Tag Archives: motivation

The (Un)Surprising Truth About What Motivates Us

My post yesterday got some interesting reaction, including a link to this video called “The surprising truth about what motivates us.” The video is by Dan Pink, a journalist (but not scientist) who seems to be trying the best he can to express scientific theories about motivation.

I don’t want to pick on Pink too much, so I’ll use an early howler in the video to discuss motivation a bit

We are not as endlessly manipulable and as predictable as you would think. There’s a whole set of unbelieably interesting studies — I want to give you two — that call into question this idea that if you reward something you get more of the behavior you want and if you punish something you get less of it.

There’s a lot to unpack here, and most of it I won’t bother to try. I don’t know who Pink came across who told him people are “endlessly manipulable,” but it was more likely to be a crazy person on the street than a scientist or economist. A “reward” is often defined as a stimulus that leads to an increase in the targeted behavior, so if the “reward” does not lead to more of a behavior, it’s not a reward.

But I’ll forgive Pink for these, as the first statement just leads up to the video, and the second may be the result of terminological confusion. But the idea that a “reward” does not always leads to an increase in a behavior is painfully well known. Even the great behaviorist, B.F. Skinner, would have said that idea is stupid.

In behaviorism, animals have “drives” caused by their species-nature which they attempt to satiate through the accusition of appropriate stimuli. For instance, most species have some drive for hydration. An ant, a dog, and a human will all attempt to inbibe water in some method. But when an animal is no longer thirsty, the “reward” of water is satiated, and the animal will continue on with other drives.

Likewise, “rational choice” economics teaches that humans are driven in the persuit of well-being, and that money, leisure, social status, and other stimuli serve to satiate that drive, based on an individual’s pre-existing “preference schedule.”

But I don’t want to spend time demonstrating that Pink is completely wrong from every seriously held perspective. It’s more interesting to try to find common ground with him, and illustrate how many is important even though it is not the primary motivation for most people.

Most people are motivated to do what they are good at. Specifically, most people have a variety of goals they would like to achieve. To be powerful, popular, and respected might be three of these goals, though there are many more. Likewise, an individual has some understanding of what is required to actually achieve those goals at the given instant. Motivation can be reliable predictable in a wide variety of domains by an indiviuals’ belief, really and right now, that he can actually achieve the specific set of goals required to meet a desired objective. (This concept is referred to as self-efficacy.)

Consider the case of a man who has two objectives: he wishes to stay healthy, and he wishes to be comfortable. But this man has diabetes, and whlie he loves to eat bread, he knows the effect of bread on his insulin levels is dangerous. I don’t believe that individuals are purely driven by the calculated net present value of eating bread v. not eating bread. Rather, without the self-efficacy to avoid unhealthy foods, the individuals’ more base desire to satiate himself on bread will win.

So how would we increase self-efficacy in this case? Clinically, we would conduct a task analysis to understand what is required to really & right now maintain healike

Goal: Stay Healthy
Required Step: Recognize situations in which will-power will be limitd
Required Step: Understand socially acceptable methods of extracting oneself from temptation
Required Step: Execute pro-active steps to satiation hunger drive using less unhealthy substities


An understanding of self-efficacy allows us to understand how economic incentives work on individuals who are not “motivated by money.”

Consider a kind-hearted individuals who has several goals, such as to look-after his family, to labor in a field which is helpful to others, and to make a difference to people he has never met. Even if subject is completely non-materialistic, all of these goals are made easier through higher income. Of course, other things also make these goals easier, and in creating a reward schedule one has to be careful not to impose costs higher than the benefits of an increase in income. No one is stupid enough to believe that an increase in income always by itself leads to an increase in targetted behavior. But it is equally wrong to say that ‘good’ people are not motivated by income, or that the fact that other things can also improve self-efficacy means that financial compensation is not a part of a reasonable reward schedule.

In America, to wrap this up, we do not have a professional teaching cohort. Way too many teachers either washed out of either majors for that to be the case. Rather, we have the remnants of a professional cohort intermixed with the labor force you get when you underpay and mistreat a once-admired (if still somewhat respected) profession.

If we are serious about having students educatd by teaching professionals, we need to treat them like professionals, which includes paying them like professionals.

If we are not serious about that, we need to make idiot-proof scripts that an even-teacher teaching cohort can read during classtime.

Self-efficacy, not self-esteem

Chicago Boyz had an interesting thread comparing the self-esteem movement to China’s focus on engineering education.

It is worth while, but it’s important to note two things that I have observed teaching 400-level classes in educational psychology

1) no serious researcher takes self-esteem seriously
2) nearly every pre-service teacher takes self-esteem seriously

I suspect this is becasue of the influence of the regular education faculty, who have limited training in the theory and methods of psychology, but a lot of exposure to educational fads like emotional intelligence.

To complicated matters, there is a predctively valid concept caled “sellf-efficacy,” which unfortunately shares the same initials as self-efficayc.

Here are examples of self-esteem questions

  • I feel good about myself (True / False)
  • I am a good person (True / False)
  • I am happy with who I am (True False)
  • I am good at math (True False)

Here are examples of self efficacy questions

On a scale of 0 to 100, how confident are you that you can perform the following tasks?

  • Correctly add these two numbers: 5 and 3
  • Correctly add these two numbers: 5353 less 3349
  • Correctly subtract these two numbers: 5 less 3
  • Correctly subtractthese two numbers: 5353 and 3349
  • Correctly multiple these two numbers: 5 and 3
  • Correctly multiplethese two numbers: 5353 and 3349
  • Correctly divide these two numbers: 5 into 3
  • Correctly divide these two numbers: 5353 into 3349

No one takes self-esteem seriously. Criticizing it is like criticizing holocaust denial: an excersize in frustration.

Self-efficacy is one of the best motivational constructs we have.

Update: Using the terms ‘entitlement’ (think: ‘self esteem’) and ‘locus of control’ (think: ‘self efficacy’) the New York Times covers similar ground.

A Simple Model of Performance

Over the past year I have worked on grounding John Boyd’s OODA loop in modern psychology. Both here, and more often in academic drafts, I have described Orientation as System 1 or intuitive cognition, Decision as System 2 or deliberate cognition, Observation as perception and Action as behavior. I still think that the OODA model, or something very much like it, is probably the best high-level conceptual model of the human mind that we have available.

However, it does not help us understand what causes variance in the population, in most tasks. The reason for this is that it does not directly address the issue of Motivation. To use a computer metaphor if Orientation or System 1 is the hard drive, controller cards, and BIOS, while System 2 is RAM, then Motivation is the hypervisor, or that thing that controls the ability of everything else to engage in behaviors to achieve a goal. Motivation, or the hypervisor, is useful because it regulates System 1/Orientation/Long-Term Memory’s and System 2/Decision/Working Memory’s control of behavior, which in turn affects performance. A model of cognition that does not include performance misses both motivation’s regulation of behavior, and motivation’s direct impact on performance. A model derived from Horn et al. (1993) may give us a way forward:

The cognitive components of this model can each be broken down into sub-components. Long-term memory includes both procedural knowledge (how to ride a bike, how to tie your shoes) and declarative knowledge (how you would answer questions: what is a bike? what are shoes?). Many tasks require procedural and decalarative knowledge to operate together. Working-memory includes visual working memory, which is in tasks such as imagining the rotation of objects in three-dimensional space, and verbal working memory, which is used to remember lists, numbers or names. As far as I can tell, motivation loads from both self-efficacy, the believe that as of now you can perform specific tasks to reach a goal, and attitudes, especially the enjoyment of a thing (as it relates to consumption) and desire to block out the world (as it relates to production).

This leaves the question of where the Central Executive is. John Sweller has argueed that it exists in Long-Term Memory, and indeed that no central executive is conceivable other than one that operates through a darwinistic random process within System 1 / Orientation. Alan Baddeley asserts it is a third component of working memory, alongside visual and verbal working memory, because central executive functions appear to tax working memory capacity. Albert Bandura asserts that humans are “agents,” and their Central Executive agency must rely within their Motivation. I don’t know.

The harmonization of John Boyd’s OODA loop with this model of what actually predicts performance is an important task for the field of psychology, especially if it can account for creativity. I hope somebody does it.

Some thoughts on Albert Bandura’s Research Agenda

The work of Albert Bandura is legendary. The breadth of his theoretical work challenges description. Bandura has described subjects as theoretical as comparing the effects of goals and negative self-efficacy (Bandura & Locke, 2003), to as emotionally volatile as the roots of terrorism (Bandura, 2004). Bandura’s social cognitive theory has been described as a way of understanding mass communication (Bandura, 2001a) and human agency (Bandura, 2006). Bandura’s influence is such that he wrote the review of his own theory for the first the Annual Review of Psychology (Bandura, 2001b).

Bandura’s empirical work is also very large. In his empirical work, Bandura not only outlines theories of the world but executes research practices designed to help understand the world. Thus, in his research papers, Bandura refines both his theoretical reach and his methodological approach. In order to briefly describe these processes, I will briefly online three of Bandura’s papers. First, in Bandura, Ross, & Ross (1961), Bandura describes personality in the context of personality. Second, in gender differences article, Bandura and a co-author examining the development of self-regulatory mechanisms (Bussey & Bandura, 1984). Finally, Bandura examined moral disengagement and the state-sponsored killing of prison inmates (Osofsky, Bandura, & Zimbardo, 2005). For each of these studies, a general background is first present. The theoretical discussion and conclusion of the article is then summarized. Finally, the experimental design is discussed. Following this, the papers are compared and contrasted, to emphasize how Bandura’s theories and methods have evolved over time.

The first article to be examined, “Transmission of aggression through imitation of aggressive models,” bridges the gap between behavioralism and cognitivism. In it, Skinner and his co-authors consciously position their research as presenting original problems to the idea that there are no complex mental structure outside of stimulus-response chains. However, in their discussion of results they stick of behavioralist phrases and ideas. The study itself is part of a chain of research, describing itself as building on Bandura & Huston (1961) and quickly followed by Bandura (1965).

Bandura, Ross, & Ross (1961) argued that imitation is a generalizable form of learning, not limited to the repetition of specific actions. This learning was defined as “sufficiently novel patterns of responses which are unlikely to occur independently of the observation of the behavior of a model and if a subject reproduces these behaviors in a substantially identical form.” The research went beyond publications beyond research which argued that children’s actions were merely imitations of what they had observed (Maccoby, 1959) in order to demonstrate that children would engage in behaviors similar to, but different from, those they had seen modeled. The article challenged the ability of behaviorialism to explain this, nothing that responses to stimulus could only (in the behavioralist view of the world) could only be encouraged after similar responses were already provoked. The modeling research that interested Bandura showed that a learner could observe an action and produce a different action, behaving in a manner that required cognitive manipulation of information.

Methodologically, the study would be well received today. 72 students were examined in a 2x2x2 design with 1 control group. Likewise, half of the subjects were males and alf were females, all less than 70 months old. Half the participants saw a model adult of the same sex, while half saw a model adult of the opposite. Additionally, half saw the model act in a hostile manner towards a doll, while half did not. Along with this, Bandura and his co-authors reported results in way that are typical now and that were typical of the older behaviorists. Tests of t values, p values, and tables of results are still regular features of research journals. However, the authors talk about behavioral measures such as the probability of behavior, rather than later formulations such as the average incidence of some behavior.

The second article, published twenty-three years later, is “Influence of gender constancy and social power on sex-linked modeling.” The theoretical construction of the article is new, reflecting the cognitive revolution. The theory is expanded to develop the role of development, a phenomenon missing from Bandura, Ross, & Ross (1961). However, in accordance with Leahy’s (2004) claim of continuity with the older approach, the methods would be recognizable to a behaviorist. The article itself began a collaboration with Bussey, that would lead to three more publications on gender development (Bussey & Bandura 1992; 1999; 2004).

Theoretically, Bussey & Bandura (1984) builds on Bandura, Ross, & Ross (1961). The new paper extends the analysis of the older research into the development of sex roles in children. This is done in a way that respects the lines of research of both Bandura and his colleague, as they cite earlier work by both (including Bandura, 1969 and Bussey, 1979). The substantive focus of Bussey & Bandura (1984) thus begins to make long stretches of time a meaningful part of the experiment, as younger children are compared to older ones. Finally while some phrases are holdovers from Bandura, Ross, & Ross (such as modeling, patterns of responses or behavior, and so on), others (such as information, cognitive, and capacity) reflect the new reality of psychology after the cognitive revolution

The method of is familiar. Two experiments are conducted in Bussey & Bandura (1984), the first being a two-way analysis and the other being a 2x4x2 analysis of variance design. The second experiment in particular recapitulates Bandura, Ross, & Ross (1961): participants are split into either male or female groups, with either male or female models, in one of four treatment conditions. To make sure the reader does not miss this, Bandura & Ross tie both the theory an the procedure to Bandura’s earlier work: “Social power can exert a strong impact on modeling (Bandura, Ross, & Ross, 1963)” (p. 1297). The behavioralist concern for probability of behavior is gone, and the analysis of variance (ANOVA) test appears with its standard P and p values as results.

The third article, published twenty-one more years later, is titled “The role of moral disengagement in the execution process.” The theory in the paper would be incomprehensible to a behaviorist, rather focusing on high-level mental structures over long periods of time with no special reinforcement. The method of this paper is no longer experimental, but rather a correlational study. Osofsky, Bandura, & Zimbardo (2005) follows Bandura’s continuing work on how socialization impacts moral standards (Bandura, 1999, 2002), and is followed by work on moral disengagement in support of the military (McAlister, Bandura, & Owen, 20060.

The theory of the article continues the generalizing trend that began with the transition from Bandura, Ross, & Ross (1961) to Bussey & Bandura (1984). In the first article, the behavior of small children was study. In the second, the socialization of children as they develop into sex roles was examined. In this final piece, Bandura and his co-authors examine the socialization of adults as they do their jobs. The article also continues the trend toward increasing concern over moral behavior. The first publication discussed its findings in the context of imitation, largely ignoring the moral implications of teaching violence. The second more directly addressed concerns of social power and sex roles, and cites Kohlberg (1996). However, the third openly deals with the disquieting impact on morality of being involved in the imprisonment and death of follow human beings.

The methodology is a correlational study that does not create any new conditions, but observes how factors appear to affect participants. Nearly 250 guards three maximum-security prisons in three different states were studied. Some of the guards had served in the execution process, while others had not. Standard survey tools such as likert scales were used by the guards to rate responses to questions, such as “Murderers who receive the death penaly have forfeited the right to be considered full human beings” (p. 380). From the guards’ responses, Bandura and his colleagues were able to conclude that “Executioners made heaviest use of dehumanization, security and economic justifications and disavowal of personal responsibility” (p. 382). For those who participated in executions, moral disengagement increased with the number of executions with which a guard was involved.

The length of Bandura’s career is striking. At first glance, his experiments (ranging from imitation, to sex-role development, to capital punishment) have little to do with each other. Some have even talked about Bandura having a second and third professional life. With respect, I disagree. Bandura has consistently addressed issues at the intersection of modeling, behavior, and social rules. Bussey & Bandura’s (1984) study is a continuation of Bandura, Ross, & Ross’s (1961) examination on imitative child violence, just as Osofsky, Banduyra, & Zimbardo (2005)’s adult participatory violence naturally extends the study of childhood imitative violence. Bandura’s research style has expanded a bit more in this time, moving beyond the experimental studies of the laboratory to correlational studies of the outside world. This perhaps is the more profound shift, sacrifices the rigor of being able to manipulate variables at well for the external validity of conducting research that matters


Bandura, A., & Huston, A. C. (1961). Identification as a process of incidental learning. Journal of Abnormal and Social Psychology, 63, 311-318.

Bandura, A. (1965). Influence of models’ reinforcement contingencies on the acquisition of imitative responses. Journal of Personality and Social Psychology, 1, 589-595.

Bandura, A. (1969). Social-learning theory of identificatory processes. In D. A. Goslin (Ed.), Handbook of socialization theory and research (pp. 213-262). Chicago: Rand McNally.

Bandura, A. (1999). Moral disengagement. In I. W. Charny (Ed.), Encyclopedia of genocide (pp. 415-418). Santa Barbara, CA: ABC-Clio.

Bandura, A. (2001a). Social cognitive theory: An agentic perspective. Annual Review of Psychology, 52, 1-26. In html form from Annual Review Psychology (must be accessed from edu domain).

Bandura, A. (2001b). Social cognitive theory of mass communication. Media Psychology, 3, 265-298.

Bandura, A. (2002). Selective moral disengagement in the exercise of moral agency. Journal of Moral Education, 31, 101-119.

Bandura, A. (2004). Role of selective moral disengagement in terrorism and counterterrorism. In F. M. Mogahaddam & A. J. Marsella (Eds.), Understanding terrorism: Psychological roots, consequences and interventions (pp. 121-150). Washington, DC: American Psychological Association Press.

Bandura, A., & Locke, E. (2003). Negative self-efficacy and goal effects revisited. Journal of Applied Psychology, 88, 87-99.

Bandura, A., Ross, D., & Ross, S. A. (1961). Transmission of aggression through imitation of aggressive models. Journal of Abnormal and Social Psychology, 63, 575-582. Retrieved July 14, 2008, from http://psychclassics.yorku.ca/Bandura/bobo.htm.

Bussey, K. (1979). Same-sex imitation: Fact of fiction? Unpublished doctoral dissertation, University of Queensland, Brisbane, Australia.

Bussey, K., & Bandura, A. (1984). Influence of gender constancy and social power on sex-linked modeling. Journal of Personality and Social Psychology, 47, 1292-1302.

Bussey, K., & Bandura, A. (1992). Self-regulatory mechanisms governing gender development. Child Development, 63, 1236-1250.

Bussey, K., & Bandura, A. (1999). Social cognitive theory of gender development and differentiation, Psychological Review, 106, 676-713.

Bussey, K., & Bandura, A. (2004). Social cognitive theory of gender development and functioning. In A.H. Eagly, A. Beall, & R. Sternberg (Eds.). The psychology of gender (2nd ed., pp.92-119) New York: Guilford.

Kohlberg, L. (1966). A cognitive-developmental analysis of children’s sex-role concepts and attitudes. In E.E. Maccoby (Ed.), The development of sex differences (pp. 82-173). Stanford, CA: Stanford University Press.

Leahey, T.H. (2004). A history of psychology: Main currents in psychological thought (6 ed.). Upper Saddle River, New Jersey: Pearson Prentice Hall.

Maccoby, E.E. (1959). Role-taking in childhood and its consequences for social learning. Child Development, 30, 239-252.

McAlister, A. J., Bandura, A., & Owen, S. V. (2006). Mechanisms of moral disengagement in support of military force: The impact of September 11. Journal of Social and Clinical Psychology, 25(2), 141-165.

Osofsky, M. J., Bandura, A., & Zimbardo, P. G. (2005). The role of moral disengagement in the execution process. Law and Human Behavior, 29, 371-393.