediate intention will be affected, the basic underlying determinants need not have changed. Consider again the intention to see a particular play. Reading a negative review or being told by a FROM INTENTIONS TO ACTIONS 13 friend that the play is not worth seeing may influence the person*s beliefs such as to produce a more negative attitude toward the intended behavior and perhaps also a more negative subjective norm.
As a result the person may no longer intend to see the play on the night in question or on any other night, unless and until other events again cause a change of mind. Contrast this with the person who intends to see the play, drives to the theater, but is told that there are no more tickets available. The environmental obstacle to performance of the behavior will force a change of plan; but it need not change the person*s attitude or subjective norm with respect to seeing the play. Instead, it may merely cause the person to try again on a different night.
Note also that lack of opportunity poses a problem only when the performance of a behavior on a single occasion is to be predicted. Behavioral tendencies across occasions are relatively unaffected because appropriate opportunities are likely to be present on at least some occasions. Dependence on others. Whenever the performance of a behavior depends on the actions of other people, there exists the potential for incomplete control over behavioral goals. A good example of behavioral interdependence is the case of cooperation.
One can cooperate with another person only if that person is also willing to cooperate. Experimental studies of cooperation and competition in laboratory games have provided ample evidence for this interdependence. For example, Ajzen and Fishbein (1970) reported correlations of 0. 92. and 0. 89 between cooperative strategy choices of the players in two Prisoner*s Dilemma games. These high correlations suggest that a person*s tendency to make cooperative choices depends on reciprocation by the other player.
As is true of time and opportunity, the inability to behave in accordance with intention because of dependence on others need not affect the underlying motivation. Often an individual who encounters difficulties related to interpersonal dependence may be able to perform the desired behavior in cooperation with a different partner. Sometimes, however, this may not be a viable course of action. A wife*s adamant refusal to have more children will usually cause the husband eventually to abandon his plan to enlarge the family, rather than shift his effort to a different partner.
In short, lack of opportunity and dependence on others often lead only to temporary changes in intentions. When circumstances prevent the performance of a behavior, the person may wait for a better opportunity and, when another person fails to cooperate, a more compliant partner may be sought. However, when repeated efforts to perform the behavior result in failure, more fundamental changes in intentions can be expected. A theory of planned behavior The above discussion makes clear that many factors can disrupt the intention—behavior relation.
Although volitional control is more likely to present a problem for some behaviors than for others, personal deficiencies and external obstacles can interfere with the performance of any behavior. Collectively, these factors represent people*s actual control or lack of control over the behavior. [See also the discussions of "facilitating factors” by Triandis "the context of opportunity” by Sarver (1983), "resources” by Liska (1984) and "action control” by KuhI (1985). ] Given the problem*s ubiquity, a behavioral intention can best be interpreted as an intention to try performing a certain behavior.
A father*s plan to take his children fishing next weekend is best viewed as an intention to try to make time for this activity, to prepare the required equipment, secure a fishing license, and so forth. The successful performance of the intended behavior is contingent on the person*s control over the many factors that may prevent it. Of course, the conscious realization that we can only try to perform a given behavior will arise primarily when questions of control over the behavior are salient. Thus, people say that they will try to quit smoking or lose weight, but that they intend to go to church on Sunday.
Nevertheless, even the intention to attend Sunday worship services must be viewed as an intention to try performing this behavior since factors beyond the individual*s control can prevent its successful execution. A recent attempt to provide a conceptual framework that addresses the problem of incomplete volitional control is Ajzen*s theory of planned behavior (Ajzen, 1985; Ajzen and Madden, 1986; Schifter and Ajzen, 1985). This conceptual framework is an extension of the theory of reasoned action. As in the original model, a central factor in the theory of planned behavior is an individual*s intention to perform the ehavior of interest. In contrast to the original version, however, the theory of planned behavior postulates three, rather than two, conceptually independent determinants of intentions. The first ATTITUDES, PERSONALITY, AND BEHAVIOR 14 two — attitude toward the behavior and subjective norm — are the same as before. The third and novel antecedent of intention is the degree of perceived behavioral control. This factor, discussed in Chapter 5, refers to the perceived ease or difficulty of performing the behavior and it is assumed to reflect past experience as well as anticipated impediments and obstacles.
As a general rule, the more favorable the attitude and subjective norm with respect to a behavior, and the greater the perceived behavioral control, the stronger should be the individual*s intention to perform the behavior under consideration. Note that the theory of planned behavior does not deal directly with the amount of control a person actually has in a given situation; instead, it considers the possible effects of perceived behavioral control on achievement of behavioral goals.
Whereas intentions reflect primarily an individual*s willingness to try enacting a given behavior, perceived control is likely to take into account some of the realistic constraints that may exist. To the extent that perceptions of behavioral control correspond reasonably well to actual control, they should provide useful information over and above expressed intentions. A structural model of the theory of planned behavior is shown in Figure 6. 2. Figure 6. 2 shows two important features of the theory of planned behavior.
First, the theory assumes that perceived behavioral control has motivational implications for intentions. People who believe that they have neither the resources nor the opportunities to perform a certain behavior are unlikely to form strong behavioral intentions to engage in it even if they hold favorable attitudes toward the behavior and believe that important others would approve of their performing the behavior. We thus expect an association between perceived behavioral control and intention that is not mediated by attitude and subjective norm. In Figure 6. this expectation is represented by the arrow linking perceived behavioral control to intention. The second feature of interest is the possibility of a direct link between perceived behavioral control and behavior. As noted earlier, in many instances, the performance of a behavior depends not only on motivation to do so but also on adequate control over the behavior in question. It follows that perceived behavioral control can help predict goal attainment independent of behavioral intention to the extent that it reflects actual control with some degree of accuracy.
In other words, perceived behavioral control can FROM INTENTIONS TO ACTIONS 15 influence behavior indirectly, via intentions, and it can also be used to predict behavior directly because it may be considered a partial substitute for a measure of actual control. Of course, in some situations perceived behavioral control is not particularly realistic. This is likely to be the case when the individual has little information about the behavior, when requirements or available resources have changed, or when new and unfamiliar elements have entered into the situation.
Under those conditions a measure of perceived behavioral control may add little to the accuracy of behavioral prediction. The broken arrow in Figure 6. 2 indicates that the link between perceived behavioral control and behavior is expected to emerge only when there is some agreement between perceptions of control and the person*s actual control over the behavior. Like the theory of reasoned action, the theory of planned behavior deals with the antecedents of attitudes, subjective norms and perceived behavioral control, antecedents which in the final analysis determine intentions and actions.
Recall that, at the most basic level of explanation, behavior is assumed to be a function of salient information, or beliefs, relevant to the behavior. Three kinds of beliefs are distinguished: behavioral beliefs which are assumed to influence attitudes toward the behavior, normative beliefs which constitute the underlying determinants of subjective norms, and control beliefs which provide the basis for perceptions of behavioral control. Earlier we discussed the effects of behavioral beliefs on attitude toward the behavior, and the effects of normative beliefs on subjective norms.
In a similar fashion, control beliefs are assumed to provide the basis for perceived behavioral control. According to the theory of planned behavior, among the beliefs that ultimately determine intention and action is a set that deals with the presence or absence of requisite resources and opportunities. These beliefs may be based in part on past experience with the behavior, but they will usually also be influenced by second-hand information about the behavior, by observing the experiences of acquaintances and friends, and by other factors that increase or reduce the perceived difficulty of performing the behavior in question.
The more resources and opportunities individuals think they possess, and the fewer obstacles or impediments they anticipate, the greater should be their perceived control over the behavior. As with behavioral and normative beliefs, it is possible to separate out these control beliefs and treat them as partially independent determinants of behavior. Just as beliefs concerning consequences of a behavior are viewed as determining attitudes, and normative beliefs are viewed as determining subjective norms, so beliefs about resources and opportunities may be viewed as underlying perceived behavioral control.
Consider the case of regular attendance at class lectures in college. As part of a pilot study, Ajzen and Madden (1986) elicited salient beliefs about factors that might help or interfere with the performance of this behavior. The following ten factors were mentioned with the greatest frequency: conflicting events, sickness, family obligations, employment, being tired or listless, transportation problems, upsetting personal problems, oversleeping or forgetting, heavy load imposed by other classes, and failure to prepare class assignments.
In the experiment itself, control beliefs were assessed by asking respondents to rate, on 7-point scales, the likelihood that each of the ten factors would occur. The sum over these responses provided a belief-based measure of perceived behavioral control. In addition, Ajzen and Madden also asked students to judge more directly how much control they thought they had over regular class attendance. Specifically, the following three questions were posed at separate points in the questionnaire. 1. How much control do you have over whether you do or do not attend this class every session? omplete :___:___:___:___:___:___:___: very little control control 2. For me to attend every session of this class is easy :___:___:___:___:___:___:___: difficult 3. If I wanted to, I could easily attend this class every session likely :___:___:___:___:___:___:___: unlikely ATTITUDES, PERSONALITY, AND BEHAVIOR 16 A direct measure of perceived behavioral control was obtained by summing over responses to these three items. A correlation of 0. 54 confirmed the hypothesized link between this direct measure and the belief-based measure of perceived behavioral control described above.
The theory of planned behavior is a general model in which the theory of reasoned action represents a special case. As noted earlier, the original model was designed to deal with behaviors over which people have a high degree of volitional control and it assumed that most behaviors of interest in the domains of personality and social psychology fall into the volitional category (see Ajzen and Fishbein, 1980). The theory of planned behavior, however, explicitly recognizes the possibility that many behaviors may not be under complete control, and the concept of perceived behavioral control is added to handle behaviors of this kind.
However, when behavioral control approaches its maximum and issues of control are not among an individual*s important considerations, then the theory of planned behavior reduces to the theory of reasoned action. In those instances, neither intentions nor actions will be affected appreciably by beliefs about behavioral control and the only remaining dispositions of interest are attitude toward the behavior and subjective norm. Prediction of intentions Earlier in this chapter we reviewed some of the empirical evidence in support of the theory of reasoned action.
Clearly, this evidence is also supportive of those aspects of the theory of planned behavior that overlap with the theory of reasoned action. The theory of planned behavior, however, goes beyond the theory of reasoned action in that it introduces the concept of perceived behavioral control and proposes a direct causal effect of perceived control on intention, an effect not mediated by attitude or subjective norm. Evidence for this aspect of the theory is examined in the present section.
Schifter and Ajzen (1985) applied the theory of planned behavior to the prediction of weight loss intentions, and actual weight reduction, among female college students. Attitudes toward losing weight during the following 6 weeks were assessed by means of several evaluative semantic differential scales. To measure subjective norms, participants were asked to indicate, again on 7-point scales, whether people who were important to them thought they should lose weight over the next 6 weeks, and whether those people would approve or disapprove of their losing weight.
As a measure of perceived behavioral control, participants indicated, on a scale from 0 to 100, the likelihood that if they tried they would manage to reduce their weight over the next 6 weeks and their estimates that an attempt on their part to lose weight would be successful. The final measure of interest for present purposes dealt with intentions to lose weight over the following 6 weeks. Each woman indicated, on several 7-point scales, her intention to try to reduce weight and the intensity of her decision. The first row in Table 6. shows the correlations of intentions to lose weight with attitudes, subjective norms and perceived behavioral control. It can be seen that all three predictors correlated significantly with intention. A hierarchical regression analysis was performed on intentions to lose weight in which attitudes and subjective norms were entered on the first step, and perceived behavioral control on the second. 2 This analysis reveals the effect of perceived behavioral control on intentions after the effects of attitude and subjective norm have been statistically removed.
Thus, the hierarchical regression analysis tests the idea that perceived behavioral control contributes to intentions over and above the influence of the two factors contained in the original theory of reasoned action. The results of the analysis confirmed the importance of perceived behavioral control as a third determinant of intentions to lose weight. Although the multiple correlation of intentions with attitudes and subjective norms alone was quite high (r = 0. 65), it increased significantly — to 0. 72 — with the addition of perceived behavioral control.
All three independent variables had significant regression coefficients, indicating that each made an independent contribution to the prediction of weight loss intentions. The importance of perceived control over a behavioral goal has also been demonstrated in the context of scholastic performance (Ajzen and Madden, 1986). In one part of the investigation, undergraduate college students enrolled in upper division courses expressed, at the beginning of the semester, their intentions to attempt getting an “A” grade in the course, as well as their attitudes, subjective norms and perceived control over this behavioral goal.
Attitudes toward getting an “A,” subjective norms, and perceived behavioral control were each assessed by means of several direct questions and on the basis of FROM INTENTIONS TO ACTIONS 17 a set of relevant salient beliefs. The measure of intention was a set of three direct questions dealing with intentions to try to get an “A. ” Before turning to the prediction of intentions it is worth noting that the study provided support for the hypothesized relation between direct and belief-based measures of attitude, subjective norm, and perceived behavioral control. The correlations between the two types of measures ranged from 0. 7 to 0. 57 (p ; 0. 01). The second row in Table 6. 7 shows the correlations of intentions to get an “A” with the direct measures of attitudes, subjective norms, and perceived behavioral control. A hierarchical regression analysis revealed that attitudes and perceived behavioral control each had a significant effect on intention. On the basis of attitude toward the behavior and subjective norm alone, the multiple correlation with intention was 0. 48 (P ; 0. 01). The introduction of perceived behavioral control on the second step of the regression analysis raised the multiple correlation significantly to the level of 0. 5. Losing weight and getting an “A” in a course are both behavioral goals over which people clearly have only limited volitional control. In addition to the desire to lose weight, people have to be familiar with an appropriate diet or exercise regimen, and they have to be capable of adhering to the diet or exercise program in the face of distractions and temptations. Similarly, getting an “A” in a course depends not only on strong motivation but also on intellectual ability, availability of sufficient time for study, resisting temptations to engage in activities more attractive than studying, and so on.
It is not surprising, therefore, that perceived behavioral control is found to influence intentions to pursue or not to pursue these behavioral goals. There is also evidence, however, that even when problems of volitional control are much less apparent, people*s intentions are affected by their control beliefs. In the investigation by Ajzen and Madden (1986) records were kept of students* attendance of eight class lectures following administration of a questionnaire. The questionnaire contained measures of intention to attend classes regularly, attitudes toward this behavior, subjective norms, and perceived behavioral control.
The latter three variables were again assessed by means of direct questions and, more indirectly, on the basis of sets of salient beliefs. The correlations between the belief indices and the direct measures were significant, ranging from 0. 47 to 0. 54 (p ; 0. 01). As to the prediction of intentions from the direct measures, in the third row of Table 6. 7 it can be seen that perceived behavioral control correlated significantly with intentions, as did attitudes and subjective norms. A hierarchical regression analysis showed that on the basis of attitudes and subjective norms alone, the multiple correlation with intentions was (P ; 0. 1). However, the addition of perceived behavioral control on the second step improved the prediction significantly, resulting in a multiple correlation of 0. 68. The findings presented up to this point indicate that the original theory of reasoned action, with its implication that perceived behavioral control can influence intention only indirectly via attitude or subjective norms, is not sufficient. The addition of perceived behavioral control as a direct determinant of intention improved prediction of several behaviors, and this effect was independent of attitudes and subjective norms. ATTITUDES, PERSONALITY, AND BEHAVIOR 18 Prediction of goal attainment The theory of planned behavior also suggests the possibility that perceived behavioral control may be related to behavior not only indirectly, via its effect on intentions, but also directly, over and above the effect due to intentions. This possibility was explored in the studies described above in which attempts were made to predict attainment of three behavioral goals: attending lectures on a regular basis, getting an “A” in a course, and losing weight. Table 6. shows the correlations between intentions and perceived behavioral control on the one hand and attainment of the behavioral goal on the other. With respect to regular class attendance, both intentions and perceived control correlated significantly with actual behavior. A hierarchical regression analysis, however, showed that the addition of perceived behavioral control did not improve prediction of behavior significantly. This was expected since class attendance is a behavior over which students have considerable volitional control.
The addition of a (subjective) measure of control thus added little information of value in the prediction of actual behavior. In contrast, losing weight does pose problems of volitional control. As would therefore be expected, the results with respect to attainment of this goal showed the relevance of perceived behavioral control quite dramatically. As can be seen in the second row of Table 6. 8, both intentions and perceived control correlated significantly with goal attainment, but perceived control was the better predictor of the two.
The addition of perceived behavioral control on the second step of a hierarchical regression analysis improved prediction significantly, raising the multiple correlation with goal attainment from 0. 25 to 0. 44. Perhaps the most interesting results, however, emerged in the study on getting an “A” in a course. The questionnaire assessing the different constructs of the theory of planned behavior was administered twice, once at the beginning of the semester and again toward the end. Perception of control over getting an “A” should, of course, become more accurate as the end of the semester approaches.
As an addition to intentions, the later measure of perceived behavioral control should therefore contribute to the prediction of course grades more than the earlier measure. The data presented in the last two rows of Table 6. 8 lend support to this hypothesis. Although both measures, intentions and perceived control, gained in predictive accuracy, the more dramatic gain was observed with respect to the latter. Moreover, hierarchical regression analysis showed that whereas with the data obtained early in the semester, only intentions had a significant effect on behavior, with the later data, both ntentions and perceived behavioral control had significant regression coefficients. Thus, the addition of perceived behavioral control had no effect on the accuracy of behavioral prediction for the data obtained early in the semester, but it raised the correlation significantly from 0. 39 to 0. 45 for the data obtained toward the end of the semester. 4 FROM INTENTIONS TO ACTIONS 19 Before concluding this discussion it may be instructive to take a closer look at the way in which the examination of control beliefs can aid our understanding of the factors that determine behavioral performance.
We shall use academic achievement as an example. This analysis parallels our earlier discussion of behavioral and normative beliefs as determinants of a mother*s choice to breast-feed or bottle-feed her baby. In a pilot study conducted prior to the main experiment, Ajzen and Madden (1986) asked college students to list any factors that could help them get an “A” in a course and any factors that might make it difficult for them to get an “A. Four potential facilitating factors mentioned frequently were stimulating subject matter, clear and organized lectures, possession of required skills and background, and availability of help from the instructor. Four frequently mentioned factors whose presence would hamper attaining a good grade were taking other demanding classes, extracurricular activities, arduous text and reading materials, and difficult exams and course requirements.
In the second wave of the main experiment, toward the end of the semester, college students were asked to judge, with respect to each of these eight factors, how much the factor was likely to influence their ability to get an “A” in a particular course they were taking at the time. Table 6. 9 shows the average control beliefs scored in the direction of facilitation (x = factor hinders attaining a good grade, 7 = factor facilitates attaining a good grade) as well as the correlation of each belief with the intention to get an “A” and with actual grades attained. Inspection of the mean control beliefs reveals that the students who took art in the experiment thought they would be helped by the subject matter of the course which was stimulating enough to motivate them, by the lectures which they considered to be sufficiently clear and organized, by their possessing the required skills and background, and by the ready availability of help from the instructor. On the other hand, the students also believed that they would encounter certain obstacles, especially in the form of demands on their time and energy imposed by other classes they were taking and in the form of extracurricular activities. The correlations displayed in Table 6. demonstrate the impact of these different control beliefs on intentions to make an effort to get an “A” in the course and on actual grades attained. Of special importance were perceptions concerning the course*s subject matter, lecture organization, possession of required skills and background, and the nature of the exams and other course requirements. The more that students saw these factors as facilitating their performance in the course, the stronger were their intentions to try for an “A” and the higher were the grades they actually attained. ATTITUDES, PERSONALITY, AND BEHAVIOR 20
In conclusion, the experiments reviewed above have provided some initial support for the theory of planned behavior. The addition of perceived behavioral control to the variables contained in the original theory of reasoned action was found greatly to improve the prediction of behavioral intentions. This finding indicates that perception of behavioral control, like attitude toward the behavior and subjective norm, can have an important impact on a person*s behavioral motivation. The more that attainment of a behavioral goal is viewed as being under volitional control, the stronger is the person*s intention to try.
In addition, perceived behavioral control can also improve the prediction of actual behavior beyond the level obtained on the basis of intentions alone. This is the case, however, only under certain conditions. First, the behavior must at least in part be determined by factors beyond a person*s control. When the behavior is largely under volitional control, intentions alone are found to be sufficient to predict it. Secondly, perceived behavioral control must be fairly realistic, reflecting actual control to a reasonable degree.
This condition was apparently met in the study on weight loss, and it was also fulfilled toward the end of the semester in the study on academic performance. Summary and conclusions This chapter discussed a theoretical framework, the theory of planned behavior, that can help us predict and understand the performance of specific action tendencies. We examined some of the factors that influence deliberate performance of willful actions as well as additional factors that must be taken into account when we are dealing with behaviors or behavioral go