6.1. A MODEL OF HUMAN BEHAVIOR

For our purposes here an act is a short sequence of behaviors that eventually results in some outcome, such as the publication of a paper or the development of a good research design. In other words, we are using the word "act" in a very specific way. Hundreds of these acts are necessary to produce a publication or to develop a product. What we are trying to understand is what makes these small acts more or less probable.

Actions have results that are evaluated, and considered as the outcomes of action that may satisfy individual needs. This motivational model has received some support in the literature (Pritchard and Youngcourt, 2008).

Two variables are important in this case: previous habits and self-instruction. For example, when a person says, "I should look up these references," that is a self-instruction or behavioral intention. Research has shown that behavioral intentions predict behaviors quite well (Triandis, 1977, 1980).

The model thus states that the probability of an act is dependent on two kinds of variables: habits and behavioral intentions. However, even when people have the proper habits and intentions to carry out a particular act, they may fail to do so because external conditions may not be favorable. We utilize the concept of facilitating conditions in order to explain the phenomenon that even though the individual may have all that is required, the act may not occur. Reasons beyond the intentions of the individual may not allow it. For example, there may be a lack of proper equipment or there may be distractions in the environment. Facilitating conditions can be measured both with data obtained "outside the individual" (e.g., by asking objective observers, who know the conditions of work well, to judge if the act can occur) and with data obtained from "inside the individual," by measuring the individual's sense of "self-efficacy." This can be measured by asking the individual, "Can you do that?" A scale can be constructed that measures the individual's beliefs that the behavior can take place under different kinds of circumstances. The circumstances described in the scale can be more and more difficult. Those who think that they can do the behavior under the most difficult circumstances are highest in self-efficacy. Thus, a high sense of self-efficacy is an especially important facilitating condition. For instance, we can ask, "Can you solve this equation?" A person who says no is very low in self-efficacy. Those who answer yes are higher in self-efficacy. A person who says yes when the question is "Can you solve this equation when you are waiting to board a plane in a noisy airport?" is very high in self-efficacy.

Consider a more specific example. If a person said, "I will look up this reference," but the book that contains the particular reference is not around, the probability that the act will occur decreases. Facilitating conditions modify the probability that habit and intention in themselves will result in the act. They reflect the situation within which behavior may occur.

For those who enjoy the precision that mathematical statements provide, the first equation of the model is:

P_a = (W_H - H + W_I - I )F (1)

where P_a is the probability of an act, W_H and W_I are weights that are positive numbers between 0 and 1.00 and sum to 1.00, H is a measure of habit, I is a measure of intention, and F is a measure of the facilitating conditions.

The weights depend on the novelty of the act for the individual. When the individual is faced with a new situation, the weight for intention is 1.00 and the weight for habit is zero. However, as the person performs the act over and over again the weight for habit keeps increasing until it becomes 1.00 and then the weight for intention is zero. For instance, when one learns a new skill (e.g., riding a bicycle) in the early phases, one's behavior is under the control of intentions, but at the end it is entirely under the control of habits. Once behavior is under the control of habits, it is difficult to "explain" it to others without actually carrying out the act and observing one's own behavior.

Another variable that shifts behavior to habit control instead of intention control is stress. When people are under stress, as in an emergency or under time pressure, their behavior is under habit control. That is why there is so much drilling of emergency procedures in the military or on ships. In an emergency one cannot depend on an intellectual analysis of the situation. One must have the right habits.

To make a behavior automatic under habit control, one needs several hundred trials (Schneider, 1993). But for some jobs the advantages of having a behavior under habit control are immense. Consider these examples. Suppose the job requires a two-category judgment or a four-category judgment (e.g., is this an airplane, a missile, a bird, a shadow?) To match a new stimulus with the correct response on a two-category task takes 0.7 seconds; on a four-category task it takes about 2 seconds if the response has not yet become automatic. If it has become automatic (after several hundred trials) the two-category task or the four-category task take the same time: 0.002 seconds! If the workload changes, the responses that are not automatic require more time, but the responses that are automatic do not require more time. Of course, to get the response to be automatic requires about 10 hours of practice. Thus, it is only in some critical jobs, such as air traffic controller, that this type of training is economically justified.

Another advantage of responses under habit control is that they are not forgotten. Once you learn to ride a bicycle, you can ride one 10 years later. In short, automatic responses are faster, do not degrade with time, and stay in long-term memory. But only for some jobs is it justified to spend the time to make responses automatic. While in most research work the development of automatic responses is unnecessary, there are components of jobs, such as learning the value of the integral of a particular equation, that may speed up the work of scientists.

6.1.1. Determinants of Habits

What are some of the variables that determine a habit? Habits build up as a result of previous rewards. We call such rewards "reinforcements" because they reinforce the link between stimulus conditions and behavior. Behavior is a function of its consequences. As people engage in a particular behavior in the presence of a certain configuration of stimuli, and when desirable events follow the behavior, the probability increases that the configuration of stimuli will in the future produce the same behavior. The behavior eventually becomes automatic, without thinking. When this happens, we say that the act has become "overlearned" and occurs under the control of habits. In that case, behavioral intentions are not relevant as explanations of the behavior.

6.1.2. Determinants of Intentions

Let us now examine what determines behavioral intentions. Three classes of variables are relevant for the determination of behavioral intentions: social factors, act satisfaction, and perceived consequences.

6.1.2.1. Social Factors

Social factors include norms, roles, self-concept of the person, and interpersonal agreements.

6.1.2.2. Act Satisfaction

The second class of variables that determines behavioral intentions is satisfaction associated with the act itself. Many acts are enjoyable in themselves, such as eating certain types of foods, playing the piano, or working on computer problems. Often such acts associated with pleasure have been formed through classical conditioning. In other words, the activity itself was associated with pleasant events in the past and is pleasant to think about, so this factor involves affect (emotion) toward the behavior itself. This affect motivates the person to self-instruct to do the act, and this in turn becomes the behavioral intention that causes the behavior. Working on a challenging research project or working with a noted scientist could fall in this category.

6.1.2.3. perceived_consequences

Finally, the perceived consequences of the act are also important. When we do something, such as publish a paper, we perceive certain consequences. For example, when we publish a paper, we might have the perception that this could lead to a promotion, to recognition, or to a particular reward. It is obvious that each of these consequences is probabilistically associated with the act since there is no certainty that the behavior will have the particular consequence. For example, if the scientist publishes a paper, the probability of promotion may be 0.60; the probability that there will be some recognition associated with the paper may be 0.90. Thus, each act has associated with it a probability between zero and one. So the person thinks, "If I do such and such, then there is a high probability (or low probability) that x will happen." In this case, x is a consequence. Each consequence also has some value to the person. For example, some people would see a promotion as very desirable, but others might not. Obviously, if the consequence has a positive value attached to it, it will increase the probability of the behavioral intention. If the consequence is perceived as negative, it will decrease the probability that the corresponding behavioral intention will be activated. For each of the acts the person may consider, there is a whole string of consequences, each of which has some probability and some value attached to it. To obtain the total effect of these perceived consequences, each person must multiply the probability and the values for every consequence and then sum these products. Intelligent people will make better estimates of these probabilities and values than others.

Thus, we can say that behavioral intentions are a function of (1) social factors such as roles, norms, the self-concept, and interpersonal agreements, (2) the affect toward the act itself, and (3) the total value of the perceived consequences. Because some people are susceptible to social factors and others are susceptible to perceived consequences of the act, each of the three factors can now be given a weight. For example, people who have been socialized to be very sensitive to the views of others, and who have received lots of rewards and punishments in their interactions with others, develop great sensitivity to social norms. Their behavioral intentions are much more influenced by the social factors than by the other two sets of factors. On the other hand, people who have been socialized to be quite independent often give attention to how much pleasure they can get out of a particular situation. Thus, they are likely to pay a great deal of attention to the affect that is attached to the act. Still others are quite interested in the future and in the way the act is going to bring "good outcomes." Such people look at the consequences of the act and are likely to give weight to those consequences.

The consequences of the act can include job autonomy, vacations, fringe benefits, and the opportunity to use time flexibly. For example, to work at home when you otherwise are expected to be at the office can be highly rewarding. Setting difficult but reachable goals with feedback is one of the ways in which a supervisor can motivate a subordinate. In addition, it has been found that interesting work that provides both challenge and variety can be rewarding. Deadlines are like an interpersonal agreement that can be rewarding and that can function as a goal. Recognition, promotions, and the opportunity to grow, to receive more pay, or to have a more secure job can all be motivators.

It is also useful to consider situations that are demotivating. One such situation occurs when the employee feels that the organization discriminates against him or her. Other causes for demotivation are poor interpersonal relationships with a supervisor or with peers, low pay, indifference by the organization, lack of promotion or recognition, and having to work for an incompetent supervisor.

Again, for the sake of those who like mathematical formulations, what we have said above can be summarized by the following equations:

Where I is a measure in intentions

A is a measure of the affect toward the behavior itself

C is a measure of the value of the consequences

S is the social factor that reflects roles (R), norms (N), the self-concept (S_C), and interpersonal agreements (I_A)

P_c is the probability of a consequence

V_C is the value of the consequence

W_S, W_A, W_C are weights that are positive numbers between zero and one and that sum to 1.00

c stands for consequences 1 to n

If a supervisor wants to change the behavior of a subordinate, every one of these variables may be influenced, and of course combinations of these variables may be optimal. For example, the supervisor can associate pleasant events with the desired behavior, so that even a minimal quantity of the desired behavior may elicit the pleasant event (a nod, a smile, a pat on the back, etc.). A discussion of roles, norms, and the resultant interpersonal agreements can influence the S-factor. A discussion of the probable consequences of the particular behavior can influence the P_c. The association of important values of the subordinate with the desired behaviors can lead to higher C. Goals are most effective if they are specific, difficult, and attainable. Such goals can become interpersonal agreements.

6.1.3. Facilitating Conditions

There are a number of factors that facilitate the performance of a behavior. Most of them are situational, such as helpful conditions, the right setting, or access to the resources needed to carry out the behavior. However, there are also internal conditions over which the individual does not have much control, such as the person's physiological state (e.g., hormonal balance), beliefs that the behavior is possible and likely to lead to the successful reaching of goals (sense of self-efficacy), and the level of difficulty of the task relative to the person's ability. For instance, no matter how intensive a researcher's intention to invent a new product, and how brilliant the past record of inventions (habits), there are situations in which no invention will be possible because the person is feeling depressed, or he or she believes that they are not able to have a new idea, or the task is much too difficult relative to the available talent. Some of these conditions can be measured objectively, and others may be estimated by objective observers of the total situation. The point about the F component of Equation 1 is that when it is zero, it can bring the probability of the act to zero, no matter how high the levels of habits or intentions.

Links between desirable behavior and challenges, variety on the job, recognition, promotions, growth, extra pay, extra security, and so on, are too obvious to mention in detail. One can also motivate people by providing deadlines.

Of special importance in R&D labs is whether the organization rewards reasonable risk-taking, innovation, and creativity. Does the organization provide feedback and rewards for good work? What kind of facilitating conditions and environment is the organization providing for motivation of this unique group of talented individuals—the researchers?

Common sense suggests that job satisfaction results in high productivity. However, the empirical evidence is not supportive of this expectation. This happens in part because one can get high production without satisfaction (e.g., in coercive situations, such as among slaves and in prisons, where people may not be able to eat if they do not produce enough), and one might also be able to have high satisfaction without much production (as in situations in which the management lets workers do whatever they like). The evidence suggests that those who receive high pay and have supportive supervisors are high in "extrinsic" satisfaction, which in turn leads to high performance. On the other hand, high performance leads to intrinsic satisfaction—that is, to people enjoying the work itself. If we perceive the work situation as equitable—that is, if our effort is rewarded about as equally as that of others—we are more likely to be satisfied than if we feel that others are getting more or less for their effort.

Scientists crave visibility and recognition. That is certainly true in the West. However, some caution is required in East Asia. People in some East Asian cultures feel distinctly uncomfortable when they are made to "stick out" from the group, even when that means that they are being complimented. A supervisor in that part of the world should first "test the waters" by privately talking to the subordinate. "I think your accomplishments require public recognition. You did extremely well and I want everyone to know that. I am planning ..." See how the subordinate reacts. If the subordinate is clearly embarrassed and does not want public recognition, do an about face: "I respect your wishes and will not do what I had planned." But this last step is complicated. It could be compared to such East Asian occasions as Chinese banquets or the Japanese tea ceremony, where one is frequently offered a special dish or tea. The correct response in such situations is to refuse it politely. The host then insists, and after a few refusals the guest agrees to accept the tea or delicacy. Thus, offer the opportunity for a special recognition again, at another time, and see what happens. The reluctant East Asian scientist may change his or her mind.

Nevertheless, in general, managers should provide opportunities for visibility (e.g., invitations to give a lecture or to make presentations to important customers) as a reward. There are many studies of compensation, and this is a topic we will not discuss in detail, except to point out the desirability of linking the scientist's behavior to the goals of the laboratory by offering personal rewards, prizes, or recognition for actions that promote such goals.

At different stages in their careers, people need a different mixture of rewards. For example, Busch, Venkitachalam, and Richards (2008) note that younger IT employees like to be recognized formally by their boss, but older IT employees seek out recognition from colleagues with regard to performance. Keep in mind that scientists and engineers need to increase their skills in order to learn more. So training, growth, and transferring to different jobs can be seen as rewards. In middle career (age 35–50), recognition, esteem, and visibility are the most important rewards. In late career (50–70 years), security, health and pension benefits, recognition, and visibility are the important rewards.

6.1.4. Support for the Model

Numerous studies support this formulation (for a review see Triandis, 1980). We will mention only two as examples. In one study, foremen were instructed by the experimenters to behave poorly. In a control group, foremen were instructed to behave normally. Half the workers were doing a new job; the other half did a job that they had been doing for a long time. The instructions to the foremen influenced the productivity of the workers only when the workers were doing a new job. In other words, when the job was under habit control, the supervisor's behavior was irrelevant, but when it was under intention control a supervisor who treated a subordinate badly depressed the subordinate's performance. In other studies (Fiedler, 1986a) the effect of the leader's intelligence (which is relevant to the utilization of intentions, as we will see below) and years of experience (which is relevant to the extent the leader's behavior is under the control of habits) were related to the effectiveness of the team under the supervision of the particular leader. Under conditions of time pressure, stress, or emergency (when habits are likely to control the behavior), the experience of the leader correlated with group effectiveness. Under those conditions the leader's IQ was unrelated to effectiveness. However, under conditions of low stress the opposite pattern of correlations was obtained. In short, when people are under stress they use their habits more than their intentions, and so they do not utilize their intelligence as much.

The example with the foremen who behaved poorly for experimental purposes makes another important point: It is possible for people to be very dissatisfied and yet be highly productive. For example, an employee who sees high productivity as a means to a promotion out of a boring job may be low in job satisfaction but extremely high in productivity.

In fact, the factors that determine productivity are not the same as the factors that determine job satisfaction. Productivity depends on how many high-effort, high-quality behaviors are attempted by the person. The model we just described indicates the factors that will lead to high productivity: beliefs that others expect high productivity; the person's beliefs that high productivity is appropriate and that he or she is the type of person who is highly productive; instructions from supervisors that point to high but attainable goals; specific goals; the availability of clear procedures for reaching goals; feedback from supervisors concerning goal attainment; enjoyment of high-effort, high-quality behaviors; beliefs that such behaviors will have desirable consequences (e.g., promotions); and the conversion of the intention to produce high-quality outputs into habits—that is, automatic behaviors that the person carries out without thinking.

Job satisfaction, on the other hand, depends on how much one gets (resources such as status, training, money, goods, services) relative to what one expects. If one gets slightly more than one expects, this may boost productivity, but the effect is likely to be short-lived. One soon rationalizes that the extra resources obtained are "well deserved." If one gets less than expected, one is dissatisfied. Expectations depend on our perception of what we bring to the job, relative to what others bring, and what we get out of the job, relative to what others get. So a researcher who believes that his or her international reputation is much greater than that of colleagues is likely to be dissatisfied with the same pay as that of colleagues. Note that we are discussing perceptions, not reality. It is perceptions that determine expectations.

In other words, the manager of an R&D laboratory who wants subordinates to be productive must ensure that the norms of the laboratory (perception of what people are expected to do, what is "proper" behavior) call for high effort and high quality and must present difficult but attainable, specific goals to subordinates. Furthermore, the environment should be structured in such a way that there are clear procedures for reaching such goals, and feedback is provided when the goal is reached. Rewards should be given liberally, for both minor accomplishments (a nod, a pat on the back) and major accomplishments (special prizes and awards). This will link the high productivity behaviors to enjoyable situations and to beliefs that such behaviors result in benefits.

A manager who wants subordinates to be satisfied should provide as many rewards as feasible (see below for varieties of rewards) and also realistic expectations concerning such rewards. Publishing wage surveys that indicate that the laboratory pays better than average, for instance, would be helpful. The fact that a famous scientist at another laboratory is underpaid is worth mentioning to one's subordinates. Discussing how much value is placed on various factors that one brings to the laboratory—advanced degrees, years of experience, publication record, editorships, listings on the masthead of specific journals, honorary degrees, elections to high-status positions in scientific societies, and so on—can be helpful. This is true because a subordinate may think that one of these factors is worth much more than does a supervisor, thus creating discrepancy between what the subordinate expects and what they are likely to get. For example, does the laboratory really care if a scientist gets an honorary degree? Usually such events promote the individual but not the lab and may not improve the lab's productivity. Or is a book summarizing a program of research worth as much as N refereed publications? Again, a major discrepancy between subordinate and supervisor perceptions can occur, and clear discussion of such issues can be most beneficial.