DePaulo deception 1996 classic experiment

Journal of Personality and Social Psychology Colwright 1996 by the American Psychological Association, Inc. 1996, Vol. 71, No. 4, 703-716 0022-3514196/$3.00

Truth and Investment: Lies Are Told to T h o s e W h o Care

Bella M. DePaulo University o f Virginia

Kathy L. Bell Texas Tech University

Participants discussed paintings they liked and disliked with artists who were or were not personally invested in them. Participants were urged to be honest or polite or were given no special instructions. There were no conditions under which the artists received totally honest feedback about the paintings they cared about. As predicted by the defensibility postulate, participants stonewalled, amassed misleading evidence, and conveyed positive evaluations by implication. They also told some outright lies. But the participants also communicated clearly their relative degrees of liking for the different special paintings. The results provide new answers to the question of why beliefs about other people's appraisals do not always correspond well with their actual appraisals.

I n their formal roles as p a r e n t s and supervisors, a n d in their i n f o r m a l roles as colleagues and friends, people often provide us with evaluative feedback. T h e y c o m m e n t on o u r work, o u r behavior, o u r friends, a n d o u r lovers. These appraisals are im- p o r t a n t for m a n y reasons, including t h r e e i n t e r d e p e n d e n t ones. First, evaluative feedback can be o f great e m o t i o n a l signifi- cance. Second, it can have i n s t r u m e n t a l value; for e x a m p l e , it can shape p e r f o r m a n c e a n d guide i m p o r t a n t life decisions. T h i r d , the appraisals o f o t h e r s - - o r o u r perceptions o f t h e m - - can f o r m and i n f o r m o u r sense o f self(e.g., Baldwin, 1992; Fel- son, 1992; Jussim, Soffin, Brown, Ley, & K o h l h e p p , 1992; M c N u l t y & Swarm, 1994; Mead, 1934). According to the sym- bolic interactionists, the self t h a t develops is a " l o o k i n g glass s e l f " (Cooley, 1902) f o r m e d b y o u r perceptions o f others' re- sponses t o us.

The looking glass m e t a p h o r seems to i m p l y t h a t the question Of the accuracy o f perceptions is n o n p r o b l e m a t i c ; we can sim- ply l o o k to others and see their opinion o f us reflected b a c k to us directly (Felson, 1992; Glaser & Strauss, 1964). Yet the pre- ponderance o f evidence suggests t h a t there is considerable e r r o r in o u r perceptions o f how others view us (e.g., DePaulo, Kenny, Hoover, Webb, & Oliver, 1987; K e n n y & DePaulo, 1993). O u r perceptions o f others' appraisals c o r r e s p o n d imperfectly with

Bella M. DePaulo, Department of Psychology, University of Virginia; Kathy L. Bell, Department of Psychology, Texas Tech University.

This investigation was supported in part by a Research Scientist De- velopment Award from the Alcohol, Drug Abuse, and Mental Health Administration and the National Institute of Mental Health (NIMH) and an R01 Award from NIMH. We thank Malaika Almarode, Dan Alter, Lori Cuatina, Robert Cooley, Alexandra Dahne, Irene Dalton, Laura Daniel, Naomi Driesen, Dave Gawrylowicz, Darby Gibbs, Salina Guliani, Joan Hairfield, Joe Ho, Ken Hodge, Alicia Hughes, Melinda Lantz, Traci Mann, Joanne Moak, Nikki Picerno, Lynne Robinson, Scan Robinson, Laura Rogers, Karen Swarmer, Laura Tuck, Benita Watson, Ken Watson, and Cheryl Witt for their help with this research.

Correspondence concerning this article should be addressed to BeUa M. DePauio, Department of Psychology, Gilmer Hall, University of Vir- ginia, Charlottesville, Virginia 22903. Electronic mail may be sent via the Internet to [email protected].

their actual appraisals, a n d o u r self-perceptions are m o r e closely linked to o u r perceptions o f how others view us t h a n to their actual views o f us (Felson, 1992 ).

W h e n we err in our perceptions o f how others view us, we may do so because others d i d n o t c o m m u n i c a t e their views o f us openly and honestly, or because we m i s i n t e r p r e t e d their ap- praisals. T h r e e elements are i m p o r t a n t : w h a t the evaluators t r i e d to convey (as i n d i c a t e d b y their own r e p o r t s ) , w h a t t h e y actually d i d convey (as indicated, for example, by t r a n s c r i p t s o f w h a t they s a i d ) , and how their c o m m u n i c a t i o n s were perceived (as i n d i c a t e d b y people's i m p r e s s i o n s ) . Studies o f meta-accu- r a c y typically o m i t the m i d d l e element: There is n o precise r e c o r d o f w h a t evaluators actually said, or the r e c o r d is never analyzed. In the present research we assessed all t h r e e components.

A n i m p o r t a n t reason for dishonesty in evaluative c o m m u n i - cations may be t h a t evaluators c a r e m o r e a b o u t the e m o t i o n a l i m p a c t o f their feedback t h a n its i n s t r u m e n t a l value. The feed- b a c k t h a t supervisors can provide t o floundering employees, for e x a m p l e , is potentially o f i n s t r u m e n t a l value b o t h to the em- ployees and to the organization, yet supervisors are reluctant to provide feedback to those employees a n d often delay doing so (Larson, 1989). Honesty a n d openness are highly p r i z e d char- acteristics o f friendships, yet even friends are reluctant to share their unflattering appraisals o f each other (Blumbergo 1972; Mayer, 1957). The persons directly affected by b a d news have the greatest need to know t h a t news, yet people are m o r e in- d i n e d to c o m m u n i c a t e such news to uninvolved t h i r d p a r t i e s t h a n to the targets (Felson, 1992; Tesser & Rosen, 1975).

As an individual's personal investment in an object increases, b o t h the i n s t r u m e n t a l and the e m o t i o n a l significance o f evalua- tive feedback are likely to increase as well. For example, when an a r t student is discussing paintings with other people, the a p - praisals t h a t they can provide are m o r e e m o t i o n a l l y i m p a c t f u l and also m o r e useful when the paintings are the a r t student's own work t h a n when they are the creations o f other artists. Yet we t h i n k t h a t e m o t i o n a l considerations will prevail, and evaiu- ators will be less honest a b o u t the paintings when they are the a r t student's own w o r k - - p a r t i c u l a r l y when they dislike the work---even though it would be especially useful to the a r t stu-

703

7 0 4 DEPAULO AND BELL

dents t o know how their work really is perceived b y other peo- ple. For instance, a r t students whose work is p o o r yet who never h e a r t h a t from others may pursue a m a j o r or even a career to which they a r e ill suited.

In the present research, p a r t i c i p a n t s looked over a set o f paintings, chose the two they liked the best and the two they liked the least, a n d i n d i c a t e d j u s t how m u c h they liked each o f those four paintings. T h e y also wrote o u t w h a t they liked and disliked a b o u t each painting. O n l y then d i d they learn t h a t they would be discussing those paintings with an a r t student who was personally invested in one o f the liked and one o f the disliked paintings. Those conversations were v i d e o t a p e d ( a n d later t r a n s c r i b e d ) . After each conversation, we asked the p a r t i c i - p a n t s how honest a n d how c o m f o r t a b l e t h e y h a d been and how m u c h liking they h a d t r i e d to convey. We then showed the vid- eotapes to judges who i n d i c a t e d their perceptions o f the p a r t i c - i p a n t s ' honesty and actual liking for the paintings. We p r e d i c t e d t h a t the p a r t i c i p a n t s would be m o r e dishonest and m o r e un- comfortable, and would exaggerate their liking more, when they were discussing the paintings t h a t were special to the a r t stu- d e n t - e s p e c i a l l y when they disliked those paintings.

G o f f m a n (1967, 1971 ) p r o v i d e d a perspective for under- standing people's reluctance to say exactly w h a t they feel. He argued t h a t in o r d e r for everyday social life to p r o c e e d smoothly, it is i m p o r t a n t for people to give deference to the "faces" (identities) t h a t others seem to be claiming. As politeness the- o r y has d o c u m e n t e d (P. Brown & Levinson, 1987; R. Brown & G i l m a n , 1989; Holtgraves, 1992), people understand this and act accordingly. N o one needs to tell us to be polite when dis- cussing an ugly p a i n t i n g with the artist who created it. Dis- agreements and criticisms are face-threatening and will be c o m - m u n i c a t e d only very politely, i f at a l l - - b u t even positive c o m - munications, P. Brown a n d Levinson argued, can be face- t h r e a t e n i n g (e.g.,. c o m p l i m e n t s t h a t cause e m b a r r a s s m e n t ) .

Telling people explicitly to be polite and to t r y to avoid hurt- ing another person's feelings, then, should result in c o m m u n i - cations t h a t are n o different t h a n i f no instructions h a d been given. In b o t h instances, people will dishonestly convey overly positive appraisals. To b r e a k down the s t u r d y b a r r i e r s to the c o m m u n i c a t i o n o f negative evaluations, it may be i m p o r t a n t to underscore explicitly the i m p o r t a n c e o f honesty. I n the present research, we explicitly i n s t r u c t e d some o f o u r p a r t i c i p a n t s to be honest a b o u t their appraisals. O n l y f r o m such honest evalua- tions, we said, c o u l d the a r t students really l e a r n a b o u t other people's perceptions o f art. We p r e d i c t e d t h a t these instructions would d a m p e n p a r t i c i p a n t s ' exaggerations, relative to condi- tions in which p a r t i c i p a n t s were i n s t r u c t e d to be polite o r were given n o special instructions, b u t we were unsure as to whether they would elicit evaluations t h a t were totally honest.

T h e situation we created was a very difficult one for the par- ticipants, especially when t h e y were discussing paintings they disliked with the a r t student who p a i n t e d them. Bavelas and her colleagues (Bavelas, Black, Chovil, & Mullett, 1990) character- ized this situation as the m o s t c o m m o n sort o f " c o m m u n i c a t i v e a v o i d a n c e - a v o i d a n c e conflict: [ Participants h a d ] a choice be- tween saying something false b u t k i n d and something t r u e b u t h u r t f u l " (p. 5 8 ) . O n the basis o f m o r e t h a n a dozen experi- ments, Bavelas et al. concluded that, in these situations, people

equivocate. T h e y avoid answering the question t h a t is asked, they avoid describing their own opinion, they are unclear in the answers they do give, and they sometimes even avoid addressing the person who posed the question. Bavelas et al.'s research, then, tells us what people do not say in a v o i d - a v o i d situations ( o r at least in role-play versions o f t h e m ) , b u t it stops short o f telling us w h a t they do say. Even their conclusions a b o u t w h a t people d o n o t say are based n o t on content analyses o f the c o m - m u n i c a t i o n s b u t on judges' global impressions.

We agree with Bavelas et al. (1990) t h a t people prefer to avoid telling either o u t r i g h t lies o r hurtful truths. Therefore, we pre- dict, as they did, t h a t the rate o f telling outright lies will be low. However, we t h i n k t h a t the rate o f lying, though low, will still be responsive to o u r experimental manipulations. Specifically, we predict t h a t p a r t i c i p a n t s will be m o s t likely to lie when discuss- ing paintings they dislike with a r t students who are personally invested in t h e m - - e s p e c i a l l y i f the participants had been in- structed to be polite.

In the difficult situation we created, we t h i n k that p a r t i c i - p a n t s have two goals: T h e y want to mislead the a r t student a b o u t how they feel, b u t they also want to be able to deny that they lied. T h e i r c o m m u n i c a t i o n s will be governed by w h a t we will call the defensibility postulate, t h a t is, p a r t i c i p a n t s ' inclina- tion to exaggerate their liking for the paintings and to convey dishonest appraisals o f t h e m will be t e m p e r e d by considerations o f defensibility (see also Schlenker, 1980). Participants will craft c o m m u n i c a t i o n s which, i f challenged, can be defended as either truthful or at least n o t clearly deceptive. In the context o f this experiment, we t h i n k t h a t one way they can do this is to amass misleading evidence. As the a r t student continues to p r o b e t h e m a b o u t their opinions o f the paintings, they can men- tion m o r e and m o r e o f the things t h a t they really do like a b o u t the paintings, while being a bit m o r e restrained in enumerating the aspects o f the paintings t h a t they really do dislike. The result is a c o m m u n i c a t i o n t h a t is likely to succeed in conveying a mis- leadingly positive i m p r e s s i o n yet can still be defended as t r u t h - f u l - - a f t e r all, all o f the positive aspects m e n t i o n e d were ones t h a t the p a r t i c i p a n t s really d i d like a b o u t the paintings. N o t mentioning all o f the disliked aspects, they might argue, is n o t d i s h o n e s t - - t h e y j u s t d i d n o t m e n t i o n them.

We t h i n k t h a t the participants will also come u p with entirely new aspects o f the paintings t h a t they will c l a i m to l i k e - - a s p e c t s t h a t they h a d n o t written down when we first asked t h e m to describe what they liked and disliked a b o u t the paintings. Per- haps they will tell themselves t h a t they j u s t noticed these new virtues o f the p a i n t i n g d u r i n g the conversation with the a r t stu- dent. Defensibility is especially likely to r e m a i n intact i f they also notice some new aspects o f the p a i n t i n g that they dislike. Again, though, the newly discovered disliked aspects will be far o u t n u m b e r e d by the new liked aspects.

The prediction made by Bavelas et al. (1990) that people will avoid stating their own opinion is consistent with the defensibility postulate and was directly tested by the coding o f participants' explicit expressions o f liking or disliking for the paintings. W h e n participants are discussing a painting they dislike, especially one that is special to the a r t student, they might s t o n e w a l l - - t h a t is, avoid making any explicit evaluations at all. They might also mention fewer aspects o f the paintings that they like or dislike.

LYING AND CARING 7 0 5

T h e r e is a n o t h e r very clever way t h a t p a r t i c i p a n t s c a n defen- sibly i m p l y m o r e l i k i n g t h a n t h e y really do feel for t h e p a i n t i n g s , a n d t h a t is b y m a n i p u l a t i n g w h a t t h e y say a b o u t t h e p a i n t i n g s i n w h i c h t h e a r t s t u d e n t s a r e not p e r s o n a l l y invested. T h a t is, a t t h e s a m e t i m e t h a t p a r t i c i p a n t s t r y t o avoid saying explicitly t h a t t h e y dislike t h e a r t s t u d e n t ' s o w n p a i n t i n g s t h a t t h e y detest, t h e y c a n b e far less r e t i c e n t i n v o i c i n g t h e i r distaste for the p a i n t - ings c r e a t e d b y o t h e r a r t students. T h e strategy is o n e o f social c o m p a r i s o n b y i m p l i c a t i o n . I n c o m p a r i s o n to t h e negative a p - praisals t h a t were explicitly stated a b o u t t h e o t h e r a r t i s t s ' work, t h e w i t h h o l d i n g o f a n y explicit a p p r a i s a l s o f t h e a r t s t u d e n t ' s o w n w o r k will s e e m r a t h e r positive. T h o s e c o m m u n i c a t i o n s a r e also defensibly positive: I f pressed, t h e p a r t i c i p a n t s c a n c l a i m t h a t t h e y d i d not say t h a t t h e y liked t h e a r t s t u d e n t ' s o w n work; t h e y s i m p l y said t h a t t h e y d i d n o t like t h e o t h e r artists' work.

W h e n we showed t h e v i d e o t a p e s o f t h e c o n v e r s a t i o n s to t h e judges, we gave t h e m t h e s a m e i n f o r m a t i o n t h a t t h e artists w o u l d b e likely to have i n t h e c o m p a r a b l e real life situations. T h a t is, the j u d g e s k n e w w h e t h e r t h e p a i n t i n g s were special to t h e artists, b u t t h e y d i d n o t k n o w w h a t t h e p a r t i c i p a n t s really d i d t h i n k o f t h e p a i n t i n g s . T h e y also d i d n o t k n o w t h e partici- p a n t s ' i n t e n t i o n s - - t h a t is, w h e t h e r t h e y were m a k i n g a n y spe- cial effort to b e h o n e s t o r polite.

W e p r e d i c t e d t h a t t h e j u d g e s w o u l d r e p o r t s o m e o f t h e s a m e t h i n g s t h a t t h e p a r t i c i p a n t s w o u l d say t h e m s e l v e s - - t h a t t h e p a r - t i c i p a n t s were less h o n e s t a n d less c o m f o r t a b l e , a n d exaggerated m o r e , w h e n discussing the p a i n t i n g s i n w h i c h the artists were m o r e i n v e s t e d (cf. D e P a u l o & K i r k e n d o l , 1989; D e P a u l o , La- nier, & Davis, 1983; D e P a u l o , Stone, & Lassiter, 1 9 8 5 b ) . I f t h a t were all t h a t t h e j u d g e s n o t i c e d , t h e n those results m i g h t s i m p l y reflect j u d g e s ' t h e o r i e s a b o u t how people c o m m u n i c a t e to peo- ple w h o care, r a t h e r t h a n a n y real d i s c e r n m e n t . However, b e - cause t h e j u d g e s d i d n o t k n o w w h e t h e r the p a r t i c i p a n t s liked or disliked a p a i n t i n g , i f t h e y also t h o u g h t t h a t t h e p a r t i c i p a n t s s e e m e d especially less h o n e s t w h e n d i s c u s s i n g t h e special p a i n t - ings w h e n t h e p a r t i c i p a n t s disliked t h o s e p a i n t i n g s , t h e n t h e y w o u l d b e showing s o m e i n s i g h t i n t o p a r t i c i p a n t s ' t r u e feelings.

It is i m p o r t a n t to n o t e t h a t we asked t h e j u d g e s directly j u s t how m u c h t h e y t h o u g h t t h e p a r t i c i p a n t s really d i d like each o f t h e p a i n t i n g s . I f t h e y d i s c o u n t e d t h e p a r t i c i p a n t s ' expressions o f l i k i n g t o o m u c h i n t h e special c o n d i t i o n s ( b e c a u s e t h e y k n e w t h a t t h e p a r t i c i p a n t s were t a l k i n g t o artists w h o were p e r s o n a l l y i n v e s t e d i n t h e p a i n t i n g s ) , t h e y w o u l d b e w r o n g a b o u t t h e par- t i c i p a n t s ' a c t u a l feelings (cf. G i l b e r t & M a l o n e , 1995; S n y d e r & F r a n k e l , 1976). I f i n s t e a d t h e y were t o o i n c l i n e d to take w h a t the p a r t i c i p a n t s said at face v a l u e (e.g., D e P a u l o , 1992, 1994; D e P a u l o , Stone, & Lassiter, 1985a; G i l b e r t & M a l o n e , 1995; J o n e s , 1 9 9 0 ) , t h e y w o u l d a g a i n b e wrong, b u t i n a different di- rection. T h a t is w h a t we predicted. Because we expected t h e p a r t i c i p a n t s ' verbal strategies to b e effective i n c r e a t i n g m i s l e a d - ingly positive i m p r e s s i o n s , we expected t h e j u d g e s to believe t h a t t h e p a r t i c i p a n t s really d i d like t h e special p a i n t i n g s m o r e t h a n t h e n o t - s p e c i a l ones.

M e t h o d

P a r t i c i p a n t s a n d A r t S t u d e n t s

Participants were 47 male and 47 female introductory psychology students who participated for partial fulfillment of a course requirement

in an experiment that was ostensibly about psychology and art. Five other participants were excluded: 2 men and 1 woman who surmised the purpose of the experiment, 1 man whose speech could not be un- derstood, and t woman who completed the forms improperly. Partici- pants were randomly assigned to the six between-subjects cells formed by the crossing of the two degrees of investment (paintings were de- scribed as the art student's favorites or her own) with the three kinds of instructions (honest, no instructions, polite). There were 7 or 8 men and 7 or 8 women in each of the cells.

Three women alternated in the role of the art student, and 3 women and 2 men served a~ experimenters. Preliminary analyses in which art students and experimenters were included as a factor in the design showed fewer significant effects involving the factor than would be ex- pected by chance.

P r o c e d u r e

Participants were r u n individually and were told that the experiment was designed to help art students learn more about how art is perceived by people who are not experts. Participants were then left alone in a room to choose the 2 paintings they liked the most and the 2 they liked the least from 19 paintings that were displayed. (The paintings had been painted by undergraduates in an introductory painting course.) Partic- ipants rated each of these 4 paintings on 9-point scales of liking, with higher numbers indicating greater liking. The experimenter then re- turned and gave the participant a second questionnaire on which the participant was asked to describe briefly, in an open-ended format, what he or she liked and disliked about each of the4 paintings.

The experimenter then told the participant that he or she would now discuss the four paintings with the art student. The experimenter men- tioned that the art student may have actually painted some of the paint- ings herself, and she would tell the participant if she had. The experi- menter also informed the participant that the art student would not know that the four paintings were ones that the participant selected and that she would not ever see the participant's ratings of liking for the paintings or the brief descriptions of what the participant liked and dis- liked aboui the paintings.

The art student always claimed that one of the participant's two most favorite paintings (randomly selected ) and one of the participant's least favorite paintings (also randomly selected) were special to her in some way. The two types of specialness, or degrees of investment, were ran- domly assigned. In the moderate investment condition, the art student claimed that the painting was one of her favorites ("This is one of my favorites" ); in the high investment condition, she claimed that the paint- ing was one of her own ("This is one that I d i d " ) . She introduced this information just before asking the participant what he or she thought of the painting.

Participants were randomly assigned to one of three instructional conditions. One third of them were instructed to be honest when dis- cussing the paintings with the art student. Specifically, they were told:

If it turns out that the art student did paint some of these paintings, you should still be very honest in describing your own opinions about those paintings. Tell her truthfully what you liked and what you disliked about each painting you discuss, even if the paintings are ones she painted herself. This is supposed to be a learning ex- perience for the students. For them to really learn about people's perceptions of art, they have to hear unbiased descriptions of those perceptions. They need to know what you really did like and really did dislike about each painting you discuss.

Some of the art students like to mention which ones they really liked of the ones that are NOT theirs. Again, be sure to be honest about your own opinions oftbe paintings. Tell her what you really

7 0 6 DEPAULO A N D BELL

t h i n k o f the painting, regardless o f what her opinions might be. They will learn more i f they hear your t r u e opinions.

A n o t h e r t h i r d o f the participants were instructed to be polite to the a r t student and to try not to h u r t her feelings. Specifically, they were told:

I f it turns o u t t h a t the student did paint some o f these paintings, try to convince her t h a t you really did like the ones she painted so t h a t her feelings won't be hurt. It is O K to mention things you dislike about her paintings when she asks, b u t j u s t try to convey the im- pression t h a t overall, you like the ones she did. This study is sup- posed to be a learning experience for the participants, but we d o n ' t want any of t h e m to end up feding badly because of it.

Some o f the art students like to mention which ones they liked o f the ones t h a t are NOTtheirs. Again, it is O K if you d o n ' t agree with h e r - - y o u can say that, but j u s t try to be real nice about it.

In the no-instructions condition, participants were not given any par- ticular instructions about what to do.

After determining t h a t the participant understood the instructions, the experimenter left the room, t u r n e d on a hidden video recorder, then returned with the art student. After introducing the participant to the a r t student, the experimenter left the room.

The art student, who was unaware of the participant's instructional condition ( b u t did know which paintings the participants liked and disliked), then proceeded to interview the participant a b o u t each of the four paintings, in counterbalanced order. She asked the following questions about each painting, giving the participant ample time to an- swer each question before moving on to the next: " W h a t do you t h i n k o f it? W h a t are some o f the specific things you like about it? (Anything else?) W h a t are some o f the specific things you dislike about it? (Anything else?)" Participants were instructed by the experimenter n o t to ask the art student about her opinions. The art students were trained to deflect any such questions.

After the discussion o f each painting, the a r t student left the room while the participant completed a questionnaire about the discussion. O n 9-point scales, participants indicated how m u c h liking they tried to convey to the art student, how honest and straightforward they had been, and how comfortable they felt while discussing what they liked a n d disliked a b o u t the painting. Participants were debriefed, and all o f t h e m signed a consent form allowing us to use their videotapes.

J u d g e s a n d V i d e o t a p e s

Seven male a n d 14 female undergraduates were recruited to rate vid- eotapes (with sound) o f the discussions o f the paintings. The tapes were rated by j u s t one judge at a time. N o t all o f the judges rated all o f the tapes; on the average, each tape was rated by 6 men (the range was 5 - 7 ) and 12 women ( 1 1 - 1 4 ) .

The discussions o f the paintings were edited onto 17 videotapes o f about 1 hr each. A nearly equal n u m b e r o f participants from each con- dition appeared on each tape. After each discussion o f each painting, there was a 10-s rating pause. During the pause, judges rated the partic- ipant on 9-point scales o f honesty, actual liking for the painting, degree of liking that the participant was trying to convey to the art student, a n d comfort, with higher numbers indicating more of each attribute. Reliabilities (alphas) were .78 for honesty, .94 for actual and conveyed liking, and .71 for comfort.

T r a n s c r i p t s a n d C o d e r s

Exact typed transcripts were made o f all o f the discussions o f all o f the paintings. Three undergraduates coded the transcripts. One coded

the conversations o f h a l f o f the participants in each condition ( 188 conversations), and a second coded the other half. We used these cod- ings in the analyses. The third person coded 48 o f the conversations coded by the first person and 44 o f the conversations coded by the sec- ond person. The conversations in each set included approximately equal numbers from each condition. The codings o f the third person were compared with those o f the first two to assess reliability.

Verbal S t r a t e g i e s

Each coder coded three verbal strategies separately for each discus- sion o f each painting.

1. Explicit evaluations o f liking and disliking. Coders indicated whether the participants explicitly said t h a t they liked the paintings and whether they explicitly said that they disliked them.

2. Total number o f liked and disliked aspects that were mentioned. Coders counted the total n u m b e r of different aspects o f the paintings t h a t participants said t h a t they liked and the total n u m b e r they said they disliked. For example, i f participants said they liked the color, the shading, and the originality, they would get a score o f 3 for total n u m b e r of liked aspects.

3. Number o f new liked and disliked aspects that were mentioned. Coders counted the n u m b e r of aspects of the paintings t h a t participants said they liked and disliked t h a t were different from the aspects t h a t the participants had described in writing before they knew they would be meeting an art student. To code this variable, coders first identified each o f the liked and disliked aspects that participants described in writing, then they identified the liked and disliked aspects from the transcripts of the discussions, then they compared the two sets.

R e l i a b i l i t i e s

There were two intraclass correlations for each variable: One was the correlation between the first coder and the third, and the other was be- tween the second coder and the third. For explicit evaluations of liking the reliabilities were .96 and .96; for explicit evaluations o f disliking they were .85 and 1.00; for total n u m b e r of liked aspects, .88 and .92; for total n u m b e r o f disliked aspects, .88 a n d . 8 0 ; for new liked aspects, .75 and .80; and for new disliked aspects, .71 and .59.

R e s u l t s

M a n i p u l a t i o n C h e c k s

O n a m a n i p u l a t i o n c h e c k q u e s t i o n n a i r e , all p a r t i c i p a n t s i n t h e h o n e s t c o n d i t i o n i n d i c a t e d t h a t t h e i r g o a l was t o b e h o n e s t a b o u t t h e i r feelings a b o u t t h e p a i n t i n g s . All p a r t i c i p a n t s i n t h e p o l i t e c o n d i t i o n i n d i c a t e d t h a t t h e i r g o a l was t o t r y t o b e n i c e a n d a v o i d h u r t i n g t h e a r t s t u d e n t ' s feelings. W h e n q u e s t i o n e d a b o u t t h e i r u n d e r s t a n d i n g o f t h e i n s t r u c t i o n s j u s t b e f o r e b e g i n - n i n g t h e d i s c u s s i o n o f t h e p a i n t i n g s , all p a r t i c i p a n t s c o r r e c t l y r e p o r t e d t h a t t h e a r t s t u d e n t m a y h a v e p a i n t e d s o m e o f t h e p a i n t i n g s ( o r t h a t s o m e w e r e t h e a r t s t u d e n t ' s f a v o r i t e s ) a n d t h a t t h e y w o u l d n o t k n o w w h e t h e r t h e a r t s t u d e n t h a d p a i n t e d a n y o f t h e p a i n t i n g s u n t i l t h e y m e t her. N i n e t y - t w o o f t h e 9 4 p a r t i c - i p a n t s c o r r e c t l y i n d i c a t e d t h a t t h e y w o u l d b e d i s c u s s i n g t h e f o u r p a i n t i n g s t h e y h a d s e l e c t e d . ( B e c a u s e t h e 2 p a r t i c i p a n t s w h o i n i t i a l l y v o l u n t e e r e d t h e w r o n g a n s w e r c o r r e c t e d i t a f t e r f u r t h e r p r o b i n g , a n d b e c a u s e all o f t h e i r o t h e r m a n i p u l a t i o n c h e c k d a t a w e r e c o r r e c t , t h e i r d a t a w e r e r e t a i n e d i n t h e a n a l y s e s . ) All o f t h e p a r t i c i p a n t s u n d e r s t o o d t h a t t h e a r t s t u d e n t s w o u l d b e l i e v e t h a t t h e p a i n t i n g s p i c k e d f o r d i s c u s s i o n w e r e s e l e c t e d a t r a n d o m b y

LYING AND CARING

Table I

Effects o f Participants" Liking for the Paintings and Artists ' Investment on Participants" and Judges" Ratings

7 0 7

Type of painting

Disliked Liked Fs (1, 82)

Not Not Artists' Participants' Ratings special Special Difference special Special Difference investment liking Interaction

Participants Comfort 6.47 5 . 0 5 1.42"*** 6.88 6.61 0.27 60.19"*** 58.69**** 19.52"*** Honesty 7.66 6.38 1.28"*** 7.99 7.98 0.01 48.50**** 67.68**** 47.07**** Actual liking 2.30 2.34 0.04 7.35 7.35 0.00 0.03 1392.96**** 0.04 Conveyed liking 3.18 3.96 0.78** 6.44 6.79 0.35 26.89**** 176.90"*** 2.90* Exaggeration 0.88 1 . 6 2 0.74*** -0.91 -0.56 0.35 19.94"*** 74.85**** 1.64

Judges Comfort 5.85 5.75 0.10"* 6.08 6.06 0.02 4.62** 73.70**** 3.00* Honesty 6.56 6.22 0.34**** 6.51 6.37 0.14"* 46.94**** 1.17 6.65*** Actual liking 3.67 3.84 0.17* 5.91 5.98 0.07 4.85** 766.96**** 1.11 Conveyed liking 4.14 4.65 0.51"*** 6.41 6.70 0.29*** 45.19**** 670.52**** 4.24** Exaggeration 0.47 0 . 8 1 0.34**** 0.50 0.72 0.22**** 84.72**** 1.02 3.40*

Note. The special paintings were the ones in which the artists were invested. Difference is not special minus special for comfort and honesty, and special minus not special for actual liking, conveyed liking, and exaggeration. MSEs for artists' investment were, from top to bottom, 2.22, 1.60, 0.82, 2.27, 1.42, 0.26, 0.23, 0.60, 0.66, and 0.08. For participants' liking they were 3.11, 2.59, 1.70, 9.82, 4.93, 0.37, 0.41, 1.18, 1.30, and 0.12. For the interaction they were 3.14, 1.59, 0.77, 2.82, 2.07, 0.24, 0.28, 0.47, 0.54, and 0.09. * p < . 1 0 . * * p < . 0 5 . ***p<.01. ****p<.001.

the experimenter. All p a r t i c i p a n t s also u n d e r s t o o d t h a t the a r t student would n o t see w h a t t h e y h a d written a b o u t the paintings.

Analyses o f p a r t i c i p a n t s ' initial ratings o f their liking for the paintings i n d i c a t e d t h a t they liked their two favorite paintings far m o r e t h a n their two least favorite ones ( M s -- 7.35 a n d 2.32). Because the paintings t h a t were described as special to the a r t student were r a n d o m l y assigned, p a r t i c i p a n t s should n o t have liked t h e m any better t h a n the ones t h a t were n o t special, and in fact they d i d n o t ( M s = 4.82 a n d 4.84; see Table 1 for significance tests).

D e s i g n a n d M e a s u r e s

D a t a were analyzed with a mixed-design analysis o f variance ( A N O V A ) . The between-subjects factors were the instructional m a n i p u l a t i o n ( p a r t i c i p a n t s were t o l d t o be honest o r polite o r t h e y were given no i n s t r u c t i o n s ) and m i s t s ' degree o f invest- m e n t in the special paintings (those paintings were described as either the artists' f a v o r i t e s - - t h e m o d e r a t e i n v e s t m e n t condi- tion, o r as their own w o r k - - t h e high investment c o n d i t i o n ) . T h e within-subjects factors were the artists' i n v e s t m e n t in the paintings ( t h e y were invested in the special paintings and n o t invested in the not-special ones) a n d the p a r t i c i p a n t s ' liking for the paintings (disliking or liking), i

P a r t i c i p a n t s ' reports o f how honest a n d straightforward they h a d been in their discussions were highly correlated, r ( 9 2 ) = .86, p < .001, and so they were averaged to f o r m a single m e a - sure o f honesty. T h e m e a s u r e o f p a r t i c i p a n t s ' actual liking for the paintings was their ratings o f their liking for each o f the paintings before they knew t h a t they would be meeting an a r t student. T h e i r reports o f how m u c h liking they h a d tried to con-

vey to the a r t student were collected after the discussions. We assessed the degree to which the p a r t i c i p a n t s h a d exaggerated their liking b y subtracting p a r t i c i p a n t s ' actual liking for each p a i n t i n g from the degree o f liking t h a t they t r i e d to convey. The other dependent m e a s u r e was p a r t i c i p a n t s ' ratings o f their com- fort d u r i n g each o f the discussions. Similarly, for the analyses o f the judges' impressions, d e p e n d e n t m e a s u r e s were judges' per- ceptions o f the p a r t i c i p a n t s ' honesty, comfort, actual liking for the paintings, a n d degree o f liking t h a t they seemed to be t r y i n g t o convey. We c o m p u t e d exaggeration scores b y subtracting per- ceptions o f actual liking f r o m perceptions o f conveyed liking.

I Sex of participant also was included as a factor in the design, but the results of that factor are not of central relevance to the theme of the present report and therefore are not included. They are currently avail- able from BcUa M. DcPaulo and will be reported in a subsequent article that will include several studies in addition to the data from this re- search (Witt, Bell, & DcPaulo, 1996). The significant effects for partic- ipant sex that did occur in the present research generally indicated that the overall effects were characteristic of both the men and the women, but they were even more characteristic of the women. For example, the judges believed that both the men and the women were trying to convey more liking for the special paintings than for the not-special ones, but they saw a bigger difference for the women than for the men. The degree- of-investment factor was included to test whether our predictions for investment would be qualified by degree of investment. Although those results will not be presented, significant interactions did occur for par- ticipants' self-reports and judges' impressions. In all instances, the in- teractions indicated that the effects of investment were even stronger when the art students were highly invested in the paintings (the paint- ings were their own work) than when they were moderately invested in them (the paintings wcrc their favorites). Complete results are available from Bclla M. DcPaulo.

7 0 8 DEPAULO AND BELL

Participants" Self-Reports a n d Judges" I m p r e s s i o n s

Paintings that were liked and disliked, special and not special As we predicted, the m a i n effect of investment was significant for all relevant dependent measures (see Table 1 for statistical tests a n d significance levels). W h e n the paintings were special to the a r t students (second and fifth c o l u m n s of Table 1 ), com- pared to when they were n o t (first a n d fourth c o l u m n s ) , the participants reported being more uncomfortable a n d more dis- honest. They also tried to convey more liking, a n d they exagger- ated their liking more. Similarly, all m a i n effects of liking for the painting were significant. Participants said they were less comfortable and less honest when discussing the paintings they disliked than the ones they liked. They tried to convey more liking for the paintings they liked, b u t they exaggerated their liking more for the paintings they disliked (i.e., they tried to convey more liking t h a n they really did feel). In fact, according to their self-reports, participants actually understated their liking for the liked paintings. Also as predicted, the effects of the artists' investment on participants' honesty a n d comfort depended significantly on whether the participants liked the paintings. Participants were significantly less honest and less comfortable when discussing the special paintings than the not- special ones only when they disliked the paintings.

The judges also thought that the participants were more u n - comfortable and dishonest when discussing the special paint- ings than the not-special ones a n d that they tried to convey more liking, and more exaggerated liking, for the special paint- ings. The differences in honesty and conveyed liking that they noted were even more striking when the participants disliked the paintings t h a n when they liked them.

There was a significant m a i n effect of investment, b u t n o sig- nificant interaction with liking for the painting, on judges' im- pressions of participants' actual liking for the paintings. Partic- ipants liked the special paintings almost exactly the same as the not-special ones. The judges did n o t know this, and from watch- ing the tapes, their impression was that the participants really did like the special paintings even more than the not-special ones. 2

There was one other way in which the judges' impressions departed from the participants' self-reports. The participants said they exaggerated their liking for the disliked paintings b u t understated their liking for the liked paintings. The judges thought that the participants were always exaggerating their lik- ing (especially so for the special paintings). Tests of whether the exaggeration (or understatement) scores differed from zero were significant for all four paintings for the participants' self- reports (all p s < .05 or smaller) and the judges' impressions (all p s < .001 ).

Finally, although the judges were n o t told whether the partic- ipants liked the paintings, their impressions of the discussions o f the liked and disliked paintings were accurate. They thought the participants really did like the liked paintings more, and were trying to convey more liking for them, and that they felt less comfortable discussing the disliked paintings.

Honesty and politeness. Did the participants who were in- structed to be honest or to be polite behave and feel differently than those who were left to their own devices? Significant m a i n

effects of the instructional m a n i p u l a t i o n for the measures of honesty, F ( 2 , 82 ) = 8.42, p < .001, M S E = 6.92, and exaggera- tion, F ( 2 , 82) = 4.66, p = .01, M S E = 4.89, indicated that they did. The means for self-reported honesty in the honest, no- instructions, and polite conditions, were 7.71, 7.85, and 6.95, respectively. The difference between the honest and the no-in- structions conditions was n o t significant. The difference be- tween the no-instructions and the polite conditions was signifi- cant, F ( 1, 82) = 14.31, p < .001. In their reports of their own honesty, then, participants given n o special instructions were more similar to the participants instructed to be honest than to those who were urged to be polite. The judges' impressions of the participants' honesty showed the same thing, F ( 2 , 82) = 3.03,p = .05, M S E = 0.85. The judges thought that the partici- pants were no less honest in the no-instructions condition ( M = 6.49) than in the honest condition ( M = 6.46, F < 1 ), b u t they thought the participants were significantly less honest in the po- lite condition ( M = 6.30) than in the no-instructions condition, F ( I , 82) = 5.33,p = .02.

However, in the degree to which they reported exaggerating their liking for the paintings, participants in the no-instructions condition were more similar to participants who were told to be polite. (The u n i n s t r u c t e d participants did not differ signifi- cantly from the polite participants [ F < 1], but they did differ significantly from the honest participants, F [ 1, 82 ] = 4.30, p = .04.) In fact, participants in both the no-instructions and the polite conditions said that they tried to convey more liking than they really did feel (Ms = 0.37 and 0.61 for the no-instructions and polite conditions, respectively), b u t participants in the

2 We thought that if our judges had instead been completely unaware of the most important constraint in the present research--when the participants were and were not talking to artists who cared--they might have been even more taken by participants' expressions of liking. To test this, we prepared exact typed transcripts of the four conversations of 8 of the participants in the no-instructions and polite conditions who were talking to the artists about paintings that were or were not the artists' own. We recruited 65 raters (32 men and 33 women) to report their impressions of how much the participants really did like the paintings in each conversation, on the same 9-point scale used by our judges. Approximately half of the raters (n = 33 ) rated the conversations with the same information that our judges had--that is, they knew when the artists claimed that the paintings were their own. For the other raters (randomly assigned), that critical information was removed from the transcripts. The key interaction between whether the paintings were or were not special, and whether the judges knew that they were special, was significant, F( 1, 61 ) = 33.54, p < .001, MSE = 0.94. When the paintings were not special, raters perceived almost exactly the same amount of liking when they knew that they were not special (M = 4.75 ) as when they did not know that (M = 4.79). However, when the paint- ings were special and the raters knew that they were, they thought that the participants liked those paintings much less (M = 4.66) than when they did not have that information (M = 5.40). That is, raters dis- counted some of the liking that participants expressed when they knew that the participants were talking to an artist who cared. The implica- tion for understanding the ratings made by our judges, who did know when the paintings were special, is that they may have (inaccurately) perceived even greater differences in liking between the special and not- special paintings if they had not had that crucial information.

LYING AND CARING

Table 2 Effects oflnstructions and Artists" Investment on Participants" and Judges'Ratings

7 0 9

Instructions

Honest No instructions Polite

N o t Not Not Ratings Fs(1, 82) special Special Difference special Special Difference special Special Difference

Participants Honesty 7.42**** 7.86 7.56 0.30 8.09 7.59 0.50** 7.52 6.39 1.13"*** Conveyed liking 6.44*** 4.73 4.97 0.24 5.07 5.42 0.35 4.62 5.73 1. l l**** Exaggeration 8.58**** - 0 . 2 5 - 0 . 1 7 0.08 0.21 0.53 0.32*** - 0 . 0 2 1.23 1.25"***

Judges Actual liking 2.96* 4.86 4.84 - 0 . 0 2 4.77 4.86 0.09 4.73 5.04 0.3 l*** Conveyed liking 6.30*** 5.36 5.50 0.14 5.23 5.64 0.41"*** 5.24 5.88 0.64**** Exaggeration 3.72** 0.50 0.66 0 . 1 6 " * * 0.46 0.78 0.32**** 0.51 0.84 0.33****

Note. Difference is not special m i n u s special for honesty, special minus not special for conveyed liking, exaggeration, and actual liking. * p < . 1 0 . * * p < . 0 5 . * * * p ~ . 0 1 . * * * * p < . 0 0 1 .

honest condition said t h a t they conveyed slightly less liking t h a n they felt ( M = - 0 . 2 1 ) .

D i d the i n s t r u c t i o n a l m a n i p u l a t i o n influence the way the par- ticipants discussed the paintings t h a t were or were n o t special to the a r t students? Significant interactions between the instruc- tional m a n i p u l a t i o n and the investment variable for the m e a - sures o f honesty, conveyed liking, and exaggeration, i n d i c a t e d t h a t it did. As shown in Table 2, in all three i n s t r u c t i o n a l condi- tions, p a r t i c i p a n t s said t h a t they were less honest when the art- ists were invested in the paintings t h a n when they were not, and they also said t h a t they t r i e d t o convey m o r e liking and t h a t they exaggerated their liking m o r e when the artists were invested. The degree to which they showed these effects, however, in- creased f r o m the honest t o the no-instructions to the polite con- dition. (See the c o l u m n s in Table 2 labeled Difference. ) For the exaggeration measure, for e x a m p l e , the degree to which partic- i p a n t s exaggerated their liking m o r e for the special t h a n for the not-special paintings was only 0.08 ( a n d n o t significant) in the honest condition; it increased to 0.32 in the no-instructions condition and to 1.25 in the polite condition. I n fact, for all t h r e e measures, p a r t i c i p a n t s in the no-instructions condition were m o r e similar to the p a r t i c i p a n t s in the honest condition t h a n they were to the p a r t i c i p a n t s in the polite condition. Con- trast analyses showed t h a t the difference between the special paintings and the not-special ones was the same for the honest condition and the n o - i n s t r u c t i o n s condition for all three mea- sures ( F s < 1 ); b u t the special versus not-special difference was significantly greater in the polite condition t h a n in the no-in- structions condition for all measures (all p s = .007 or smaller).

T h e judges also thought t h a t the ways t h a t the p a r t i c i p a n t s h a n d l e d the discussions o f the special ( c o m p a r e d to the not- special) paintings were influenced b y their a t t e m p t s to be honest or polite. T h e judges thought t h a t the p a r t i c i p a n t s t r i e d t o convey m o r e liking and m o r e exaggerated liking for the spe- cial paintings t h a n for the not-special ones ( a n d they tended t o t h i n k t h a t the p a r t i c i p a n t s really d i d like the special paintings more, which they d i d n o t ) , a n d t h e y also noticed t h a t the degree to which p a r t i c i p a n t s t r i e d to favor the special paintings in-

creased f r o m the honest to the n o - i n s t r u c t i o n s to the polite con- dition (see Table 2).

T h e way t h a t the u n i n s t r u c t e d p a r t i c i p a n t s c o m p a r e d to the others was different for the judges' ratings t h a n for the partici- p a n t s ' own reports. I n the self-report data, the degree to which the p a r t i c i p a n t s favored the special over the not-special p a i n t - ings was essentially the same for the p a r t i c i p a n t s who were t o l d to be honest as for those who were left to their own d e v i c e s - - c o n t r a r y to o u r predictions. The judges, in contrast, thought t h a t u n i n s t r u c t e d p a r t i c i p a n t s were no different from the polite p a r t i c i p a n t s in the degree to which they favored the special paintings. ( T h e F s were < I for exaggeration, and 2.52 a n d 2.51, b o t h p s = . 12, for actual and conveyed liking, respectively.) The judges also thought t h a t the u n i n s t r u c t e d p a r t i c i p a n t s were different from the honest p a r t i c i p a n t s in the degree to which they favored the special paintings; for conveyed liking, F ( 1, 82) = 3.46, p = .07, and for exaggeration, F ( 1, 82) = 4 . 8 2 , p = .03.

Finally, the i n s t r u c t i o n a l m a n i p u l a t i o n was especially i m p o r - t a n t to the way the p a r t i c i p a n t s dealt with the artists' investment when the paintings were ones the p a r t i c i p a n t s disliked. The three-way interaction o f instructions, investment, and liking for the p a i n t i n g was significant for p a r t i c i p a n t s ' reports o f their honesty, F ( 2 , 82 ) = 12.23, p < .001, M S E = 1.59. As shown in Table 3, when p a r t i c i p a n t s liked the paintings (see last three c o l u m n s o f the t a b l e ) , the instructions they received h a d virtu- ally n o effect in any o f the conditions on how honest they were a b o u t the special c o m p a r e d to the not-special paintings. How- ever, when p a r t i c i p a n t s disliked the paintings (first three col- u m n s o f Table 3), they a d m i t t e d to being less honest a b o u t the ones t h a t were special to the artists c o m p a r e d to the ones t h a t were not. T h i s difference was significant in every instructional condition, b u t it increased from the honest ( M = 0.50) to the n o - i n s t r u c t i o n s ( M = 0.94 ) to the polite condition ( M = 2.38 ). O n c e again, the u n i n s t r u c t e d participants, in their self-reports, were m o r e similar to the p a r t i c i p a n t s who were t o l d to be honest t h a n to those who were t o l d to be polite. The difference between special and n o t special was n o t significantly greater in the u n i n - s t r u c t e d condition t h a n in the honest condition, F ( l , 82) =

7 1 0 DEPAULO AND BELL

Table 3 Effects o f Instructions, Participants' L i k i n g f o r the Paintings, and Artists' Investment on Participants" and Judges" Ratings

Type of painting

Disliked Liked

Ratings and Not Not instructions special Special Difference special Special Difference

Participants: Honesty Honest 7.58 7.08 0.50** 8.14 8.04 0.10 No instructions 7.90 6.96 0.94**** 8.28 8.22 0.06 Polite 7.48 5.10 2.38**** 7.55 7.67 -0.12

Judges Honesty,

Honest 6.52 6.39 0.13 6.54 6.37 0.17* No instructions 6.67 6 . 3 1 0.36**** 6.58 6.42 0.16* Polite 6.49 5.97 0.52**** 6.42 6.33 0.09

Conveyed liking Honest 4.27 4.30 0.03 6.46 6.71 0.25* No instructions 4.01 4.59 0.58**** 6.45 6.69 0.24* Polite 4.15 5.07 0.92**** 6.32 6.70 0.38***

Note. The special paintings were the ones in which the m i s t s were invested. Difference is not special minus special for honesty and special minus not special for conveyed liking. * p < . 1 0 . * * p < . 0 5 . ***p<.01. ****p~.001.

1.89, p = . 17, b u t it was significantly greater in the polite condi- tion t h a n in the u n i n s t r u c t e d condition, F ( 1, 82) = 20.25, p < .00 I.

The judges also n o t i c e d t h a t it was especially difficult for the p a r t i c i p a n t s to discuss the special paintings t r u t h f u l l y when the p a r t i c i p a n t s disliked the paintings b u t were trying to be polite a b o u t them, F ( 2 , 82) = 3.27, p = .04, M S E = 0.28. As shown in Table 3, the degree to which the p a r t i c i p a n t s seemed to be m o r e dishonest when discussing the disliked paintings t h a t were special to the artists ( c o m p a r e d to the disliked paintings t h a t were n o t special) increased from the honest to the n o - i n s t r u c - tions to the polite condition. The s a m e p a t t e r n o c c u r r e d for judges' perceptions o f the liking t h a t p a r t i c i p a n t s seemed to be trying to convey, F ( 2 , 82) = 4.74, p = .01, M S E = 0.54. W h e n the p a r t i c i p a n t s disliked the paintings, the judges thought t h a t they seemed to be trying to convey especially m o r e liking for the special t h a n for the not-special paintings and t h a t this tendency increased from the honest to the n o - i n s t r u c t i o n s to the polite condition. Again, the judges, in c o n t r a s t to the participants, thought t h a t the u n i n s t r u c t e d p a r t i c i p a n t s were m o r e similar to the p a r t i c i p a n t s who were told to be polite t h a n t o the p a r t i c i - p a n t s t o l d to be honest. F o r perceptions o f honesty, the special versus not-special difference for the disliked paintings was n o t significantly smaller in the n o - i n s t r u c t i o n s condition t h a n in the polite condition, F ( 1, 82) = 1.57, ns, b u t it was nearly signifi- cantly greater in the n o - i n s t r u c t i o n s condition t h a n in the honest condition, F ( 1, 82) = 3.25, p = .07. For the measure o f conveyed liking, the corresponding values were F ( 1 , 82) = 3 . 0 3 , p = .08, a n d F ( 1, 82) = 7 . 9 3 , p = .006.

Verbal S t r a t e g i e s

Design. The design for the analyses o f p a r t i c i p a n t s ' verbal strategies was the s a m e as for the p a r t i c i p a n t s ' self-reports and

the judges' ratings, except t h a t one within-subjects factor o f p r o - fessed affect (liking--disliking) was added. For the measure o f explicit evaluation, the levels were ( a ) whether the participants explicitly said t h a t they liked the p a i n t i n g and ( b ) whether they explicitly said t h a t they disliked it (see Cochran, 1950, and Ro- senthal & Rosnow, 1991, for the use o f ANOVA with dichoto- m o u s dependent variables). For total aspects m e n t i o n e d a n d for new aspects mentioned, the levels were n u m b e r o f liked aspects m e n t i o n e d and n u m b e r o f disliked aspects mentioned.

Initial likes and dislikes. To be sure t h a t participants d i d n o t like m o r e aspects o f the special paintings t h a n the not-spe- cial ones even before they met the a r t students, we analyzed the n u m b e r o f liked and disliked aspects o f all o f the paintings t h a t p a r t i c i p a n t s h a d described in writing. The interaction o f invest- m e n t with n u m b e r o f liked versus disliked aspects was n o t sig- nificant ( F < 1 ). Thus, p a r t i c i p a n t s began by listing a l m o s t exactly the same n u m b e r o f likes a n d dislikes for the special paintings as for the not-special ones.

Professed affect. The m a i n effect o f professed affect was sig- nificant for all t h r e e measures. Across the discussions o f all o f the paintings, the p a r t i c i p a n t s were almost twice as likely to say t h a t they liked a p a i n t i n g ( M = 0.44) t h a n to say t h a t they dis- liked it ( M = 0.24), even though all p a r t i c i p a n t s actually liked the exact same n u m b e r o f paintings t h a t they disliked, F ( l, 82) = 56.09, p < .001, M S E = 0.13. T h e y also mentioned m a n y m o r e things t h a t they liked t h a n disliked a b o u t the paintings ( M s = 5.26 and 3.61), F ( 1 , 82) = 66.49, p < .001, M S E = 7.64, and o f the aspects o f the paintings t h a t they m e n t i o n e d b u t had n o t originally listed, significantly m o r e o f t h e m were aspects t h a t they liked t h a n disliked ( M s = 3.26 and 2.04), F ( 1, 82) = 4 3 . 0 1 , p < .001, M S E = 6.47.

L i k e d and disliked paintings. Participants had m o r e diffi-

LYING AND CARING 71 1

culty c o m m u n i c a t i n g t r u t h f u l l y a b o u t the paintings t h a t they disliked t h a n a b o u t the ones t h a t they liked. The interactions between professed affect a n d liking for the paintings were sig- nificant for all t h r e e measures. As shown in Table 4, when par- ticipants liked a painting, they said so 81% o f the time; however, when they disliked a painting, they said so explicitly only 48% o f the time, F ( 1, 82) = 353.70, p < .001, M S E = 0.19. W h e n p a r t i c i p a n t s liked a painting, t h e y m e n t i o n e d m a n y m o r e things a b o u t it t h a t they liked t h a n t h a t they disliked, b u t when they disliked a painting, t h e y m e n t i o n e d fewer t h a n one m o r e thing a b o u t it t h a t they disliked t h a n liked, F ( 1, 82) = 217.25, p < .001, M S E = 5.05. Similarly, when discussing a p a i n t i n g t h a t they liked, p a r t i c i p a n t s m e n t i o n e d 4.18 a d d i t i o n a l things a b o u t it t h a t t h e y liked t h a t they h a d n o t a l r e a d y listed, c o m p a r e d to only 1.33 new things t h a t they disliked; in contrast, the n u m b e r o f new liked a n d disliked aspects t h a t p a r t i c i p a n t s generated when the p a i n t i n g was disliked h a r d l y differed (2.34 and 2.75), F ( I , 82) = 114.39,p < .O01,MSE = 4.33.

Stonewalling was i n d i c a t e d b y the m a i n effect for liking for the p a i n t i n g for the measures o f explicit evaluation, F ( 1, 82) = 30.37, p < .001, M S E = 0.09, and total n u m b e r o f aspects mentioned, F ( 1, 82) = 7.87, p = .007, M S E = 4.09. These re- sults showed t h a t p a r t i c i p a n t s n o t only h a d a h a r d t i m e telling the t r u t h a b o u t the disliked paintings, b u t they also h a d a h a r d t i m e saying anything at all. W h e n p a r t i c i p a n t s disliked a p a i n t - ing, they were less likely to m a k e any explicit evaluation (whether positive or negative) t h a n when they liked it ( M s = 0.28 and 0.41 for disliked and liked paintings, respectively). Participants also m e n t i o n e d fewer things t h a t t h e y liked o r dis- liked when they disliked a p a i n t i n g ( M = 4.23) t h a n when they liked it ( M = 4.64).

Not-special and special paintings. Professed affect in- teracted significantly with investment, and in the p r e d i c t e d di- rection, for all t h r e e measures. As shown in Table 5, when the paintings were special to the a r t students ( c o m p a r e d to when they were n o t ) , the p a r t i c i p a n t s were relatively m o r e likely to say t h a t they liked t h e m a n d relatively less likely to say t h a t they disliked them, F ( 1 , 82) = 7.83, p = .006, M S E = 0.13. Sim-

Table 4 Participants" Verbal Strategies Used in Discussing the Disliked and Liked Paintings

Participants' liking for the paintings

Verbal strategy Disliked Liked Difference

Explicit evaluation a Liked 0.08 0.81 0.73**** Disliked 0.48 0.00 -0.48****

Total aspects mentioned Liked 3.84 6.67 2.83**** Disliked 4.62 2 . 6 1 -2.01"***

New aspects mentioned Liked 2.34 4.18 1.84**** Disliked 2.75 !.33 - 1.42****

a Proportion of participants who explicitly said that liked the paintings. **** p -< .001.

they liked or dis-

Table 5 Participants" Verbal Strategies Used in Discussing the Not-Special and Special Paintings

Artists' investment

Verbal strategy Not special Special Difference

Explicit evaluation a Liked 0.41" 0.48 0.07** Disliked 0.28 0 . 2 1 -0.07**

Total aspects mentioned Liked 5.02 5.49 0.47** Disliked 3.92 3.30 -0.62***

New aspects mentioned Liked 3.09 3.42 0.33* Disliked 2.20 1.88 -0.32

Note. The special paintings were the ones in which the m i s t s were invested. a Proportion of participants who explicitly said that they liked or dis- liked the paintings. * p < . 1 0 . * * p < . 0 5 . ***p<.01.

ilarly, when the paintings were special, c o m p a r e d to when they were not, the p a r t i c i p a n t s m e n t i o n e d relatively m o r e things t h a t they liked a b o u t t h e m a n d relatively fewer things t h a t they dis- liked, F ( 1 , 82) = 13.14, p < .001, M S E = 4.22. Similarly, the p a r t i c i p a n t s thought o f relatively m o r e new things to like a b o u t the special paintings t h a n a b o u t the not-special ones, and rela- tively fewer things to dislike, F ( 1, 82) = 5.70, p < .05, M S E = 3.55. Thus, the ways in which the p a r t i c i p a n t s discussed the special versus the not-special paintings paralleled the ways they discussed the liked versus disliked paintings. It is i m p o r t a n t t o n o t e t h a t the interactions o f professed affect with investment were n o t qualified b y p a r t i c i p a n t s ' liking for the p a i n t i n g (except for one higher order interaction involving instructions, discussed n e x t ) . T h a t means, for example, t h a t p a r t i c i p a n t s used their strategy o f mentioning m a n y m o r e liked t h a n disliked aspects ( o f the special p a i n t i n g s ) j u s t as m u c h when discussing the paintings they disliked as the paintings t h e y liked.

Honesty and politeness. A m a i n effect o f instructions on ex- plicit evaluations i n d i c a t e d t h a t the m o r e polite the p a r t i c i p a n t s were i n s t r u c t e d to be, the less likely they were to offer any ex- plicit evaluation at all, F ( 2 , 82) = 3.30, p < .05, M S E = 0.10. The means for the honest, no-instructions, a n d polite condi- tions were 0.38, 0.35, and 0.30, respectively. The instructional m a n i p u l a t i o n also m o d e r a t e d the way the p a r t i c i p a n t s explicitly evaluated the paintings t h a t were or were n o t special to the a r t students. There was a significant interaction among the instruc- tional m a n i p u l a t i o n , investment, liking for the painting, and professed affect, F ( 2 , 82) = 4.84, p = .01 (see Table 6 ) . I f par- ticipants were being completely honest, then they would explic- itly say t h a t they disliked the paintings t h a t t h e y actually d i d dislike and t h a t t h e y liked the paintings t h a t they actually d i d like. T h a t is, the n u m b e r s in the m i d d l e two c o l u m n s o f Table 6 would all be exactly 1.00. But n o n e o f t h e m were. T h e n u m b e r s were fairly high for the liked paintings; when p a r t i c i p a n t s really d i d like the paintings, between 69% a n d 91% o f t h e m explicitly said t h a t they did. A n d virtually n o n e o f t h e m ever said t h a t

7 1 2 DEPAULO AND BELL

Table 6 Effects o f Instructions, Participants" Liking for the Paintings, and Artists" Investment on Participants" Explicit Evaluations

Disliked paintings Liked paintings

Professed Professed Professed Professed Instructions liking disliking liking disliking

Honest Not special .09 .56 .81 .03 Special .03 .62 .91 .00 Difference - . 0 6 .06 .10 - . 0 3

No instructions Not special .00 .64 .83 .00 Special .16 .40 .79 .00 Difference .16** -.24**** - . 0 4 .00

Polite Not special .03 .47 .69 .00 Special .16 .22 .81 .00 Difference .13* -.25**** .12* .00

Note. The special paintings were the ones in which the artists were invested. Entries are proportions of participants who explicitly said that they liked or disliked the paintings. * p < . 1 0 . * * p ~ . 0 5 . ****p~.001.

t h e y disliked a n y o f those paintings. Neither the artists' invest- m e n t in the paintings n o r the instructions the p a r t i c i p a n t s h a d been given m a d e m u c h o f a difference. But when p a r t i c i p a n t s disliked the paintings, t h e y often r e f r a i n e d from saying so ex- plicitly, a n d b o t h the instructions a n d the artists' i n v e s t m e n t m a t t e r e d t o them.

As shown in Table 6, p a r t i c i p a n t s strayed farthest from the t r u t h when they disliked a p a i n t i n g t h a t was special to the artist and they were trying to be polite a b o u t it. W h e n i n s t r u c t e d to be honest, 62% o f the p a r t i c i p a n t s explicitly acknowledged that t h e y disliked the p a i n t i n g t h a t was special to the artist; a m o n g the u n i n s t r u c t e d p a r t i c i p a n t s , only 4(~o d i d so, a n d a m o n g those p a r t i c i p a n t s urged to be polite, only 22% d i d so. (All differences among these t h r e e n u m b e r s were significant [ p s < .01 o r smaller] .) In the n o - i n s t r u c t i o n s a n d polite conditions, 16% o f the p a r t i c i p a n t s told o u t r i g h t lies: T h e y explicitly said t h a t they liked the p a i n t i n g t h a t they h a d a l r e a d y i n d i c a t e d in writing t h a t they hated. ( I n the honest condition, 3% o f the p a r - ticipants d i d this.)

It is also informative to c o m p a r e the relative percentages o f p a r t i c i p a n t s who explicitly said t h a t they liked and disliked the detested special paintings in each condition. I n the honest con- dition, 59% m o r e o f the p a r t i c i p a n t s said t h a t t h e y disliked t h a n liked the p a i n t i n g t h a t t h e y d i d in fact dislike. I n the n o - i n s t r u c - tions condition the difference was 24%, and in the polite condi- tion it was only 6%.

P a r t i c i p a n t s ' explicit evaluations o f the disliked paintings t h a t were not special to the artists showed t h a t p a r t i c i p a n t s used the p r e d i c t e d strategy o f evaluation b y implication. Participants were s o m e w h a t less likely t o say explicitly t h a t t h e y disliked the disliked paintings when they were i n s t r u c t e d t o b e polite t h a n when they were i n s t r u c t e d to be honest, b u t this d r o p from honest to polite was far less p r e c i p i t o u s when the paintings were n o t special to the artist (56 to 47) t h a n when they were special

(62 to 2 2 ) . The converse o c c u r r e d for explicit statements o f l i k i n g - - t h e o u t r i g h t lies. Participants in the polite condition ( a n d the no-instructions c o n d i t i o n ) told polite lies a b o u t the paintings they disliked t h a t were special to the a r t student: 16% o f t h e m said t h a t they liked those paintings, c o m p a r e d t o 3% in the honest condition. In contrast, when the disliked paintings were not special to the artist, 3% o f the participants in the polite condition ( a n d n o n e in the no-instructions condition ) explicitly said t h a t they liked them, c o m p a r e d to 9% in the honest condition.

In sum, then, when p a r t i c i p a n t s were t r a p p e d in the challeng- ing situation o f trying t o b e polite a b o u t work they disliked t h a t was special to the artists with w h o m they were interacting, they m a n i p u l a t e d b o t h their evaluation o f the work in which the a r t - ists were invested and their evaluation o f the other artwork in which the artists had no investment. T h e y refrained from say- ing explicitly t h a t t h e y disliked the paintings t h a t were special to the artists. A t the same time, they were much less r e s t r a i n e d when it c a m e to c o n d e m n i n g the paintings t h a t were n o t special. 3

D i s c u s s i o n

A L o o k i n g Glass or a Reversible Figure?

Decades o f research relevant to the reflected appraisal p r o - cess have indicated t h a t o u r perceptions o f others' views o f us are n o t strongly related to their actual views and t h a t o u r self- perceptions a r e m o r e highly related to o u r perceptions o f o t h - ers' appraisals t h a n to t h e i r actual appraisals (Felson, 1992). We began with two possible explanations for the p o o r fit be- tween actual and perceived appraisals. First, people may n o t be open and honest in c o m m u n i c a t i n g their appraisals. Second, we may misperceive those appraisals.

The present study strongly s u p p o r t e d the first explanation. T h e p a r t i c i p a n t s r e f r a i n e d from saying how they really d i d feel a b o u t the paintings, especially when t h e y disliked them. By it- self, this finding is h a r d l y new. F r o m the literatures on perfor- mance appraisals (Fisher, 1979; Larson, 1984, 1986, 1989), the M U M effect (Tesser & Rosen, 1975 ), and lying in everyday life ( D e P a u l o , Kashy, Kirkendol, Wyer, & Epstein, 1996; see also Folkes, 1982), as well as the literatures t h a t followed m o r e di- rectly from the s y m b o l i c interactionist t r a d i t i o n (Blumberg,

3 An example of a truthful answer to the question "What do you think of it?" was given by a participant discussing a disliked painting that was one of the artist's favorites: "It's ugly. It's just ugly." An example of a truthful message about a liked painting that was one of the artist's fa- vorites was: "I liked it. This was, this was my second favorite of the group. Urn, it was the, the detail that was put into, uh, some of the, you know, the, the nuances in color, the way the black is done. And urn, and it was, uh, yeah, I really liked it overall." An example of an answer that was coded as a lie (i.e., the participant claimed to like a disliked painting) was given by a participant discussing a painting that was the artist's own work: "I like this one:' All participants had more to say about each painting when asked additional questions, but these were their complete answers to the artist's first question ("What do you think of it?"). Over the entire course of the discussion of each painting, par- ticipants spoke an average of 217 words.

LYING AND CARING 7 1 3

1972; Felson, 1992; Swarm, Stein-Seroussi, & McNulty, 1992), we already knew that there are formidable barriers to the direct communication o f appraisals--especially negative ones---to the persons they concern. W h a t our work has shown that is n e w is (a) the powerful impact on appraisals o f the target person's personal investment in the object o f the appraisals, ( b ) the difficulty o f eliciting totally honest evaluations, and (c) the value o f the defensibility postulate in predicting the verbal strat- egies people use in dodging the truth. Perhaps even more im- portant, our findings suggest that ( d ) when we look to others for their appraisals, what we see is neither a looking glass nor a hopelessly distorted image, but a reversible figure.

Truth and investment. A m o n g the m a n y motives that have been postulated t o account for the reluctance to convey nega- tive evaluations, concern with the target person's feelings is per- haps the one that is most consistently cited and supported. I f the target person's feelings are most important, then the more the target person cares about the object o f evaluation, the less likely that person should be to hear a truthful appraisal, espe- cially when the truth would hurt. This is so, we predicted, even though the instrumental value o f honest appraisals should also increase with the target person's personal investment. O u r study was the first to manipulate target persons' personal investment, and our findings were strongly supportive o f our predictions. Across virtually every measure, communications were more dishonest when the target persons cared about the objects o f the appraisals than when they did not.

I s the truth ever told to those who care? We predicted that, when left to their own devices, people are practitioners o f po- liteness (P. Brown & Levinson, 1987). We are the first t o test people's strategic use o f polite dissembling by directly instruct- ing some o f the participants to behave as politeness theory pre- dicks they would and then comparing their behavior to that o f participants given no special instructions. We thought that the uninstructed participants would convey appraisals o f the paint- ings the art students cared about that were just as distorted and dishonest as those conveyed by the participants who were ex- plicitly instructed to be polite and avoid hurting the art stu- dent's feelings. According to the judges' perceptions, this is what usually happened. The participants' reports, however, were of- ten at odds with our prediction. The self-reports o f the unin- structed participants were usually more similar to those o f the participants urged to be honest than to those urged t o be polite.

We are inclined to trust the judges' perceptions. The partici- pants in the uninstructed condition may have been motivated t o describe themselves as honest. (The polite participants, in contrast, had an excuse for being dishonest--they were follow- ing instructions.) The judges had no investment in perceiving the participants as either honest or dishonest, and they made their ratings without any awareness o f the participants' instruc- tional conditions. We also trust the judges' perceptions more because they were more in line with the results o f our objective measures o f what the participants actually said. The results were clearest for our measure o f outright lies about the disliked special paintings. The percentage o f participants who explicitly said that they liked the paintings that they had already told us that they detested was identical in the no-instructions and the polite conditions ( 16%); in the honest condition it was lower

(3%). The same pattern is evident in the percentage o f partici- pants who refrained from saying explicitly that they disliked the detested paintings when they were special compared to when they were n o t special. In the honest condition, this withholding o f an explicit negative evaluation was equally likely for the spe- cial as for the not-special paintings, b u t in the uninstructed and the polite conditions participants were significantly more likely to refrain from saying that they disliked the detested painting when it was special than when it was not. Furthermore, the mag- nitude o f this difference between the special and not-special paintings was virtually identical for the uninstructed and polite participants (Table 6).

In that the participants who were urged to be honest told vir- tually n o lies about the special paintings they disliked, and were no more likely to withhold their explicit negative evaluations o f the special paintings than o f the not-special ones, were they evenhanded in their discussions o f the special and not-special paintings in every other way, too? I f so, that would indicate that there is an easy way to elicit totally honest feedback--urge oth- ers to tell the truth and give them a compelling reason for doing so (e.g., it is only by hearing totally honest reactions that art students can learn how others really do perceive particular paintings). According to the participants' self-reports, they usu- ally were evenhanded. The one important exception occurred when they were describing the paintings they disliked; in that condition they admitted that they were significantly less honest about the special paintings than the not-special ones (Table 3). The judges, too, thought that the honest participants were usu- ally just as honest when discussing the special paintings as they were when discussing the not-special ones. But again, there was an important exception. The judges thought that the honest par- ticipants exaggerated their liking more when they were discuss- ing the special paintings than the not-special ones (see Table 2).

Another condition in which it may have been possible for all participants to be just as honest about the special paintings as the not-special ones was when they liked the paintings. Accord- ins to their self-reports, participants were in fact evenhanded in their discussions o f the special and not-special paintings when they liked the paintings. O n n o measure did they report signifi- cantly less truthfulness. The judges, however, did think there were some differences. For example, they thought that the par- ticipants were trying to convey significantly more liking, and that they were exaggerating their liking more, for the paintings they liked that were special to the artists than for the liked paint- ings that were not special (see Table 1 ). The objective measures o f what the participants really did say lend support to the judges' views. The strategy o f amassing positive evidence pref- erentially for the special paintings (relative to the not-special ones) was just as evident when the participants liked the paint- ings as when they disliked them. In this study, then, there was essentially n o condition under which art students who cared about the paintings heard totally honest feedback about them.

The defensibility postulate. When participants give a paint- ing one o f the lowest possible ratings on the liking scale and then tell the art student that they like that painting, it is hard for them to defend that statement as truthful. For that reason, our defensibility postulate predicted that outright lies would occur infrequently, as in fact they did. But they also occurred exactly

7 1 4 DEPAULO AND BELL

when we expected them t o - - w h e n the participants were dis- cussing paintings they disliked that were special to the art stu- dent, and especially when they were given n o special instruc- tions or were instructed to be polite.

The strategy o f amassing misleadingly positive evidence when one's true opinion is negative is one that was noted in passing nearly four decades ago in a study o f the self-restraint o f friends (Mayer, 1957). In the present context, participants practiced this strategy by mentioning m a n y aspects o f the special paint- ings that they really did like while mentioning relatively few as- pects that they actually disliked. The resulting communications are highly defensible in that the positive qualities that the par- ticipants mentioned were ones that they really did like. Al- though participants were n o t equally forthcoming about the qualities they disliked, they did n o t deny disliking those quali- ties (which would not be defensible); they simply refrained from mentioning them.

The mentioning o f new positive qualities that were not ini- tially listed is a riskier strategy, but one that perhaps can work if participants convince themselves that they really do like these newly discovered aspects that they simply had not noticed pre- viously. Credibility is added if the participants also notice some new aspects o f the paintings that they dislike, though our pre- diction is that they will discover fewer o f these new disliked as- pects than liked aspects when the paintings are special. The re- suits supported that prediction, too.

We believe that future research will show that the strategy o f amassing misleading evidence is widely used. Most objects o f e v a l u a t i o n - - f o r example, personalities, appearances, j o b per- formances, paintings, and journal articles--are complex stim- uli that routinely elicit both positive and negative reactions. It is a fairly simple matter, then, when put on the spot to voice one's opinion, to reel off one positive c o m m e n t after another.

Also as predicted by the defensibility postulate, participants used the very clever strategy o f evaluation by implication. By explicitly stating their disliking for the paintings created by other artists, while refraining from stating their disliking for the art student's own work, they implied a favorable social compar- ison. They appeared to like the art student's own work more than the other artists' work. They never exactly said that, how- ever, so their communications can be defended as truthful.

I n contexts in which it is possible for evaluators simply to avoid communicating their appraisals, we think that they will often do just that. I n the performance appraisal literature, for example, it has been noted that supervisors sometimes delay giving negative feedback (Larson, 1989). Even when complete avoidance is no longer possible, evaluators still manage to con- vey less than the whole truth. For example, both supervisors (Larson, 1986) and football coaches (Felson, 1981 ) hedge by conveying specific appraisals rather than global ones. I n the present research, we found that participants stonewalled by offering fewer explicitly evaluative comments and mentioning fewer aspects o f the paintings that they liked or disliked when they were discussing paintings they disliked than ones they liked.

The reversible figure. O u r results suggest an unanticipated answer to the question o f why our perceptions o f others' ap- praisals are n o t strongly related to others' actual appraisals: Par-

ticipants described the paintings in ways that allowed the art students a choice as to what to hear and what to believe.

When discussing disliked paintings that the artists cared about, participants exaggerated their liking, withheld explicit expressions o f disliking, and even told some outright lies. This gave the art students the opportunity to think that the partici- pants really did like those paintings. But the participants also dropped some blatant hints as to their relative degrees o f liking for the different paintings that the artists cared about. For ex- ample, they did not even try to convey as much liking for the paintings they disliked as for the ones they liked. They rated the disliked special paintings a 2 on the 9-point scale at the begin- ning o f the study, and they tried to convey a rating o f 4 to the art students. Although this was substantially higher than the lik- ing they really did feel, it was still significantly lower than the degree o f liking they tried to convey for the special paintings they really did l i k e - - a 7.

By biasing their appraisals o f the special paintings in a posi- tive w a y - - b y mentioning relatively more things that they liked about them and relatively fewer things that they disliked about them, relative to the not-special paintings--the participants again handed the artists the option o f believing that they really did like their paintings. Still, the ratio o f liked to disliked aspects that the participants communicated was not as lopsided as when they were describing paintings that they really did like.

Participants' explicit evaluations also allowed for interpretive flexibility. Sixteen percent o f the uninstructed and polite partic- ipants explicitly said that they liked the special paintings that they actually hated. However, that n u m b e r was dramatically lower than the 80% o f the uninstructed and polite participants who explicitly said that they liked the special paintings that they really did like.

One o f the most important qualifications o f the symbolic in- teractionist model that has emerged from research is that the link from metaperception to self-perception is not unidirec- tional. Although it is true, as the symbolic interactionists have long maintained, that our perceptions o f how others view us can influence our self-concept, it is also true that our self-concept can influence our beliefs about how others view us (Felson, 1992; Jussim et al., 1992; McNulty & Swarm, 1994). The pres- ent research shows how the latter effect might occur. The feed- back that people receive, even in the very difficult situations like the ones we created in this research, is unlikely to be totally dishonest and univalent. Instead, it is complex and multifac- eted, offering plausible evidence for very different inter- pretations. This leaves lots o f r o o m for self-perceptions to in- fluence the interpretation that is selected.

Qualifications

O u r results are qualified in three ways. First, we recruited only women as artists, to keep the size o f the study manageable. What was not known then, but is known now, is that people tell more lies to protect the other person's feelings when they are talking to women than to men (DePaulo et al., 1996 ). I f we had included male artists, the overall rate o f lying would probably have been lower. But we think that the key relationship between

LYING AND CARING 715

truth and investment would have remained unaffected. Still, that is a question for future research.

Second, we argued that participants were being less than truthful when they were more effusive about the special paint- ings than the not-special ones, after indicating in their initial ratings of the paintings that they liked the special a n d not-spe- cial paintings just the same. However, it is possible that over the course o f the conversations, participants changed their minds about the special paintings and really did come to like them more. However, even if the process were one of genuine attitude change rather t h a n strategic dissembling, the consequences for the artists r e m a i n the same: They heard evaluations of the paintings they cared about that were strikingly more positive t h a n they would have heard if the paintings were n o t special to them. They could n o t count on gleaning equally glowing ap- praisals from evaluators who did n o t know, or did n o t care, whether the paintings were special to them.

In fact, although we have argued that lies are told to those who care, we must also acknowledge that sometimes brutal truths are told instead. I n our paradigm, this did n o t occur. Per- haps participants would have told the truth if they cared more about the long-term consequences for the artists of hearing mis- leading feedback, rather t h a n the short-term consequences of hurting the artists' feelings and feeling badly for having done so. Some evaluators are in roles that d e m a n d that they pay atten- tion to the long-term consequences. Pre-med advisors, for ex- ample, are duty b o u n d to warn their low-achieving advisees that their career plans may be unrealistic. Still, we suspect they would often find a way to do so politely. Perhaps evaluators will be harshly critical when they are intellectually insecure (Amabile, 1983). But even this effect, we think, is likely to oc- cur only when the evaluators are either protected by a n o n y m i t y or when they do n o t need to deliver their feedback to the person they are criticizing in a face-to-face interaction. We t h i n k that naked truths will be told when the conventions of polite society have n o t yet been fully internalized (as in the story of the em- peror's new clothes), or when they have been temporarily aban- doned, as when adults are caught in the throes of anger or ha- tred. Other possibilities are better left to research t h a n to speculation.

R e f e r e n c e s

Amabile, T. M. ( 1983 ). Brilliant but cruel: Perceptions of negative eval- uators. Journal of Experimental Social Psychology, 19, 146-156.

Baldwin, M. W. (1992). Relational schemas and the processing of social information. Psychological Bulletin, I 12, 461-484.

Bavelas, J. B., Black, A., Chovil, N., & Mullett, J. (1990). Equivocal communication. Newbury Park, CA: Sage.

Blumberg, H. H. (1972). Communication of interpersonal evaluations. Journal o f Personality and Social Psychology, 23, 157-162.

Brown, P., & Levinson, S. (1987). Politeness: Some universals in lan- guage usage. Cambridge, England: Cambridge University Press.

Brown, R., & Gilman, A. (1989). Politeness theory and Shakespeare's four major tragedies. Language in Society, 18, 159-212.

Cochran, W. G. (1950). The comparison of percentages in matched samples. Biometrika, 37, 256-266.

Cooley, C. H. (1902). Human nature and the social order. New York: Scribner's.

DePaulo, B. M. (1992). Nonverbal behavior and self-presentation. Psy- chological Bulletin, 111, 203-243.

DePaulo, B. M. (1994). Spotting lies: Can humans learn to do better? Current Directions in Psychological Science, 3, 83-86.

DePaulo, B. M., Kashy, D. A., Kirkendol, S. E., Wyer, M. M., &Epstein, J. E. (1996). Lying in everyday life. Journal o f Personafity and Social Psychology, 70, 979-995.

DePaulo, B. M., Kenny, D. A., Hoover, C., Webb, W., & Oliver, P. (1987). Accuracy of person perception: Do people know what kinds of impressions they convey? Journal o f Personality and Social Psy- chology, 52, 303-315.

DePaulo, B. M., & Kirkendol, S. E. (1989). The motivational impair- ment effect in the communication of deception. In J. Yuille (Ed.), Credibility assessment (pp. 51-70). Norwell, MA: Kluwer Academic.

DePaulo, B. M., Lanier, K., & Davis, T. (1983). Detecting the deceit of the motivated liar. Journal o f Personality and Social Psychology, 45, 1096-1103.

DePaulo, B. M., Stone, J. I., & Lassiter, G. D. (1985a). Deceiving and detecting deceit. In B. R. Schlenker (Ed.), The self and social life (pp. 323-370). New York: McGraw-Hill.

DePaulo, B. M., Stone, J. I., & Lassiter, G. D. (1985b). Telling ingrati- ating lies: Effects of target sex and target attractiveness on verbal and nonverbal deceptive success. Journal o f Personality and Social Psy- chology, 48, 1191-1203.

Felson, R. B. (1981). Self andreflected appraisals among football play- ers: A test of the Meadian hypothesis. Social Psychology Quarterly, 44, 116-126.

Felson, R. B. (1992). Coming to see ourselves: Social sources of self- appraisals. In E. J. Lawler, B. Markovsky, C. Ridgeway, & H. A. Walker (Eds.), Advances in group processes (Vol. 9, pp. 185-205). Greenwich, CT: JAI Press.

Fisher, C. D. (1979). Transmission of positive and negative feedback to subordinates: A laboratory investigation. Journal o f Applied Psychol- ogy, 64, 533-540.

Folkes, V. S. (1982). Communicating the reasons for social rejection. Journal o f Experimental Social Psycholog~, 18, 235-252.

Gilbert, D. T., & Malone, P. S. (1995). The correspondence bias. Psy- chological Bulletin, 117, 21-38.

Giaser, B. G., & Strauss, A. L. (1964). Awareness contexts and social interaction. American Sociological Revie~, 29, 669-679.

Goffman, E. ( 1967 ). Interaction ritual: Essays on face-to-face behavior. Garden City, NJ: Anchor.

Goffman, E. ( 1971 ). Relations in public. New York: Harper & Row. Holtgraves, T. (1992). The linguistic realization of face management:

Implications for language production and comprehension, person perception, and cross-cultural communication. Social Psychology Quarterly, 55, 141-159.

Jones, E. E. (1990). Interpersonalperception. New York: Freeman. Jussim, L., Sot~n, S., Brown, R., Ley, J., & Kohlhepp, K. (1992). Un-

derstanding reactions to feedback by integrating ideas from symbolic interactionism and cognitive evaluation theory. Journal o f Personal- ity and Social Psychology, 62, 402-421.

Kenny, D. A., & DePaulo, B. M. (1993). Do we know how others view us? An empirical and theoretical account. Psychological Bulletin, 114, 145-161.

Larson, J. R., Jr. (1984). The performance feedback process: A prelim- inary model. Organizational Behavior and Human Performance, 33, 42-76.

Larson, J. R., Jr. (1986). Supervisors' performance feedback to subor- dinates: The impact of subordinate performance valence and out- come dependence. Organizational Behavior and Human Decision Processes, 37, 391-408.

716 DEPAULO AND BELL

Larson, J. R., Jr. (i989). The dynamic interplay between employees' feedback-seeking strategies and supervisors' delivery of performance feedback. Academy of Management Review, 14, 408-422.

Mayer, J. E. (1957). The self-restraint of friends: A mechanism in fam- ily transition. Social Forces, 35, 230-238.

McNulty, S. E., &Swann, W. B., Jr. (1994). Identity negotiation in roommate relationships: The self as architect and consequence of so- cial reality. Journal of Personality and Social Psychology, 67, 1012- 1023.

Mead, G. H. (1934). Mind, self, and society Chicago: University of Chicago Press.

Rosenthal, R., & Rosnow, R. L. ( 1991 ). Essentials of behavioral re- search (2nd ed.). New York: McGraw-Hill.

Schlenker, B. R. 1980). Impression management. Monterey, CA: Brooks/Cole.

Snyder, M. L., & Frankel, A. (1976). Observer bias: A stringent test of behavior engulfing the field. Journal of Personality and Social Psy- chology, 34, 857-864.

Swann, W. B., Jr., Stein-Seroussi, A., & McNnlty, S. E. (1992). Outcasts in a white lie society: The enigmatic worlds of people with negative self-conceptions. Journal of Personality and Social Psychology, 62, 618-624.

Tesser, A., & Rosen, S. ( 1975 ). The reluctance to transmit bad news. In L. Berkowitz (Ed.), Advances in experimental social psychology ( Vol. 8, pp. 193-232). New York: Academic Press.

Witt, C. L., Bell, K. L., & DePaulo, B. M. (1996). Sex differences in deceptiveness: Real or perceived? Unpublished manuscript.

Received December 1, 1994 Revision received April 8, 1996

Accepted April 16, 1996 •

New Editors Appointed, 1998-2003

The Publications and Communications Board o f the American Psychological Association announces the appointment o f five new editors for 6-year terms beginning in 1998.

As o f January 1, t997, manuscripts should be directed as follows:

For the Journal o f Experimental Psychology: Animal Behavior Processes, submit manuscripts to Mark E. Bouton, PhD, Department o f Psychology, University o f Vermont, Burlington, V T 05405-0134.

For the Journal o f Family Psychology, submit manuscripts to Ross D. Parke, PhD, Department o f Psychology and Center for Family Studies-075, 1419 Life Sciences, University o f California, Riverside, C A 92521-0426.

For the Personality Processes and Individual Differences section o f the Journal o f Personality and Social Psychology, submit manuscripts to Ed Diener, PhD, Department o f Psychology, University o f IUinois, 603 East Daniel, Champaign, I L 61820.

For Psychological Assessment, submit manuscripts to Stephen N. Haynes, PhD, Department o f Psychology, University o f Hawaii, 2430 C a m p u s Road, Honolulu, H I 96822.

For Psychology andAging, submit manuscripts to Leah L. Light, PhD, Pitzer College, 1050 North Mills Avenue, Claremont, C A 91711-6110.

Manuscript submission patterns make the precise date o f completion o f the 1997 volumes uncertain. Current editors, Stewart H. Hulse, PhD; Ronald F. Levant, EdD; Russell G. Geen, PhD; James N. Butcher, PhD; and Timothy A. Salthouse, PhD, respectively, will receive and consider manuscripts until December 31, 1996. Should 1997 volumes be completed before that date, manuscripts will be redirected to the new editors for consider- ation in 1998 volumes.