M5-

Psychology as the Science of Self-Reports and Finger Movements Whatever Happened to Actual Behavior? Roy F. Baumeister,1 Kathleen D. Vohs,2 and David C. Funder3

1Florida State University, 2Marketing Department, Carlson School of Management, University of Minnesota, and 3University of California, Riverside

ABSTRACT—Psychology calls itself the science of behavior,

and the American Psychological Association’s current

‘‘Decade of Behavior’’ was intended to increase awareness

and appreciation of this aspect of the science. Yet some

psychological subdisciplines have never directly studied

behavior, and studies on behavior are dwindling rapidly in

other subdisciplines. We discuss the eclipse of behavior in

personality and social psychology, in which direct obser-

vation of behavior has been increasingly supplanted by

introspective self-reports, hypothetical scenarios, and

questionnaire ratings. We advocate a renewed commit-

ment to including direct observation of behavior whenever

possible and in at least a healthy minority of research

projects.

For decades now, psychology students have been taught from the

first day of class that psychology is the science of behavior and

that its ultimate goal is to describe and explain what people do.

Is that a fair description? The answer varies with the specific

area of psychology. Neuroscience and cognitive psychology have

never had much to say about the meaningful activities people

perform in their daily lives, nor have they really intended to.

These fields are more interested in understanding the internal

workings of the mind and brain rather than behavioral outcomes.

In contrast, animal learning and developmental psychology

have consistently focused on behavior, perhaps because

participants studied by these fields generally cannot fill out

questionnaires or read prompts on a computer screen, and

their studies have ranged from bar pressing as a function of

rewards to behavioral coordination between small children and

their parents.

The fields of social and personality psychology, however, offer

a special and discouraging case. Both of these related fields have

a mandate to study the important social behaviors that compose

the very texture of human life, with personality psychology fo-

cusing on individual differences in those behaviors and social

psychology exploring situational influences. But personality

psychology has long relied heavily on questionnaires in lieu of

behavioral observation, a state of affairs that has begun to

change only recently and ever so slowly, at that. Even worse,

social psychology has actually moved in the opposite direction.

At one time focused on direct observations of behaviors that

were both fascinating and important—a focus that attracted

many researchers to the field in the first place—social psy-

chology has turned in recent years to the study of reaction times

and questionnaire responses. These techniques, which prom-

ised to help to explain behavior, appear instead to have largely

supplanted it. The result is that current research in social and

personality psychology pays remarkably little attention to the

important things that people do.

The 1990s was named the ‘‘Decade of the Brain’’ by the

American Psychological Association (APA). This widely-ad-

vertised rubric, promoted heavily by the APA, focused attention

on the importance of and advances in research on brain pro-

cesses. It was wildly successful, to the extent that many funding

agencies jettisoned many other research priorities as they

poured money into expensive brain research and articles and

conference sessions on brain studies proliferated. Brain

researchers have always been more interested in brain and

nervous system functioning than in behavioral implications.

Address correspondence to Kathleen Vohs, Suite 3-150, 321 19th Avenue South, Marketing Department, Carlson School of Manage- ment, University of Minnesota, Minneapolis, MN 55455; e-mail: [email protected].

PERSPECTIVES ON PSYCHOLOGICAL SCIENCE

396 Volume 2—Number 4Copyright r 2007 Association for Psychological Science

Ironically, however, their research has benefited hugely from the

conviction by funding agencies and the public at large that

anything a neuron does must be behaviorally important. Such

relevance has been demonstrated once in a great while (e.g., in

the work by Damasio, 1994, on the interaction between emo-

tional and cognitive systems in decision making), but more often

it has merely been taken on faith. Meanwhile, the increase in

study of the brain has helped erode interest in the actual ob-

servation of behavior.

It seemed an extremely wise move therefore when, impressed

by the success of the brain decade, APA came up with the idea of

making the first decade of the new century ‘‘The Decade of

Behavior.’’ The goal was to focus attention on the contributions

of psychology toward understanding and affecting important

behaviors and consequent life outcomes, thereby adding rele-

vance, credibility, and (one hoped) big research budgets to the

enterprise. This emphasis was—or at least should have been—

especially welcome to social and personality psychologists,

whose research programs would seem to be in a position to

benefit greatly from a renewed recognition of the importance of

behavior.

It is now past halfway through the putative Decade of Behavior

and is therefore a fair time to ask, ‘‘How’s it going?’’ In particular,

how are social and personality psychologists doing? To antici-

pate our answer, we think they are doing fine in many respects—

but not in respect to studying behavior.

LOOKING FOR BEHAVIOR

With that question in mind, we picked up a recent (January

2006) issue of the Journal of Personality and Social Psychology

(JPSP), by consensus the premier journal in our subdisciplines

(we are all social and personality psychologists). It is undeniably

a fine issue, offering important advances in the topics the arti-

cles address. The methods are rigorous, and the discussions are

thoughtful. The editors, reviewers, and authors did their jobs

well.

But behavior is hard to find. Or if it is there, it is rather

different than what we had imagined it to be. If this issue offers a

representative sample, then human behavior is almost always

performed in a seated position, usually seated in front of a

computer. Finger movements, as in keystrokes and pencil

marks, constitute the vast majority of human action.

In fact, a remarkable amount of ‘‘behavior’’ turns out to be

really just marks on a self-report questionnaire. Sometimes

these questionnaires ask people to report what they have done,

will do, or would do. More often, they ask people to report what

they think, how they feel, or why they do what they do. In other

words, most personality and social psychological studies gather

self-reports of inner states.

Nisbett and Wilson (1977) thought they had discredited in-

trospection back in the 1970s, when they demonstrated that the

factors that drive behavior are often invisible to the people who

perform it. As their title expressed, most introspective reporting

involves ‘‘telling more than we can know.’’ Although aspects of

this research became controversial, it is abundantly clear from

their studies, other research, and everyday observation that

people have not always done what they say they have done, will

not always do what they say they will do, and often do not even

know the real causes of the things they do. These discrepancies

mean that self-reports of past behaviors, hypothetical future

behaviors, or causes of behavior are not necessarily accurate.1

Nonetheless, self-report appears to have all but crowded out

all other forms of behavior. Behavioral science today, at least as

represented in JPSP, mostly involves asking people to report on

their thoughts, feelings, memories, and attitudes. Occasionally

they are asked to report on recent or hypothetical actions. Or,

somewhat differently (and more rarely), reaction times, implicit

associations, or memory recall might be assessed in the service

of illuminating a cognitive process. But that is as close as most

research gets. Direct observation of meaningful behavior is

apparently passé.

This is certainly quite an ironic turnabout from Nisbett and

Wilson’s (1977) critical stance. In fact, Wilson’s more recent

work has shown that when people introspect to analyze the

reasons for their actions, they often mislead themselves (Wilson,

2002). In a choice rhetorical flourish, he advises people who

seek self-knowledge to eschew direct inspection and instead

consult books on social psychology. Yet the books on social

psychology are increasingly based on research that itself is

heavily based on introspection.

The move from behavior to an emphasis on introspective self-

report and hypothetical responses to imagined events is poten-

tially a hugely important shift in the very nature of psychology.

Psychological science started out in the 1800s with introspec-

tion (e.g., Wundt, 1894). One major development of the 20th

century was the shift from introspection to direct observation

of behavior, widely regarded as an advance in the development

of scientific methodology. Did someone, somewhere, decide that

that had been a mistake and that we should now go back to in-

trospection?

Let’s take a closer look at this recent issue of JPSP, which

was chosen only for convenience and is presumably represen-

tative. It contains 11 articles reporting 38 studies. The closest

thing to direct observation of behavior in the dependent mea-

sures of any of these studies was a participant making a decision.

That is, one study asked participants to choose between two

stimulus persons (who were made known to participants via

photographs) to give them the postexperimental interview. Apart

from that borderline case, not a single one of those 38 studies

contained direct observation of behavior. The dependent mea-

sures consisted entirely of ratings, either on paper question-

1This does not mean they are never accurate, but rather that there is no way to know whether they are accurate or not without direct observation of behavior.

Volume 2—Number 4 397

Roy F. Baumeister, Kathleen D. Vohs, and David C. Funder

naires or computer-administered stimuli. The ratings were

mainly introspective self-reports.

Some of the procedures included hints of behavior along the

way. One study had participants read a fictional police report

about a violent act and express a (nonbinding) opinion as to the

appropriate prison sentence for the perpetrator. (So at least they

read about someone else’s behavior, albeit fictional behavior.)

Four studies had participants take tests: one for the purpose of

legitimizing bogus feedback, the other three as a basis for as-

sessing the accuracy of self-ratings of performance. Some of the

questionnaires asked people to report on their past behaviors.

Several asked people to read things, such as descriptions of

hypothetical behavior. One study had participants cross out all

instances of the letter e in a page of printed text.

So that is behavior today in the leading journal of social and

personality psychology: ratings and more ratings, occasionally

making a choice, reading and taking a test, and crossing out

the letter e.

Behavior fared only slightly better in the previous issue: Out

of 38 studies in 13 articles, there was one that measured nego-

tiation moves (Bowles, Babcock, & McGinn, 2005) and one that

studied ‘‘how an individual actually behaves during an induced

conflict’’ (quoted from Knee, Lonsbary, Canevello, & Patrick,

2005, p. 998). Note the authors’ use of the term ‘‘actually,’’ which

illustrates their awareness of how unusual it was to observe

behavior directly.2 That study induced and videotaped a dis-

agreement between romantic partners and then coded for un-

derstanding versus defensive behaviors (mainly speech acts;

Knee et al., 2005). Those two studies are about real behavior, but

again they are only 2 out of 38 studies. One additional study

included a behavioral independent variable (of sorts) that con-

sisted of people reading their e-mail message aloud before

sending it (as opposed to just sending it; Kruger, Epley, Parker,

& Ng, 2005). There was also a study that used a questionnaire for

self-report of behavior.

We do not doubt that other surveys would yield similar

numbers. Social and personality psychologists do not report

much actual behavior in their premier journal or elsewhere.

WAS IT ALWAYS THUS?

Our impression is that this trend has been building for a while.

In what psychologists from the baby-boom generation may re-

member as a golden age, social psychology for a time was

characterized by studies that directly observed important be-

haviors in vividly evocative contexts (see Aronson, Brewer, &

Carlsmith, 1985). We suspect that more than one social psy-

chologist was inspired onto his or her career by an undergrad-

uate class lecture on John Darley’s and Bibb Latané’s studies

of bystander intervention (e.g., Darley & Latané, 1968) or on

Stanley Milgram’s (1975) obedience studies, which put real

people into emotionally powerful situations and then watched

what they did. Even many classic studies of inner variables such

as attitudes and guilt contained dramatic behavioral experi-

ences prior to the self-reporting of inner states (e.g., Aronson &

Mills, 1959; Festinger & Carlsmith, 1959). Studies like this

faded from view, however, when the field embraced the cognitive

revolution in the 1980s, and the success (and apparent rigor and

prestige, see Rozin, 2001) of cognitively-framed studies may

have encouraged many researchers to concentrate on the self-

report measures that were appropriate for those studies rather

than struggle for difficult and expensive behavioral observation.

The impressively successful Decade of the Brain, as we have

already noted, also demonstrated to researchers the success,

prestige, and funding that could accrue to studies exploring

inner psychological processes while postponing, perhaps

indefinitely, examination of the behavioral results of these

processes.

In personality psychology, meanwhile, the interest in behavior

was weaker all along in some ways, especially given the core

emphasis on measurement of traits, for which self-report has

long been the prominent method. Although sophisticated psy-

chometric methods have been developed, the primary method

for validating a personality questionnaire is still to demonstrate

how it correlates with other questionnaires (Funder, 2001).

Hundreds of studies on the ‘‘structure of personality’’ seek that

structure amongst the correlations between questionnaires. One

major recent project that aimed to compare the utility of several

major personality inventories in their ability to predict behavior

did so by seeing how well each one predicted self-reports of

behavior (in other words, yet another questionnaire). Even

personality psychologists who evince skepticism about person-

ality traits are not immune; their research, too, is based almost

exclusively on self-reports or hypothetical predictions, with only

rare (and therefore highly notable) exceptions (e.g., Wright &

Mischel, 1987). In recent years, a few personality psychologists

have begun to look again at behavior in the laboratory (e.g.,

Borkenau, Riemann, Angleitner, & Spinath, 2001), assess be-

havioral residue (e.g., the condition of a student’s dorm room or

a worker’s office cubicle; see, Gosling, Kos, Mannarelli, &

Morris, 2002), and develop methods to code directly observed

behavior along meaningful dimensions (e.g., the Riverside Be-

havioral Q-sort; Funder, Furr, & Colvin, 2000). These efforts

remain rare, however, and have not been particularly influential

so far. The dominant method throughout personality psychology,

to this day, is the questionnaire (Funder, 2001).

We were moved by these questions and reflections to take a

slightly more systematic and quantified look at behavior in JPSP

over the decades. Two of the authors went to libraries and se-

lected issues from 1966, 1976, 1986, 1996, and 2006 for coding.

We used our birth months (March and May) for each year, and

we coded all studies reported in the issue as to whether they

included any direct observation of behavior. Coding was

2Since we wrote this, we have noticed the increasing frequency with which behavior is called ‘‘actual behavior,’’ presumably to distinguish from the other, more commonly studied kinds or perhaps just to dramatize its rarity.

398 Volume 2—Number 4

Eclipse of Behavior

deliberately liberal, such that a study qualified as having be-

havior if any element involved behavior—that is, if the study

used any manipulation, any dependent measure, or even used

behavior as the conduit for manipulating the independent

variable (e.g., taking a test and getting feedback on it). Hence,

each study in each paper had a score of 0 (no behavior) or 1 (one

or more behavioral element). Some articles were not included

in the final tally because there were no empirical data reported,

such as in conceptual or statistical papers or commentaries.

Self-reports of past behaviors or of hypothetical behaviors did

not count. The use of archival behavioral data (e.g., crime sta-

tistics; donating blood) qualified. Reading about someone else’s

behavior was not counted as behavior.

The two issues for each same year furnished quite similar

numbers except for 1966, in which the very young journal had

several peculiar features (including mostly one-study papers

and some short reports, one of which oddly reported four brief

and very similar studies that were ultimately counted as one),

and so we coded two additional issues for that year to furnish a

broader base; in addition, one of the originally coded issues was

lost because of catastrophic computer failure. In all, we coded

304 studies across the 11 issues of JPSP.

Figure 1 shows the results of this coding. Back in 1966, when

most articles contained only a single study, about half of these

involved actual behavior. The study of behavior increased its

share of the journal into the 1970s. But the use and study of

behavior dropped sharply in 1986, and the subsequent decades

have seen a continued downward trend. Apparently, the study of

behavior has been in a steady decline since the early 1980s.

CAVEAT

We want to be very clear that we see nothing wrong with what

social and personality psychologists are doing, as far as it goes.

Self-reports of behavior, emotion, intention, and thoughts

are often illuminating, may be the appropriate method for

certain topics (e.g., studies of attitudes or emotional experi-

ence), and sometimes are all that is possible. Such measures can

and do lead to important and interesting knowledge that will

advance theory. But the restriction of methods also serves to

constrain the topics that are addressed in the first place (Rozin,

2001). In other words, our chief complaint is with what social

and personality psychologists, and perhaps others in the field,

are not doing.

Surely some important behavior involves standing up?

Or actually talking to another live person, even beyond

getting instructions about how to sign a consent form and

activate the computer program? Whatever happened to help-

ing, hurting, playing, working, taking, eating, risking, waiting,

flirting, goofing off, showing off, giving up, screwing up,

compromising, selling, persevering, pleading, tricking, out-

hustling, sandbagging, refusing, and the rest? Can’t psychol-

ogy find ways to observe and explain these acts, at least once

in a while?

WONDERING WHY

There are probably many reasons for the sorry state of behavioral

study during these last few years of the Decade of Behavior. For

example, sometimes direct behavioral observations are unethi-

cal, unfeasible, or impossible. For the foreseeable future, most

studies of everyday behaviors ranging from eating to sexual

behavior and from exercise habits to drug use will have to rely

primarily on self-report despite the obvious disadvantages.

Moreover, if one wants to know what a participant is thinking or

feeling, there is little alternative but to ask. And even under the

best of circumstances, observing actual social behavior is more

difficult, challenging, and inconvenient than just asking for

ratings or sitting a participant in front of a computer screen and

measuring his or her keystrokes or reaction times. The field is

highly competitive, and the top journals require multiple stud-

ies, so the struggle to observe behavior, even when it would be

possible, may well make it harder to crank out the volume of data

that academic success now requires. Moreover, the sad fact is

that many studies fail to show meaningful significant differ-

ences. A failed behavioral study is an expensive failure and

could even be a major career setback. Last, and perhaps most

centrally, journals do not seem to give extra points or consid-

eration to studies that observe behavior instead of just getting

ratings, so why bother?3

Our data on JPSP across the years points to the early 1980s as

the beginning of a huge decline for observed behavior in studies.

Several things had changed in the journal between 1976 and0

10

20

30

40

50

60

70

80

90

1966

Year

1976 1986 1996 2006

Fig. 1. Percentage of studies from Journal of Personality and Social Psychology that included behavior (1966–2006).

3One of the authors of this article, in his previous capacities as Associate Editor of one personality journal and as Editor of another, did follow an (un- advertised) policy of trying very hard to accept any article that included any direct measurement of behavior, but he received very few such submissions.

Volume 2—Number 4 399

Roy F. Baumeister, Kathleen D. Vohs, and David C. Funder

1986. First, the journal had been split into three sections, and

they were allocated in ways that might be taken as at best in-

different to behavior. Second, editorial policy changed toward

explicitly favoring articles with multiple studies, and as we have

suggested, it is far easier to do many studies by seating groups in

front of computers or questionnaires than to measure behavior

over and over. To be sure, other trends in the field at that time

may also have contributed. These would include the cognitive

revolution (itself a strong push to focus on inner process and

downplay so-called actual behavior) and the steady increase in

restrictive Institutional Review Boards (IRBs).

IRBs may be more likely to raise objections to behavioral

measures than to ratings. From an IRB perspective, it seems far

less intrusive to ask someone what she would eat than to observe

how much she actually eats.4 The problem is, of course, that

hypothetical behavioral responses may be wildly inaccurate,

and in any case, there is no way to know unless actual behavior is

also measured. In a similar vein, it may seem less intrusive to

measure ‘‘prejudice’’ as indicated by speed of key presses in

response to prompts on a computer screen than to actually see

how a person interacts with a person against whom he or she

might or might not be biased. And it is certainly easier. But, as

psychologists bow to pressure from the IRBs to avoid anything

that might have the remotest chance of slightly upsetting their

research participants and to competitive pressures to produce

lots of studies per paper, they sacrifice the scientific quality of

the discipline. Not to mention abandoning its original goal of

being the science of behavior.

Ratings, reaction times, and similar measures are surely

necessary. The maturation of the field has required a great rise of

interest in inner process. In the 1960s, a researcher could ma-

nipulate independent variables, measure behavior, and simply

speculate about the internal mediating process. Now the re-

searcher is required to demonstrate the inner process as well.

Adding ratings surely made for better science. But in principle,

the ratings and self-reports were supposed to shed light on the

behavior, not replace it.

To put it another way: Once upon a time, perhaps, psycholo-

gists observed behavior and reported what they saw, along

with their theories about why it happened. The emergence of

competing theories, and therefore competing explanations, led

psychologists to push each other to show what happened inside

the person to produce the behavior. Gradually the focus shifted

on these debates about inner processes, and journals started

publishing studies that made significant contributions about

demonstrating inner processes. Somewhere along the way, it

became acceptable to publish data on inner processes without

any real behavior included at all, which eventually became

the norm.

WHAT’S THE PROBLEM?

One might ask, ‘‘Is there anything wrong with learning about

inner process?’’ We think not. But behavior matters too. It

cannot be blithely assumed that responding to questionnaires

is enough to tell us all we need to know about actual life. It is

necessary to study actual behavior sometimes. For example,

West and Brown (1975) conducted the same experiment two

different ways: once by asking people what they would do in this

situation, and once by actually staging the event. The experi-

ment involved an ostensible accident victim standing on the

street, asking passersby for money to help pay for medical care

at a nearby clinic. Actual and hypothetical behavior differed in

two major ways. First, the levels of help were dramatically

different. Asked how they would react to such a request, par-

ticipants said they would give fairly generously, but when the

experiment was conducted live, actual donations averaged

barely over 10 cents in some conditions.

The other difference is even more important for psychology’s

goal of building sound theories about behavior. The victim’s

attractiveness did not have a significant effect on hypothetical

donations, but it did have a significant effect on real donations.

This occurred despite the lower overall generosity in actual

behavior (hence, overcoming any possible floor effect). A re-

searcher who followed the common contemporary method of

relying solely on hypothetical behavior would draw a false

conclusion that would omit an important and significant con-

tributor to actual behavior.

In recent years, the reasons for doubting self-reports and the

resulting need to observe actual behavior have increased.

Affective forecasting studies systematically show the inaccu-

racies of people’s predictions about how they will react and feel

(e.g., Wilson & Gilbert, 2003). Studies on judgment and deci-

sion making have likewise shown that people’s predictions are

inaccurate and that hypothetical decisions do not reliably match

actual ones. For example, in hypothetical decisions, people are

moderately risk averse regardless of the amount of money they

imagine is at stake, but when actual money is used, people

become dramatically more risk averse as amounts increase (Holt

& Laury, 2002).

This issue has arisen before. During the 1960s and 1970s,

both personality traits and attitudes came under vigorous attack

on precisely these grounds. Mischel (1968) wrote a famous and

influential critique arguing that personality trait measures sel-

dom have appreciable correlations with measures of actual be-

havior. Wicker (1969) went so far as to suggest that attitudes,

which had been touted by Gordon Allport as social psychology’s

most important concept, should be abandoned. The fields re-

sponded with impressive programs designed to show that, yes,

one could predict some of the people’s behavior some of the time

4One of the authors of this article spent part of her career studying how much ice cream dieters and nondieters would eat under different conditions. One university’s IRB feared that measuring eating behavior—namely, having to admit to participants that their eating was being recorded—might cause dieters to go into a tailspin and develop disordered eating habits.

400 Volume 2—Number 4

Eclipse of Behavior

from attitudes and trait scales (e.g., Ajzen, 2000; Glasman &

Albarracı́n, 2006, Kenrick & Funder, 1988).

The issues remain today, and indeed the question of attitude–

behavior consistency and behavioral predictability remain im-

portant research topics. But in a recent class devoted to exam-

ining articles published in JPSP, the students were surprised to

find that the latest article on attitude–behavior consistency did

not bother measuring behavior—it asked participants to imag-

ine what they might do in hypothetical situations. Similarly, we

have already mentioned how at least one study has sought to

assess the predictive validity of personality trait questionnaires

via their correlations with a measure of behavior that was itself

a questionnaire. This hardly seems fair. Social psychologists

cannot claim that attitudes or personality predict behavior if by

behavior they do not really mean behavior.

The problem goes even deeper. Although self-reports, reac-

tion times, implicit associations, and the like are good and even

ideal methods for examining certain topics, we believe that

psychology has tilted towards examining precisely those topics

for which these methods are appropriate and away from every-

thing else (Rozin, 2001). This leads us to wonder, in the rush to

test competing theories of internal processes as our discipline

moves away from the description of important behavioral phe-

nomena, what goes missing? What questions are researchers

forgetting to even ask? A partial list is not difficult to generate:

� How do people with different degrees of a personality trait

behave differently?

� How do situational variables such as physical aspects, social

relationships, and cultural structures affect what people do?

� How do prejudiced individuals actually treat the objects of

their prejudice?

� How do those who have been discriminated against respond

behaviorally?

� How and when do women and men act differently in situations

ranging from the first date to leading a meeting?

The reader is invited to make his or her own additions to this

list—it is not difficult to do. Again, our point is not that topics

like these are or never were addressed (e.g., see Nisbett &

Cohen, 1996, for fascinating studies of the relationship between

culture and aggressive behavior), but that they are neglected

relative to the study of cognitive process and certainly relative to

their intrinsic importance. This is why APA’s Decade of Behavior

risks becoming a laughingstock for those in social and person-

ality psychology, the fields that should benefit the most from

this initiative.

Also, it is very possible that the abandonment of behavior

could be seriously detrimental to the field’s goals and broad

influence. A recent president of the Society for Personality and

Social Psychology articulated in the society newsletter that

these fields are suffering from failing to get their message across

to outsiders. They are seen as not making much interesting

progress, even though insiders know that the conferences and

journals are filled with exciting new work. But perhaps scholars

in other fields and even undergraduates find it difficult to ap-

preciate the excitement of the work when it rests on correlations

among questionnaire items or significant differences in reaction

times. The dramatic behaviors of the early years of social psy-

chology experiments are still featured in the textbooks, and

probably for good reason.

TODAY’S DILEMMA

Ironically, psychologists who study behavior today find them-

selves at a disadvantage. Everyone would probably agree that

the ideal paper would report both direct observation of behavior

and measurement of inner processes that mediate and produce

those behaviors. But if you only have one without the other,

preferences are lopsided. Data on behavior without inner pro-

cess are regarded as unpublishable by most journals. Grant

reviewers often behave similarly. One of the present authors

submitted a grant proposal for a behavioral study that a reviewer

criticized on the grounds that it did not include ‘‘psychological

variables,’’ apparently meaning internal process measures.5 By

this definition, behavior is not even a psychological variable!

Behavior by itself is regarded as only a beginning, an unsolved

puzzle. Meanwhile, however, a study of inner process without

behavior is acceptable.

When confronted with a study reporting behavior but not inner

process, reviewers will immediately ask, why did this happen?

Researchers need to show what goes on inside. But when con-

fronted with a study reporting inner process but no overt be-

havior, reviewers almost never ask, ‘‘Would this actually alter

behavior?’’ Inner process is considered interesting and impor-

tant in its own right, without any proof that it has any implica-

tions for what people do.

Given those unequal contingencies, it is not surprising that

researchers have turned away from behavior. It is apparently

more trouble than it is worth. Ratings are the keys to success,

and they are publishable with or without behavior. Behavior,

meanwhile, is not publishable by itself without ratings, and

moreover, behavior often has a nasty way of complicating the

cleaner, more elegant picture that one can get from ratings alone.

AFFIRMATIVE ACTION FOR ACTION?

We wish to suggest, gently and respectfully,6 that social and

personality psychology try to put a bit more behavior back into

the science of behavior (as psychology still advertises itself).

There is no need to stop asking for ratings or analyzing reaction

times, but perhaps psychologists could all push themselves to

include an occasional study that includes direct observation

5 This is not just sour grapes; the study was funded despite a negative review, which is itself an unusual outcome.

6Any attentive reader can tell by now that we are nothing if not gentle and respectful.

Volume 2—Number 4 401

Roy F. Baumeister, Kathleen D. Vohs, and David C. Funder

of what Knee et al. (2005) poignantly called ‘‘actual behavior.’’

To be sure, behavioral observation is not always ethical or fea-

sible, as we have mentioned. But when it is, why not include it?

Researchers could put a bit more effort into developing methods

for observing behavior directly (e.g., Furr & Funder, 2007). We

could do more to build on efforts such as the behavioral batteries

for personality assessment designed by Jack and Jeanne Block

(Block, 1993), the behavior-sampling technologies pioneered by

Matthias Mehl and his colleagues (Mehl, Gosling, & Penne-

baker, 2006), behavioral observations in real life such as those

used by James Dabbs, Chris Fraley, and their colleagues (Dabbs,

Hargrove, & Heusel, 1996; Fraley & Shaver, 1998), work on

performance as a function of approach–avoidance frames

by Andrew Elliot and Carol Dweck (e.g., Elliott & Dweck, 1988),

self-regulation work by Baumeister and colleagues (e.g.,

Baumeister, Bratslavsky, Muraven, & Tice, 1998; Vohs,

Baumeister, & Ciarocco, 2005), the studies on eating behavior

by Heatherton and colleagues (Heatherton, Herman, & Polivy,

1991; Vohs & Heatherton, 2000), and speech-sampling methods

such as those developed by Lisa Feldman Barrett, James

Pennebaker, Lisa Fast, and their colleagues (Fast & Funder, in

press; Feldman Barrett, Williams, & Fong, 2002; Pennebaker,

Mehl, & Niederhoffer, 2003).

Perhaps reviewers, editors, and granting agencies could even

give a little extra preference to studies that contain behavior, in

the spirit of affirmative action for the promotion of method-

ological diversity. If others share our view that the current sys-

tem discourages scientists from observing behavior, then

perhaps more vigorous changes can be made to redress the

imbalance. Maybe a new section of JPSP could be earmarked for

studies of behavior. Or perhaps one of the new journals that the

Association for Psychological Science is introducing could be

devoted to behavioral work. Having such a devoted outlet would

reduce the (apparently crippling) demand that behavioral

studies must compete for space with the easier-to-conduct, and

therefore correspondingly more rigorous and plentiful, studies

that use only ratings.

Let us stress that we are not criticizing APA’s initiative on the

Decade of Behavior. We support the goal wholeheartedly. But if

social and personality psychology has given up on behavior, how

can the field expect society as a whole to embrace it? In fact,

even if society (or funding agencies at least) were to embrace the

Decade of Behavior idea, would that benefit the field? The

saddest outcome would be for the powerful and fund-granting

authorities to decide that behavior is important after all and then

to use that as a reason to disrespect the field. They might say, and

not without reason, ‘‘We want to support the study of human

behavior, but personality and social psychologists don’t study

behavior.’’

Acknowledgments— Portions of this article were originally

published by Roy F. Baumeister and Kathleen D. Vohs in

Dialogue, the newsletter for the Society for Personality and

Social Psychology (incorporated from Division 8 of the Ameri-

can Psychological Association). Preparation of this article was

supported in part by National Institute of Health Grant MH

57039 to Roy F. Baumeister, McKnight Land-Grant Professor-

ship funds to Kathleen D. Vohs, and a Visiting Research

Fellowship from the Max Planck Institute for Human Develop-

ment, Berlin, and by National Science Foundation Grant

BCS-0642243 to David C. Funder.

REFERENCES

Ajzen, I. (2000). Nature and operation of attitudes. Annual Review of Psychology, 52, 27–58.

Aronson, E., Brewer, M., & Carlsmith, J.M. (1985). Experimentation in

social psychology. In G. Lindzy & E. Aronson (Eds.), Handbook of social psychology (3rd ed., Vol. 1., pp. 441–486). New York:

Random House.

Aronson, E., & Mills, J. (1959). The effect of severity of initiation on

liking for a group. Journal of Abnormal and Social Psychology,

59, 177–181.

Baumeister, R.F., Bratslavsky, E., Muraven, M., & Tice, D.M. (1998).

Ego depletion: Is the active self a limited resource? Journal of Personality and Social Psychology, 74, 1252–1265.

Block, J. (1993). Studying personality the long way. In D.C. Funder,

R.D. Parke, C. Tomlinson-Keasey, & K. Widaman (Eds.),

Studying lives through time: Personality and development (pp. 9–

41). Washington, DC: American Psychological Association.

Borkenau, P., Riemann, R., Angleitner, A., & Spinath, F.M. (2001).

Genetic and environmental influences on observed personality:

Evidence from the German observational study of adult twins.

Journal of Personality and Social Psychology, 80, 655–668.

Bowles, H.R., Babcock, L., & McGinn, K.L. (2005). Constraints and

triggers: Situational mechanics of gender in negotiation. Journal of Personality and Social Psychology, 89, 951–965.

Dabbs, J.M., Jr., Hargrove, M.F., & Heusel, C. (1996). Testosterone

differences among college fraternities: Well-behaved vs. ram-

bunctious. Personality and Individual Differences, 20, 157–161.

Damasio, A.R. (1994). Descartes’ error: Emotion, reason, and the hu- man brain. New York: Putnam.

Darley, J.M., & Latané, B. (1968). Bystander intervention in emer-

gencies: Diffusion of responsibility. Journal of Personality and Social Psychology, 28, 377–383.

Elliott, E.S., & Dweck, C.S. (1988). Goals: An approach to motivation

and achievement. Journal of Personality and Social Psychology,

54, 5–12.

Fast, L.G., & Funder, D.C. (in press) Personality as manifest in word

use: Correlations with self-report, acquaintance-report, and be-

havior. Journal of Personality and Social Psychology.

Feldman Barrett, L., Williams, N.L., & Fong, G.T. (2002). Defensive

verbal behavior assessment. Personality and Social Psychology Bulletin, 28, 776–788.

Festinger, L., & Carlsmith, J.M. (1959). Cognitive consequences of

forced compliance. Journal of Abnormal and Social Psychology,

58, 203–210.

Fraley, R.C., & Shaver, P.R. (1998). Airport separations: A naturalistic

study of adult attachment dynamics in separating couples.

Journal of Personality and Social Psychology, 75, 1198–1212.

402 Volume 2—Number 4

Eclipse of Behavior

Funder, D.C. (2001). Personality. Annual Review of Psychology, 52,

197–221.

Funder, D.C., Furr, R.M., & Colvin, C.R. (2000). The Riverside Be-

havioral Q-sort: A tool for the description of social behavior.

Journal of Personality, 68, 450–489.

Furr, R.M., & Funder, D.C. (2007). Behavioral observation. In R.

Robins, C. Fraley, & R. Krueger (Eds.), Handbook of research methods in personality psychology (pp. 273–291). New York:

Guilford Press.

Glasman, L.R., & Albarracı́n, D. (2006). Forming attitudes that predict

future behavior: A meta-analysis of the attitude-behavior rela-

tion. Psychological Bulletin, 132, 778–822.

Gosling, S.D., Kos, S.J., Mannarelli, T., & Morris, M.E. (2002). A room

with a cue: Personality judgments based on offices and bedrooms.

Journal of Personality and Social Psychology, 82, 379–398.

Heatherton, T.F., Herman, C.P., & Polivy, J. (1991). The effects of

physical threat and ego threat on eating. Journal of Personality and Social Psychology, 60, 138–143.

Holt, C.A., & Laury, S. (2002). Risk aversion and incentive effects (Andrew Young School of Policy Studies Research Paper Series

No. 06-12). Retrieved July 14, 2007, from http://ssrn.com/

abstract=893797

Kenrick, D.T., & Funder, D.C. (1988). Profiting from controversy:

Lessons from the person–situation debate. American Psycholo- gist, 43, 23–34.

Knee, C.R., Lonsbary, C., Canevello, A., & Patrick, H. (2005). Self-

determination and conflict in romantic relationships. Journal of Personality and Social Psychology, 89, 997–1009.

Kruger, J., Epley, N., Parker, J., & Ng, Z. (2005). Egocentrism over e-

mail: Can we communicate as well as we think? Journal of Per- sonality and Social Psychology, 89, 925–936.

Mehl, M.R., Gosling, S.D., & Pennebaker, J.W. (2006). Personality in

its natural habitat: Manifestations and implicit folk theories of

personality in daily life. Journal of Personality and Social Psy- chology, 90, 862–877.

Milgram, S. (1975). Obedience to authority. New York: Harper & Row.

Mischel, W. (1968). Personality and assessment. New York: Wiley.

Nisbett, R.E., & Cohen, D. (1996). Culture of honor: The psychology of violence in the south. Boulder, CO: Westview.

Nisbett, R.E., & Wilson, T.D. (1977). Telling more than we can know:

Verbal reports on mental processes. Psychological Review, 84,

231–259.

Pennebaker, J., Mehl, M., & Niederhoffer, K. (2003). Psychological

aspects of natural language use: Our words, our selves. Annual Review of Psychology, 54, 547–577.

Rozin, P. (2001). Social psychology and science: Some lessons from

Solomon Asch. Personality and Social Psychology Review, 5, 2–

14.

Vohs, K.D., Baumeister, R.F., & Ciarocco, N. (2005). Self-regulation

and self-presentation: Regulatory resource depletion impairs

impression management and effortful self-presentation depletes

regulatory resources. Journal of Personality and Social Psychol- ogy, 88, 632–657.

Vohs, K.D., & Heatherton, T.F. (2000). Self-regulatory failure:

A resource-depletion approach. Psychological Science, 11,

249–254.

West, S.G., & Brown, T.J. (1975). Physical attractiveness, the severity

of the emergency and helping: A field experiment and interper-

sonal simulation. Journal of Experimental Social Psychology, 11,

531–538.

Wicker, A.W. (1969). Attitudes vs. actions: The relationship of verbal

and overt behavioral responses to attitude objects. Journal of Social Issues, 25, 41–78.

Wilson, T.D. (2002). Strangers to ourselves: Discovering the adaptive unconscious. Cambridge, MA: Harvard University Press.

Wilson, T.D., & Gilbert, D.T. (2003). Affective forecasting. In M.

Zanna (Ed.), Advances in experimental social psychology (Vol. 35,

pp. 345–411). New York: Elsevier.

Wright, J.C., & Mischel, W. (1987). A conditional approach to

dispositional constructs: The local predictability of social

behavior. Journal of Personality and Social Psychology, 53,

1159–1177.

Wundt, W. (1894). Lectures on human and animal psychology (J.E.,

Creighton & E.B., Titchener, Trans.). New York: Macmillan.

Volume 2—Number 4 403

Roy F. Baumeister, Kathleen D. Vohs, and David C. Funder