What do people think when they observe an unknown person s behavior? Social psychology research has investigated this question and found that people implicitly and unintentionally infer the person s traits, motives, and situational circumstances from his/her behavior（Hassin et al. 2005; Lupfer et al. 1990; Winter & Uleman 1984）. This phenomenon is called spontaneous
inference （see Uleman et al. 1996）. In the p r e s e n t s t u d y , w e e x a m i n e d w h e t h e r s p o n t a n e o u s t r a i t i n f e r e n c e s o c c u r v i a spontaneous motive inferences.
Past research has focused on detecting various kinds of spontaneous inferences but has not focused on the relationship between motive, trait, and situational inferences, mainly
Practice & Discussion
Can Spontaneous Trait Inferences be Based on
Spontaneous Motive Inferences?
TANIGUCHI Yuri and YAGI Yasuki
（Graduate School of Literature and Human Sciences, Osaka City University / College of Comprehensive Psychology, Ritsumeikan University）
In the present study, we examined whether spontaneous motive inferences are formed based on the situational circumstance in which an actor behaves, and if they are influenced by spontaneous trait inferences. Fifty-three Japanese undergraduates participated in the study, and they were randomly assigned to three between-subjects conditions. One third of the participants were presented with target helping behavior in the free-choice situation, and another one third were presented the behavior in the no-choice situation, which they were asked to memorize. The rest of the participants were presented neither of these behaviors. After the behavior was presented, we measured spontaneous motive and trait inferences about the actor using the Go/No-Go Association Task（GNAT）. Results revealed that when the perceivers observed the actor voluntarily helping a person, they spontaneously inferred motives based on situational circumstances, and spontaneously inferred traits based on the inferred motives. In contrast, the perceivers spontaneously inferred neither motives nor traits when they observed that the actor helped a person involuntarily. These findings suggested that even under implicit processes, perceivers are able to determine whether the behavior is voluntary or involuntary, based on the situational circumstance. Only if the target acts on his/her own will would spontaneous motive inference be formed based on the situational circumstance, and the spontaneous trait inferences emerging from inferred motives would occur.
Key Words : Spontaneous Inferences, Go/No-Go Association Task, Situational Circumstance,
Motive, Trait
because, to date, it has been supposed that people do not have sufficient cognitive capacity to integrate different types of inferences（e.g., Gilbert & Malone 1995; Gilbert 1989）. In addition, they considered that when perceivers observe a target person s behavior, owing to implicit inferences, they infer a trait from the behavioral information and do not consider the situational circumstance which the target behavior occurred because the vividness of the perception of the behavior is stronger than that of situational circumstances（Gilbert & Malone 1995）. However, we contend that although implicit inferences occur, trait inferences take into account not only a person s behavior but also his/her motive or situational circumstances in which the behavior occurs. This is because u n d e r s t a n d i n g a n a c t o r s m o t i v e s i s fundamental to a person s perception. Previous studies argued that trait inferences are adjusted for situational forces, that is, when we see a target hitting another person, we do not infer violence from his behavior if they are fighting in a game of boxing（e.g., Gilbert 1989; Heider 1958）. In sum, trait inferences are discounted by situational forces, because, situational information is a useful clue for understanding the target s motive. Even if we saw the same two behaviors, the meaning of these behaviors differs depending on the situation which these behaviors occurred. We learn the relationship between the meaning of behavior and the motives of behavior since birth. Thus, to understand the target person, we consider the target s motive based on our past experiences（Trope & Gaunt 1999）.
Reeder, Kumar, Hesson-McInnis, & Trafimow （2002）showed that an inference of a target s
motive mediates the relationship between a situational force and a trait inference under explicit processes. In addition, research on impression formation, that is, explicit processes, has also argued the importance of an actor s motive in person perception. Reeder, Vonk, Ronk, Ham, and Lawrence（2004）explored the functions of an actor s motive. They revealed that even if perceivers observed the same behavior, they were likely to infer different motives about an actor based on whether the actor behaved voluntarily or is forced to behave in a particular way, and whether the behavior benefitted the actor. They also found that inferences about different motives induced inferences concerning different traits of the actor. Thus, the perceived motive of an actor significantly influences the evaluation of the behavior.
Regarding implicit processes, the speed of occurrence of motive and trait inferences differs. Van Overwalle, Van Duynslaeger, Coomans, and Timmermans（2012）compared the speed of motive and trait inference occurrences using the recognition probe paradigm（McKoon & Ratcliff 1986）. In their experiment, participants were exposed to behavior descriptions which implied either motive or trait, or both. Immediately after exposure to the descriptions, the participants were asked to indicate whether there were i m p l i e d m o t i v e o r t r a i t w o r d s i n t h e descriptions. As a result, when they were exposed to descriptions that implied motive or both motive and trait, they falsely recognized the words of the implied motive as a part of the descriptions within at least 650 ms after the word appeared. In other words, perceivers
formed spontaneous motive inferences from behavior descriptions, and when implied motive words appeared, they re-activated motive inferences within at least 650ms, spontaneously. With reference to trait inferences, however, only when they were exposed to descriptions that implied both motive and trait did they falsely recognize the words of implied trait as being a part of the descriptions at least after 1000 ms of the appearance of the words. In sum, for the occurrence of spontaneous trait inferences from behavior descriptions, re-activation of spontaneous trait inferences requires 1000 ms or more after the implied trait word appears, and motive inferences are necessary for trait inferences. This means that the speed of reaching a high level of re-activation is faster for motive inferences than for trait inferences. These findings suggest that trait inferences are formed not only from behavioral information but also from knowing the actor s motive, even during implicit processing.
Considering the aforementioned results, spontaneous motive inferences would be formed from behavioral information and prompt spontaneous trait inferences. If, like explicit processing, situational information is taken into consideration in information processing of the target s behavior in implicit processing, the target s motives are inferred from situational circumstances spontaneously, and then the spontaneous motive inferences serve as a cue about spontaneous trait inferences. In other words, it is possible that a motive is inferred spontaneously based on the situational circumstance in which the target b e h a v i o r o c c u r r e d , a n d a l s o i n d u c e s a
spontaneous trait inference corresponding to the inferred motive.
To shed light on this possibility, the present study tested the hypothesis that a spontaneous motive inference is formed based not only on an actor s behavior but also on the situation in w h i c h t h e a c t o r b e h a v e s , a n d t h a t a spontaneous trait inference is formed based on the inferred motive.
Detecting spontaneous inferences
In the present study, we used the Go/No-Go Association Task（GNAT）to detect both spontaneous motive and trait inferences（Nosek & Banaji 2001）. The GNAT measures which attribution is associated with a specific target category. In the GNAT, the participant is required to discriminate stimulus words that belong to either a target category or attribution from a distracter simultaneously. The GNAT is based on the assumption that if a target category and attribution are associated, it w o u l d b e e a s y f o r t h e p a r t i c i p a n t s t o discriminate stimulus words that belong to e i t h e r t a r g e t c a t e g o r y o r a t t r i b u t i o n simultaneously. Spontaneous inferences are considered as processing either the inferred trait or motive that is attributed to a target person（Uleman et al. 1996）. In sum, when perceivers form motive or trait inferences spontaneously, they easily discriminate stimulus words that belong to either the target person or inferred motive, or the trait in the GNAT. In the present study, the target person served as the target category, and inferred motives and traits were the applied target attributions. Accordingly, participants were presented with a behavior in one of different
situational circumstances that implied both motive and trait, and were asked to memorize the behavior. Subsequently, spontaneous inferences were investigated using the GNAT.
I. Method
1. Participants
Fifty-three Japanese undergraduates from Ritsumeikan University participated in the study. They received 500 yen as a reward for participation in the experiment. Participants were randomly assigned to three between-subjects conditions.
2. Materials
Target behavior was defined as a target s helping behavior for a person who was cleaning up a room. The target person was identified as X. X was placed in two situations. One was a free-choice situation, and the content was as follows: X and another person played a table game with a referee. After the game, X looked at the referee who was cleaning up the room and said Can I help you ? The second was a no-choice situation, and the content was as follows: X and another person played a table game. The game rule was that the person who lost the game had to help the referee to clean up the room. X lost the game. Both target behaviors were captured on videotape.
In a pilot study（ = 12）, participants were presented with the target behavior and asked to rate to what extent the behavior fitted each o f t h e 1 7 m o t i v e s a n d 2 0 t r a i t s i n t h e questionnaire on a 5-point scale（1 = not at all, 5 = extremely）. As a result, to help and kindness were judged as the implied motive
a n d t r a i t f o r t h e f r e e - c h o i c e s i t u a t i o n , respectively, whereas to conform to the rule and obedience were judged as the implied motive and trait for the no-choice situation, respectively.
During the GNAT, we used terms connoting to help and to conform to the rule as target motive attributes and terms connoting kindness and obedience as target trait attributes. We also used the terms connoting opposite motive and trait categories as distracters. In each category, 8 terms were included. In addition, we used 8 terms pertaining to the target person as a target category and 8 terms pertaining to a non-target person s information as distracters.
3. Procedure
A t t h e b e g i n n i n g o f t h e e x p e r i m e n t , participants were told that they would join a study examining cognitions about a person s behavior. In addition, they were informed as follows:（1）Their task was to observe a target s behavior and to memorize them:（2） T h e a m o u n t o f t i m e r e q u i r e d f o r t h e experiment was about 30 minutes:（3）This experiment was conducted by a participant s voluntary agreement, and they have a right to stop this experiment. There is no disadvantages for them in case that they stop the experiment: （4）Because this study was reported using
their data analyzed statistically, someone did not specify an individual person. When participants agree to a participation in the experiment, the experiment was conducted.
First, participants were presented with information about the target person and asked to memorize it for the purpose of a later task.
The information included what the target had eaten on the previous day, handedness, the manner of going to school, and so on. This information was not indicative of either the person s traits or his/her motives. Second, one third of the participants watched X s behaviors in a free-choice situation, and another one third watched X s behaviors in a no-choice situation. They were asked to memorize them. We did not request them to form an impression of X or think about X s motives. Thus, any inferences about X s traits or motives can be assumed to have occurred spontaneously. The rest of the participants in the control condition observed neither of these behaviors.
Along with the presentation of target b e h a v i o r , s p o n t a n e o u s i n f e r e n c e s w e r e measured using the GNAT. The GNAT consisted of four blocks. In two blocks, Target Person X and inferred motive（ to help or to conform to the rule ）were the the target category and attribute, respectively. This measured the strength of associations between Target Person X and each inferred motive. In the remaining blocks, Target Person X and the inferred trait（ kindness or obedience ） were the target category and attribute to measure the strength of associations between Target Person X and each inferred trait. Each block consisted of 84 trials. The first 24 trials were practice trials and the remaining 60 trials were the main trials. The target person s picture and a target attribute label appeared and remained on the screen in the upper left and right quadrants, as reminders of the target category and attribute for that block. The trials began with the appearance of a single stimulus item from one of the four categories（target
category, non-target category, target attribute, and opposite attribute）. Participants were instructed to either press the space bar as quickly as possible for items belonging to either of the labeled categories, or do nothing for items that did not belong to them. Feedback on performance accuracy was provided for 100 ms, when the participant pressed the space bar, or the response deadline was reached, w h i c h e v e r o c c u r r e d f i r s t . A f t e r t h e presentation of the feedback and a 150 ms inter-stimulus interval, a subsequent trial began. Based on Nosek & Banaji（2001）, the response deadline was 700 ms in the two blocks in which the target attributes were inferred motives and 666 ms in the two blocks in which the target attributes were inferred traits. Van Overwalle et al.（2011）indicated that forming of spontaneous trait inferences requires at least 1000 ms after the presentation of the trait word. However, in the GNAT, 1000 ms is too long for the measurement of implicit inferences（Nosek & Banaji, 2001）. Thus, in the experiment, we followed the procedure used in the study by Nosek & Banaji（2001）. After these task, they were debriefed the true aim of the experiment which was to examine implicit impression formation, were explained the structure of GNAT, and were apologized for telling a false aim.
II. Results
Based on the signal detection theory, the rates of correct response to stimulus words （hits）and the rates of false response to
distracters（false alarms）were each converted to z-scores. The difference between the z-score
values for hits and false alarms was d′ （Table 1）. Higher d′ values represent a
stronger association between a target person and an attribution. In other words, higher d′ values indicate a stronger formation of s p o n t a n e o u s i n f e r e n c e s . T h e d a t a o f 7 participants were excluded from the analysis because some of their d′ values were 0 or less.
Our hypothesis was that different motive inferences are formed spontaneously depending on situational circumstances, and these induce different spontaneous trait inferences. In other words, participants would infer a motive about to help spontaneously from the free-choice situation, and in turn, infer a trait about kindness spontaneously, based on the inferred motive. They would make a stronger motive inference about to help and trait inference about kindness than in the other two situations, i.e., in the no-choice situation and control condition. Meanwhile, participants would infer a motive about to conform to the rule spontaneously from the no-choice
situation, and in turn, infer a trait about obedience spontaneously, based on the inferred motive. They would make a stronger motive inference about to conform to the rule and trait inference about obedience than in the other two situations, i.e., in the free-choice situation and control condition. Using d′ values, planned contrasts confirmed our hypothesis. The contrasts showed that, as predicted, the strength of an association between target person, an inferred motive about to help, and the strength of an association between target person and an inferred trait about kindness were significantly greater in the free-choice situation（coded +1）than in the other two situations（in each, no-choice coded -1/2; control coded -1/2; （42）= 2.27 and 1.60, respectively; = .01 and .06, respectively）.
Next, the contrasts showed that neither the strength of an association between target person and the inferred motive about to conform to the rule, nor the strength of the association between target person and an
Table 1 Means of d′ values.
Measure Free-choice （N = 16） No-choice （N = 15） Control （N = 15） Motive
Target person "X" + to help
1.27 0.92 0.90
0.60 0.49 0.41
Target person "X" + To conform to the rule
1.84 1.74 1.59
0.58 0.72 0.66
Trait
Target person "X" + kindness
1.87 1.47 1.53
0.81 0.58 0.79
Target person "X" + obedience
1.39 1.38 1.28
0.82 0.63 0.64
i n f e r r e d t r a i t a b o u t o b e d i e n c e w e r e significantly different between the no-choice condition（coded +1）and the other two situations（in each, free-choice coded -1/2; control coded -1/2; （42）= 0.10 and 0.24, respectively; = .46 and .40, respectively）.
The results indicated that target person and to help were more strongly associated in the free-choice situation than in the other two situations. Similarly, target person and kindness were more strongly associated in the no-choice condition than in the other two situations. In other words, the spontaneous motive inference about to help and the spontaneous trait inference about kindness occurred in the free-choice situation. In contrast, these spontaneous inferences did not occur in the no-choice situation.
We conducted a path analysis on the data obtained from the free-choice and control condition to test whether the motive inference about to help was spontaneously formed from the free-choice situation and if they induced the spontaneous trait inference about kindness （see Fig. 1 upper panel）. We used a dummy
variable for the conditions（0 = control, 1 = free-choice）. Bootstrapping results indicated that when Target Person X and to help were signaled, the indirect effect for d′ was marginally significant because the 90% confidence interval did not include zero（0.002‒ 0.464）, but the 95% confidence interval did （-0.034‒0.526）. However, when Target Person
X and to conform to the rule were signaled, the indirect effect for d′ was not significant （the 95% confidence interval was -0.045‒0.306）.
This suggests that participants spontaneously inferred the motive to help from the
free-choice situation, and in turn, spontaneously inferred the trait kindness based on this inferred motive.
Finally, we conducted a path analysis on the data obtained from the no-choice situation and control condition to test whether the motive inference about to conform to the rule was spontaneously formed from the no-choice situation and induced the spontaneous trait inference about obedience （see Fig. 1 lower panel）. We used a dummy variable for the conditions（0 = control, 1 = no-choice）. Bootstrapping results indicated that neither indirect effect was significant（d′ of Target Person X and motive to help : -0.070‒0.180; Target Person X and motive to conform to the rule : -0.192‒0.417）. This suggests that situational circumstances influenced neither motive nor trait inferences.
III. Discussion
Previous studies examined that what kinds of inferences are formed from behavior descriptions. They revealed that perceivers usually form motive and trait inferences spontaneously（e.g., Hassin et al. 2005; Winter & Uleman 1984）, and these spontaneous inferences co-occur（Ham & Vonk 2003; 2011; Todd et al. 2011）. However, these studies did not focus on how motive and trait inferences are related. In the present study, we tested whether spontaneous motive inferences are formed based not only on an actor s behavior but also on the situation in which the actor behaves, and that spontaneous trait inferences are formed based on the inferred motive.
target behavior in which the target person helped another person in a free-choice situation, an association between the target person and implied motive to help（another person）, and an association between the target person and an implied trait of kindness was stronger
than those in the other two situations, i.e., the no-choice and control conditions, respectively. In addition, an association between the target person and the implied motive to help mediated the relationship between the situational circumstance and implied trait of
Situational force
(0 = control,
1 = free-choice)
Target person
+ to help
Target person
+ kindness
Target person
+ to conform to
the rule
Target person
+ obedience
.35* .38* .09 .05 .20 .14(Ȥ
2(2) = 0.81, ns., GFI = .99, RMSEA = .00)
e
e
e
.24e
.49*Situational force
(0 = control,
1 = no-choice)
Target person
+ to help
Target person
+ kindness
Target person
+ to conform to
the rule
Target person
+ obedience
.03 .10 -.05 .03 .11 .58***(Ȥ
2(2) = 0.03, ns., GFI = 1.00, RMSEA = .00)
e
e
e
.31e
.42*Fig. 1. Inferred motives as mediators of inferences about inferred traits（free choice vs. control） are indicated in the upper panel. The lower panel shows no-choice vs. control. Coefficients followed by an asterisk are statistically significant.
kindness. It was indicated that when they saw the voluntary and kind behavior of the target person, they formed a spontaneous motive inference about to help another person and a spontaneous trait inference about kindness. In addition, the spontaneous motive inference about to help（another person） was formed from not only target s kind behavior but also from the situational information that the target person acted under no-situational force. Subsequently, the spontaneous trait inference about kindness was formed based on the inferred motive to help another person. In sum, our hypothesis was supported. A spontaneous motive inference was formed based not only on an actor s behavior but also on the situation in which the actor behaves, and that a spontaneous trait inference was formed based on the inferred motive.
While our hypothesis was supported in the free-choice situation, it was not supported in the no-choice situation. When participants observed the kind behavior of the target person that he was instructed to help another person, an association between the target person and the implied motive to conform to the rule, and an association between the target person and the implied trait about obedience were not different than those in the other two situations, i.e., the free-choice and control conditions. It is indicated that they inferred neither motive nor trait spontaneously. In light of these results for spontaneous inferences, it should be necessary that the target person acts on his free will. Perceivers determine whether the target behavior is voluntary or not in light of situational circumstances. That is, perceivers
would assess whether the target person makes his own decision to behave in light of the situational circumstance in which the target b e h a v i o r o c c u r r e d a n d t h e n t h e y f o r m spontaneous motive and trait inferences, only when they perceive the target behavior as voluntary. Thus, perceivers do not merely form spontaneous inferences from a target s behavior. Rather, they take into account both the behavior and situational circumstances. A spontaneous trait inference would be formed from a spontaneous motive inference only when the target person behaves voluntarily. Even in implicit processing, we consider the situational circumstances, and the spontaneous motive inference mediates the relationship between the situational circumstances and spontaneous trait inference. Motive inferences, even if implicitly made, are important for trait inferences, and situational circumstances are important for spontaneous inferences.
Previous studies have argued that processing to integrate inferences does not occur under implicit processing（Gilbert 1989）. However, the present study suggests that perceivers spontaneously infer an actor s traits based on their inferences of the actor s motives. This depends on not only the target behavior but also situational circumstances. From this perspective, we can integrate various inferences even when under implicit processing. However, there is an issue to examine. In a previous study on explicit processing, participants inferred motives from the behavior that the supervisor forced the actor to perform（Reeder et al. 2004）. However, in the present study, in the no-choice situation, motive inferences did not occur spontaneously. This suggests that
perceivers are able to judge whether a person is behaving voluntarily or involuntarily, but, during implicit processing, they are unable to infer alternative motives from situational circumstances. Reeder et al.（2004）showed that participants formed various motive inferences corresponding to situational circumstances when they observed the same behavior. In future research, in line with the work of Reeder et al.（2004）, it is necessary to explore whether perceivers form various spontaneous motive inferences corresponding to different situational circumstances when they observe the same behavior occurring in different situational circumstances, and if they also form various spontaneous trait inferences c o r r e s p o n d i n g t o s p o n t a n e o u s m o t i v e inferences.
What type of function does implicit processing have in our decision making? Recent studies have argued that spontaneous trait inferences are utilized when perceivers make conscious judgments as an index for predictions about the future behavior of the actor（McCarthy & Skowronski 2011）. They revealed that perceivers predicted subsequent target s behavior in a trait-consistent manner when they were exposed to trait-implicative behaviors describing the target person. That is, perceivers would predict the target person s behavior based on spontaneous trait inferences. In future research, it would be fruitful to explore how spontaneous inferences affect conscious judgments if implicit processing about the target, and how this is altered by situational circumstances and the target s motives.
Acknowledgements
This study was supported by Strategic Research Foundation at Private Universities, 2010―2012, from the Ministry of Education, Culture, Sports, Science and Technology, Japan（program number: S1002008）.
References
Gilbert, D. T.（1989） Thinking lightly about others: Automatic components of the social inference process. In J. S. Uleman & J. A. Bargh（eds.） Unintended thought. New York: Guilford Press, 189―211.
G i l b e r t , D . T . , & M a l o n e , P . S . （ 1 9 9 5 ） T h e
correspondence bias. , ,
21―38.
Gilbert, D. T., Pelham, B. W., & Krull, D. S.（1988） On cognitive busyness: When person perceivers meet persons perceived.
, , 733―740.
Ham, J., & Vonk, R.（2003） Smart and easy: Co-occurring activation of spontaneous trait i n f e r e n c e s a n d s p o n t a n e o u s s i t u a t i o n a l inferences.
, , 434―447.
Ham, J., & Vonk, R.（2011） Impressions of impression management: Evidence of spontaneous suspicion of ulterior motivation.
, , 466―471.
Hassin, R. R., Aarts, H., & Ferguson, M. J.（2005） A u t o m a t i c g o a l i n f e r e n c e s .
, , 129―140. Heider, F.（1958）
. New York: John Wiley.
Lupfer, M. B., Clark, L. F., & Hutcherson, H. W.（1990） Impact of context on spontaneous trait and situational attributions.
, , 239―249.
McCarthy, R. J., & Skowronski, J. J.（2011） What will Phil do next? Spontaneously inferred traits influence predictions of behavior.
McKoon, G., & Ratcliff, R.（1986） Inferences about predictable events.
, , 331―343.
Nosek, B. A., & Banaji, M. R.（2001） The go/no-go association task. , , 625―664. Reeder, G.D., Kumar, S., Hesson-McInnis, M. S., &
Trafimow, D.（2002） Inferences about the morality of an aggressor: The role of perceived motive.
, , 789―803.
Reeder, G. D., Vonk, R., Ronk, M. J., Ham, J., & Lawrence, M.（2004） Dispositional attribution: Multiple inferences about motive-related traits. , , 530―544.
Todd, A. R., Molden, D. C., Ham, J., & Vonk, R.（2011） The automatic and co-occurring activation of multiple social inferences.
, , 37―49.
Trope, Y. & Gaunt, R.（1999） A dual-process model
of overconfident attributional inferences. In S. Chaiken & Y. Trope（eds.）,
New York: Guilford Press, 161―178.
Uleman, J. S., Newman, L., & Moskowitz, G. B.（1996） People as flexible interpreters: Evidence and issues from spontaneous trait inference. 28, 211―279.
Van Overwalle, F., Van Duynslaeger, M., Coomans, D., & Timmermans.（2012） Spontaneous goal inferences are often inferred faster than s p o n t a n e o u s t r a i t i n f e r e n c e s .
, , 13―18. Winter, L., & Uleman, J. S.（1984） When are social
j u d g m e n t s m a d e ? E v i d e n c e f o r t h e spontaneousness of trait inferences.
, , 237―252.
（Recieved: 30 May 2016） （Accepted: 6 September 2016）
