This paper examines the nature and development of expectancy-value theories. The relationship between such theories and communication in the context of group decision-making is explored. The metaphysical and epistemological requirements of such decision-making are discussed. The value of group communication arises from its function as a means of achieving the widest integration of knowledge possible and of ensuring the objectivity of one's conclusions.
Unlike the psychological theory of behaviorism, expectancy-value theories hold that consideration of the individual person is essential in understanding and predicting human actions. These theories fall together as a special subunit of general decision making principles. This paper will examine the recent historical development of this field and some of the more prominent theories in it along with their general assumptions and development. Conceptual and definitional differences and similarities of such key concepts as attitudes, beliefs, values, and expectations as described by the selected theorists will be discussed. The overarching principles which unite these theories will be presented along with relevant criticisms and shortcomings of these theories.
It will also be argued in this paper that expectancy-values need to undergo a conceptual expansion in order adequately to explain human actions. Most such theories have focussed on the psychological aspects of the individual but have dealt little or not at all with the role of interaction with other individuals in influencing values or expectations. It will be argued here that communication can -- but will not necessarily -- play an integral role in changing or confirming individual decisions for action. A model of group decision-making offered by Hirokawa and Scheerhorn (1986) will be used to illustrate the application of expectancy-value theories to communication in a group context. In addition, metatheoretical considerations of metaphysical and epistemological issues will be examined in order to demonstrate why the principles suggested for good quality group decision-making are the correct ones.
To a large extent, a person's behavior is a result of individual wants (values) and considerations of what is true about the world or oneself (beliefs) (Scheibe, 1970). Just as these values and beliefs differ from person to person, the relative importance which people attach to these factors will vary and be instrumental in determining what specific actions will or will not be taken by one individual under any particular set of circumstances. Indeed, the same person may, in essentially identical situations, choose to perform different behaviors.
Expectancy-value theories also hold that people are goal-oriented beings. The behaviors they perform in response to their beliefs and values are undertaken to achieve some end. While modern psychological expectancy-value theories are a relatively recent development, this teleological perspective on human actions has a long philosophical history dating back to Aristotle (Jones, 1952). Whether the means chosen to actualize one's decisions will be adequate to obtain that goal will be determined by objective reality and the physical laws which describe it. Rationality is a prerequisite for the consistent success of one's teleological behaviors, but it is not a prerequisite for the decision making process itself (Rand, 1964).
One of the earliest recognitions that organisms engage in purposive (goal-directed) behavior was in E. C. Tolman's Purposive Behavior in Animals (1932). He viewed organisms as being capable of anticipating behavioral consequences and of adjusting subsequent actions based upon perceived expectancies (means-ends-readinesses) about the causal connections between behavior and environmental effects.
He saw these cognitive factors as intervening variables between stimulus conditions and an animal's responses. Tolman was, however, still a student of behaviorism and viewed these intervening variables not as real but as merely theoretical generalizations designed to stand for empirical relationships he had observed in animal behavior.
The Gestalt psychologist, Kurt Lewin, developed the field theory of decision making (1936). In this theory, people are viewed as acting in a life space which is the sum of all the psychological factors impinging upon a person at any specified moment. This life space is separated into regions -- fields -- which correspond to one's conceptual organization of the world. Each of these fields has associated with it a valence or the positive or negative attractiveness of some object or activity to that person.
Lewin used the concept of subjective probability to describe the probability which a person holds in the face of uncertainty that a behavior will result in the achievement of a particular goal. This measure of the likelihood of success or failure in reaching an objective is a central idea in later expectancy-value theories.
When this subjective probability is multiplied by the valence a field has for a person, the result is a force or weighted valence which influences a person to move away from a negatively valenced field towards one more positively charged. Lewin's field theory described a process of decision making that had considerable influence on later researchers (Feather, 1959).
Von Neumann and Morgenstern's book Theory of Games and Economic Behavior (1944) dealt with risky decision making. Risky decision making occurs when the probability of an event's occurrence can be determined. This objective probability is distinguished from subjective probability (where there is uncertainty as to what an event's probability of occurrence is). An example of the difference between these two concepts would be the difference between knowing the probability of a coin coming up heads or tails and the uncertain probability that the reader will go out for a beer after finishing this paper.
In the theory of games, profits and losses are measured in terms of an individual's utility function or functions, that is, one's relative values. This theory was the first to establish a set of axioms of rational behavior which, if followed, would establish a unique, objective decision strategy for an individual to follow.
However, in this objective view of probability, there is no room for individual judgments. It is assumed that a person knows what the objective chances are for various outcomes, that is, that the expected utility equals the actual utility of an outcome. There is no evidence that people are able to do this except under limited conditions (Edward, 1954).
In 1954, L. E. Savage took the basic theory of games and modified it by substituting subjective probability for objective probability. The subjective view held that "probability measures the confidence that a particular individual has in the truth of a particular proposition" (Savage, p. 3). The magnitudes of these probabilities are, however, constrained by some rules of consistency which are stated axiomatically.
This personalistic, goal-oriented view is characteristic of later expectancy-value theories. Motivating objectives are not static but are subject to change over time. These alterations can be sudden or barely in an individual's awareness. Objectives may be abandoned or a person may be uncertain as to precisely what his or her goals are. As will be discussed later in this paper, this variability in goals over time has important implications for any attempt at persuasion which aims at influencing what those goals are to be.
Behavioral decision theory or subjective expected utility (S.E.U.) theory (Edwards, 1961) is one of the most fully developed of the expectancy-value formulations. The likelihood of an event (which is subject to human influence) occurring (the expectancy variable) is seen as the subjective probability that the outcome will occur if a behavior is undertaken. The value variable (the subjectively determined utility of the goal) is multiplied by the expectancy. The product is the subjective expected utility.
In order to maximize one's outcomes, the individual chooses the S.E.U. with the highest number. Also included in the equation are terms for the probability of the outcome occurring if no action is taken and the utility of not achieving the desired outcome (usually negative). These terms are summed over all relevant outcomes and alternatives (both positive and negative) to arrive at the final S.E.U. which will indicate whether or not a person will decide to perform a given behavior.
Some modifications which have been suggested for the basic S.E.U. theory include a focus on individual differences such as self-esteem (Pelc & Midlarsky, 1977), expectations of personal efficacy (Bandura, Adams, & Beyer, 1977), and locus of control (Bauman & Fisher, 1985).
Situational factors to be considered in the model include locus of attribution and self-focus (Carver & Scheier, 1982), environmental situations which preclude the completion of an action, and the relationship between subjective and objective probabilities (Sutton, 1982). Conceptual or methodological problems include whether S.E.U.s are additive or should be replaced by a weighted averaging rule (Lynch, 1979) and how congruent S.E.U. theories are with probability theories (Mitchell, 1977). Despite such concerns, S.E.U. theories are seen as being good predictors of how people will act.
The expectancy theory (Vroom, 1964) applies an expected-value model to an organizational setting. As in Lewin's field theory, Vroom utilized valences (defined here as affective orientations toward particular outcomes) multiplied by the strength of the expectancies of the instrumentality (utility) of those outcomes. This product yields an estimate of the force on a person to perform a particular act. The expectancy theory is most often used to predict such things as job satisfaction, one's occupational choice, the likelihood of staying in a job, and the effort one might expend at work (Behling & Starke, 1973).
Fishbein's (1967) theory of reasoned action or behavioral intentions is another widely accepted and well-developed expectancy-value theory. For Fishbein, behavior is a result of intentions. Intentions, in turn, are functions of one's attitudes to the behavior in question and one's subjective norms. Attitudes result from one's beliefs (expectations) that a behavior will lead to a particular outcome (its subjective probability) and one's evaluation of that outcome (its subjective utility). The subjective norms are a combination of a person's beliefs of how significant others feel about the normative appropriateness of the anticipated behavior and the individual's decision as to the value of those predicted norms.
Fishbein's theory is "a specialized summation model of information integration focusing on beliefs as the basic building blocks of human cognitive schemata" (Smith, 1982, p. 259). In this theory, in order to effect persuasion, the persuader must focus on the content of the message and work to alter the target's primary beliefs (Fishbein & Ajzen, 1981).
Research has suggested that Fishbein's original theory needs to be expanded to consider the effects of the time interval between intentions to act and the actual behavior, the influence of previous behavior on subsequent behavior, and the sequence of behaviors (Saltzer, 1981). There have been conflicting results as to whether attitudes directly affect intentions to behave as Fishbein predicted (Bagozzi, 1981) or in a more indirect fashion, especially in the case of those people with internal rather than external locus of control (Saltzer, 1981).
On the whole, the importance of the relationship between beliefs, expectations, attitudes, and intentions on behavior and persuasion has been well supported (Ajzen, 1971; Ajzen & Fishbein, 1969, 1977; Anderson & Fishbein, 1965; Fishen & Ajzen, 1974; Fishbein & Raven, 1962).
An extension of Fishbein's basic model of behavioral choices was presented by H. C. Triandis in 1979. While Fishbein started the decision process with a person's primary beliefs as the ultimate determinants of behavior, Triandis expanded this view by asking how those beliefs originated and what factors influenced the expression of those beliefs in action. Intentions are still seen as leading to behavior, but one's habit hierarchies (automatic situation-behavior sequences), relevant arousal (physiological arousal), and the presence of positively or negatively facilitating conditions (for example, environmental characteristics) are viewed as also having direct impacts on behavior.
Personality factors which influence this process result from past behavior, culture, history, and ecology. In congruence with earlier theories, Triandis utilized the concept of perceived consequences (subjective expected utility or outcome evaluations) which are composed of a summation of perceived probabilities that behaviors will have particular consequences (their subjective probabilities) and the perceived value of those consequences (subjective value). Triandis viewed these subjective evaluations as being influenced by feedback from the objective consequences of the actual behavior and the interpretations the individual makes of those results.
Rogers's protection motivation theory (1975) deals with cognitive appraisal processes which mediate attitude change. These processes consist of three parts. The first is the magnitude of the "noxiousness" of a specified event (its appraised severity or utility). Second is the probability of the event's occurrence (the expectancy one has of being exposed to that event). Third is the efficacy of the response (the belief one has in the efficacy of the coping response or its subjective expected utility). The protection motivation theory is viewed by Rogers as a special case of expectancy-value theories. It is concerned primarily with the effects fear has on attitude change (defined as the intent to adopt the recommended action). Some shortcomings of the theory which Rogers acknowledges are its neglect of the duration of the noxious events, the interval between them, and the cost involved in any response to those events.
Three other theories which are, in essence, expectancy-value theories but which are not usually recognized as such are Rotter's social learning theory (1954), Atkinson's motivation theory (1957), and Homan's social exchange theory (1961).
Rotter's (1954) social learning theory is concerned with three major antecedents of behavior. These are expectancies regarding the occurrence of specific outcomes (subjective probabilities), reinforcement values which are positively or negatively associated with outcomes (subjective utilities), and the psychological situation which presents the possible courses of behavior to a person. Taken together, the expectancies and reinforcement values determine a quantity called the behavior potential. The assumption in this theory is that a person will choose the alternative action which has the highest behavior potential. In other words, as in S.E.U. theory, the individual will seek to maximize his utility (which provides the motivation to act).
Atkinson's (1957) motivation theory deals with the effect on risk taking behavior of individual differences in motive strength. Motivation to achieve success (the S.E.U.) is viewed as a product of the subjective probability of success and the incentive value of success. Similarly, motivation to avoid failure is a product of the subjective probability of failure and the negative incentive value of failure. Cognitive and affective components reflecting these products result in behavior. While this theory is designed to deal with decision making in a social context, its correspondence to expectancy-value theories is evident.
This congruence with values and expectancies is also apparent in Homans's (1961) social exchange theory. Here, rewards in a relationship minus costs (similar to subjective probabilities of the success of actions combined with an assessment of the values of those actions) are equal to profits or motivation to remain in a relationship (the S.E.U.) which will lead to certain behaviors on the part of the individual.
Thibaut and Kelley (1959) and Blau (1964) used similar ideas in dealing with social influence and interactions. Thibaut and Kelley saw power as an exchange analysis in which gains for compliance are compared to costs or gains from alternative actions. Blau utilized the rewards a partner gives to another in a love relationship to describe the balance which is sought in an outcome-to-input ratio.
The Health Belief Model (Maiman & Becker, 1974) takes expected-value ideas into the medical realm. The value variable is the perception one has of one's reduced suspectibility to and the severity of a disease. The expectancy variable is the perception of benefits from minus the costs of the recommended behavior or treatment. The product of these two is a disposition to act (the S.E.U. or outcome variable). Research indicates the value for prediction of this version of the Health Belief Model in regard to a disposition to act on recommended health behaviors.
Some of the difficulties with operationalizing expectancy-value theories include dealing with the conversion of multiple values and expectancies in regard to an object or behavior into a single preference; the consistency or lack thereof of preference over time; the transitivity of preferences (if person X prefers A to B and B to C, will X always prefer A to C?); the gap between decisions to act and subsequent behavior (which is a function of individual and environmental factors); and what constitutes maximum utility for different individuals. All of these facts are subjects for future research.
Expectations, Values, and Communication
As noted earlier, expectancy-value theories have been used primarily in psychology to help us understand the behaviors of individuals alone or in a group setting. The concepts presented in these theories, however, have value in describing the role of communication in interpersonal or group settings and, specifically for the purposes of this paper, in dealing with effective or ineffective group decision-making.
Hirokawa and Scheerhorn (1986) present a model of group decision-making which is in implicit agreement with general expectancy-value concepts. This model presents a view of how groups reach decisions, which factors can lead to poor decisions, and how individual group members affect decision quality.
In general, this model holds that the decisions which a group makes are affected by the group's judgments as to 1) the nature of the situation in question, 2) the objectives and values which need to be achieved by the decision, and 3) the positive and negative outcomes which might result from other possible decisions.
The first factor in this model corresponds to the expectations variable which is found in expectancy-value theories. Expectations regarding behavioral outcomes result from judgments made about the likelihood that a particular behavior will result in the achievement of a particular outcome. As in Hirokawa and Scheerhorn's model, such judgments must take into account the specific situation under consideration in order to establish the causal relationship between behaviors and outcomes. Identical behaviors may result in significantly different outcomes if actualized in different contexts.
The second factor in this model corresponds rather obviously with the value component in expectancy-value theories. Values are what one acts to gain and/or keep (Rand, 1964). Given the teleological nature of human behavior, desired goals, purposes, "end-states," or objectives are seen by those desiring such things as being values. The very concept of "value" requires an answer to the question: of value to whom and for what? Values "presuppose an entity capable of acting to achieve a goal in the face of an alternative" (Rand, 1964, p. 15).
Judgments as to the valences of those alternatives (factor three in Hirokawa and Scheerhorn's model) determine whether one will be motivated to approach or avoid those alternatives as well as the strength of that motivation. In expectancy-value theories, such valencing and weighting are essential in determining how values and expectations will combine to influence behavioral choices.
Hirokawa and Scheerhorn present their model as a way to explain how poor group decisions occur. They see such faulty choices as resulting primarily from five factors. These factors are 1) incorrect judgments about the situation, 2) incorrect choices of goals, 3) incorrect judgments of the positive and negative aspects of alternatives, 4) incorrect choice of an information base, and 5) incorrect reasoning from an information base.
According to Hirokawa and Scheerhorn (1986), such problems "are facilitated or brought about principally through the communicative efforts of individual members" (p. 76). Essentially, this occurs as a result of social influence or persuasion. (Compare this to an expectancy-value theory such as Fishbein's  which deals primarily with instances of persuasion). Conversely, the prevention of decision errors can also be traced to social influence processes.
Metaphysical and Epistemological Foundations
As is evident from the descriptions of expectancy-value theories given above, most such theories are concerned chiefly with subjectively determined expectations (probabilities) and values. While recognition of the importance of such factors is crucial, this focus on the subjective unnecessarily limits the scope of expectancy-value theories. It is the contention of this paper that the function of communication in group decision-making (in terms of improving decision quality) is to bring such subjective judgments into as close agreement as possible with objective probabilities and values. In this view, group communication offers one (perhaps more efficient) means of achieving this goal.
The acceptance of such a position regarding the possibility of reaching "correct" decisions and the function of communication as a means to that end requires the acceptance of certain metaphysical and epistemological principles. The first is that the concept of objectivity applies not only to the physical world but to the realm of human activities, as well. If the reality of some human action does not have a definite existence apart from human judgments about that action, then there is no way to apply the concept of "error" to such judgments; errors or mistakes presuppose something from which a judgment deviates. (This is consonant with a correspondence view of truth which states that truth is an identification of the facts of reality [Oldroyd, 1986; Tomlin, 1963].)
While objectivity is usually thought of in this metaphysical sense (that is, things are what they are), it is also an epistemological principle, that is, it is not equivalent to "the observable" (Branden, 1967). In this sense, it refers to the way in which we discover whether or not an error has occurred. If one accepts the premise of objective existence, then the function of human consciousness becomes one of identification of that which exists (Rand, 1967). Misidentification of some part of reality may occur due to false premises, incomplete knowledge, or a fault in one's logical argument (Kelley, 1988; Rand, 1967) (cf Hirokawa and Scheerhorn, 1986). But, as noted above, the existence of error presupposes the idea that there does, in fact, exist something independent of our identification which serves as a standard for judging when our indentification is incorrect.
Since people are neither omniscient nor infallible (if they were, there would be no need for reason or logic), all knowledge is contextual (but not relative) in nature. To correctly identify something means that there is no contradiction with any available empirical evidence or logical extensions of that evidence. Because of this fact, it is important in any attempt to reach a knowledge claim that there be no dropping of context or evasion of any relevant knowledge which is available for consideration. In making a decision as to what qualifies as knowledge and what of that knowledge can be said to be known with certainty, it is necessary to consider the widest context available at any given time. Knowledge is thus contextually absolute and not relative and mutable when shifting among concurrent perspectives (Rand, 1967).
Objectivity as an epistemological principle, then, is achieved by examining the widest context of knowledge possible and applying reason (both deductively and inductively) to that knowledge in accordance with the laws of logic as identified by Aristotle (i.e., the Law of Identity [A is A]; the Law of Noncontradiction [not both A and not-A]; and the Law of Excluded Middle [either A or not-A]).
(Such a procedure should also be followed in making judgments about the objective status of the values one holds. For example, one might subjectively hold heroin to be a value, but a consideration of the widest context of knowledge available, e.g., medical facts, would prove that it is not an objective value for human life since it is destructive of such life. This conclusion follows from the fact that only life makes the concept of "value" possible. As mentioned earlier, "values" have meaning only in reference to a valuer, i.e., a living entity [Rand, 1964]. Anything which destroys the basis for values is, thus, objectively not a value.)
Given these points, it follows that when group communication acts to increase the relevant context of knowledge, when it forces participants to prove that their premises are the correct ones, and when it functions to ensure that an argument is, in fact, consistent and noncontradictory, it will help provide the conditions necessary for consistently effective choices, that is, decisions which reflect objective reality. To the extent that it fails in this role, to that extent it will more likely lead to a decrease in decision quality. Consonant with Hirokawa and Scheerhorn (1986), group communication can operate to change expectations and values in either a positive or negative direction.
"Discussion and debate are values only if they are means to the discovery of the truth" (Kelley, 1990, p. 50). If values cannot be objectively validated, if causal connections cannot be objectively determined, if the truth or falseness of opinions cannot be judged, then communication does not and cannot serve to accomplish any of the goals identified in Hirokawa and Scheerhorn's (1986) model.
Conceptual knowledge is neither automatic nor self-evident but exists in a complex, hierarchical interrelationship among its elements. Integration at the widest level possible is essential to understanding reality (Rand, 1967). "...(D)iscussion and debate provide a useful -- and in some cases practically indispensable -- means of achieving" the values of integration and objectivity (Kelley, 1990, p. 51). Integration of only the essential and relevant can be a difficult process. "Discussion with other people who have integrated things differently, who have attended to different features and patterns in the data, is a way of countering the problem" (Kelley, 1990, p. 52) of "confirmation bias" (the tendency to remember most readily that which supports our beliefs). Examples of the implementation of such principles are our adversary system of law, peer review in academia, and organized debates. The competition of ideas which can -- but will not necessarily -- occur during group decision-making is the best check possible that logical errors or faulty premises will not go uncorrected. Since cognition is an individual activity and knowledge is a relationship between an individual and reality, such errors can be detected and corrected without communication if one has the time and energy to do so. Communication, however, is usually a more efficient means to that end.
If one goal of group decision-making is to be certain of the choice(s) made, then it must be realized that such certainty -- like knowledge itself -- is contextual. The reframing and application of expectancy-value theories in accordance with the ideas outlined above allow for both rational and irrational decisions. The role of communication is to increase the probability that we will obtain the value of those rational actions which correspond to the reality we seek to understand and manipulate. Given the nature of human consciousness, there are no guarantees of success or "magic bullets" which will infallibly work. If we are to succeed, however, it is necessary to be aware of and acknowledge the requirements we should follow in attempting to reach our goals.
Ajzen, I. (1971). Attitudinal vs normative messages: An investigation of the differential effects of persuasive communication on behavior. Sociometry, 34, 263-280.
Ajzen, I., & Fishbein, M. (1977). Attitude behavior relations: A theoretical analysis and review of empirical research. Psychological Bulletin, 84, 888-918.
Ajzen, I., & Fishbein, M. (1969). The prediction of behavioral intentions in a choice situation. Journal of Experimental Social Psychology, 5, 400-416.
Anderson, L. R., & Fishbein, M. (1965). Prediction of attitude from the number, strength, and evaluative aspect of beliefs about the attitude object. Journal of Personality and Social Psychology, 2, 437-443.
Atkinson, J. W. (1957). Motivational determinants of risk-taking behaviors. Psychological Review, 64, 359-372.
Bagozzi, R. P. (1981). Attitudes, intentions, and behavior: A test of some key hypotheses. Journal of Personality and Social Psychology, 141, 607-627.
Bandura, A., Adams, N. E., & Beyer, J. (1977). Cognitive processes mediating behavioral change. Journal of Personality and Social Psychology, 35, 125-139.
Bauman, K. E., & Fisher, L. A. (1985). Subjective expected utility, locus of control, and behavior. Journal of Applied Social Psychology, 15, 606-621.
Behling, O., & Starke, F. A. (1973). The postulates of expectancy theory. Academy of Management Journal, 16, 373-388.
Blau, P. M. (1964). Exchange and power in social life. New York: Wiley.
Branden, N. (1967). The psychology of self-esteem. New York: Bantam.
Carver, C. S., & Scheier, M. F. (1982). Outcome expectancy, locus of attribution for expectancy, and self-directed attention as determinants of evaluations and performances. Journal of Experimental Social Psychology, 18, 184-200.
Edwards, W. (1954). The theory of decision making. Psychological Bulletin, 52, 380-417.
Edwards, W. (1961). Behavioral decision theory. Annual Review of Psychology, 12, 472-498.
Feather, N. T. (1959). Subjective probability and decision making under uncertainty. Psychological Review, 66, 150-164.
Fishbein, M. (1967). Attitude and the prediction of behavior. In M. Fishbein (Ed.), Readings in attitude theory and measurement. New York: Wiley.
Fishbein, M., & Azjen, I. (1974). Attitudes toward objects as predictors of single and multiple behavioral criteria. Psychological Review, 81, 59-74.
Fishbein, M., & Azjen, 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: Lawrence Erlbaum.
Fishbein, M., & Raven, B. H. (1962). The AB scales: An operational definition of belief and attitude. Human Relations, 15, 35-44.
Hirokawa, R. Y., & Scheerhorn, D. R. (1986). Communication in faulty group decision-making. In R. Y. Hirokawa & M. S. Poole (Eds.) Communication and group decision-making (pp. 63-80). Beverly Hills, CA: Sage.
Homans, G. C. (1961). Social behavior: Its elementary forms. New York: Harcourt, Brace, & World.
Jones, W. T. (1952). A history of Western philosophy. New York: Harcourt Brace.
Kelley, D. (1988). The art of reasoning. New York: W. W. Norton.
Kelley, D. (1990). Truth and toleration. Verbank, NY: Institute for Objectivist Studies.
Lewin, K. (1936). Principles of topological psychology. New York: McGraw-Hill.
Lynch, J. G., Jr. (1979). Why additive utility models fail as descriptions of choice behavior. Journal of Experimental Social Psychology, 15, 397-417.
Maiman, L. A., & Becker, M. H. (1974). The health belief model: Origins and correlates in psychological theory. Health Education Monographs, 2, 336-353.
Mitchell, T. R. (1977). Expectancy and expected value: Decision models for organizations. In E. H. Burack & A. R. Negandhi (Eds.), Organization design. (pp. 97-115). Kent, OH: Comparative Administrative Research Institute.
Oldroyd, D. (1986). The arch of knowledge: The history of the philosophy and methodology of science. New York: Methuen.
Pelc, R. E., & Midlarsky, E. (1977). Effects of evaluative feedback on the subjective probability of success. The Journal of Psychology, 96, 191-196.
Rand, A. (1964). The virtue of selfishness. New York: Signet.
Rand, A. (1967). Introduction to Objectivist epistemology. New York: Meridian.
Rogers, R. W. (1975). A protection motivation theory of fear appeals and attitude change. The Journal of Psychology, 91, 93-114.
Rotter, J. B. (1954). Social learning and clinical psychology. New York: Prentice Hall.
Saltzer, E. B. (1981). Cognitive moderators of the relationship between behavioral intentions and behavior. Journal of Personality and Social Psychology, 41, 260-271.
Savage, L. E. (1954). Foundations of statistics. New York: Wiley.
Scheibe, K. E. (1970). Beliefs and values. New York: Holt, Rinehart, & Winston.
Smith, M. J. (1982). Persuasion and human action. Belmont, CA: Wadsworth.
Sutton, S. R. (1982). Fear arousing communications: A critical examination of theory and research. In J. R. Eiser (Ed.), Social psychology and behavioral medicine (pp. 303-337). London: Wiley.
Thibaut, J., & Kelley, H. H. (1959). The social psychology of groups. New York: Wiley.
Tolman, E. C. (1932). Purposive behavior in animals and man. New York: Century.
Tomlin, E. W. F. (1963). The Western philosophers (rev. ed.). New York: Harper & Row.
Triandis, H. C. (1979). Values, attitudes, and interpersonal behavior. In H. E. How, Jr. (Ed.), Nebraska symposium on motivation (Vol. 27, pp. 195-259). Lincoln, NE: University of Nebraska Press.
Von Newmann, J., & Morgenstern, O. (1944). Theory of games and economic behavior. Princeton, N. J.: Princeton University Press.
Vroom, V. H. (1964). Work and motivation. New York: Wiley.