Jump to ContentJump to Main Navigation
The Psychology of Gratitude$

Robert A. Emmons and Michael E. McCullough

Print publication date: 2004

Print ISBN-13: 9780195150100

Published to Oxford Scholarship Online: March 2012

DOI: 10.1093/acprof:oso/9780195150100.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 30 March 2017

(p.291) Appendix Annotated Bibliography of Psychological Research on Gratitude

(p.291) Appendix Annotated Bibliography of Psychological Research on Gratitude

The Psychology of Gratitude
Oxford University Press

With the recent resurgence of positive psychology, many researchers have a renewed interest in the concept of gratitude. Though the current empirical work on gratitude is exciting and promising, such studies are few in number. It is our hope that this annotated bibliography will aid psychologists in their investigation of the nature and effects of gratitude, spurring further advances in theory and research in gratitude.

The studies summarized in this bibliography are both descriptive and experimental. Choosing from research that best illustrated the nature and effects of gratitude, we included experiments that employed vignettes, narratives, and behavioral measures, as well as correlational studies of personality and even telephone surveys. Although there is not complete consensus among these studies, they do come together to paint a rather coherent picture of the nature of gratitude: People experience gratitude in response to a valued positive outcome that another individual intentionally caused. This grateful emotion leads people to desire to act prosocially themselves, at least in the short run. Feelings of gratitude are reported to be pleasant and are experienced often in the course of everyday life.

There is much room for further research in this area of psychology. Future studies can move beyond the reliance on vignette manipulations and self-report measures evident in much previous work on gratitude. Instead, (p.292) we can look at the effects of actual gratitude experienced by research participants, both in their everyday lives and in situations experimentally manipulated to produce gratitude. Given that we have at least a rudimentary understanding of what gratitude is, we can also develop additional studies to answer questions about the effects of gratitude. For example, if gratitude does increase prosocial behavior, is this behavior directed only at the benefactor, or will those experiencing gratitude act more prosocially toward anyone? What are the effects of gratitude on health? In addition, we can look more deeply at the dispositional side of gratitude. Is there such a thing as a grateful personality? What is the relationship between the disposition toward gratitude and variables such as religiousness or the Big Five model of personality? Using the present studies as a foundation, psychologists can continue to explore the complex emotion of gratitude.

Baron, R.A. (1984). Reducing organizational conflict: An incompatible response approach. Journal of Applied Psychology, 69, 272–279.

Objective: To apply the technique of incompatible response strategy to the reduction of destructive organizational conflict.

Design: Between-subjects experiment.

Setting: This study was conducted in a laboratory setting.

Participants: Participants were 85 male and 71 female undergraduate students.

Manipulated variables: The accomplice’s behavior and the incompatible response induction served as the manipulated variables. The participant and a male accomplice of the experimenter were asked to imagine that they were executives who were employed at a large company. Their task was to discuss solutions to a particular problem in their company. The accomplice disagreed with the participant’s statements in one of two ways: In the disagreement condition, the accomplice disagreed with the participant calmly and rationally. In the condescension condition, the accomplice disagreed in a condescending and conceited manner. After the accomplice’s behavior, the experimenter asked the participant and accomplice to wait while she searched for some questionnaires. During this time, the incompatible response variable was manipulated. The accomplice either waited quietly (control condition), offered the participant a piece of candy (gift condition), apologized for his rude behavior and explained that he was stressed out by exams (sympathy condition), or asked the participant to read some cartoons in his notebook (humor condition).

Assessment of outcome variables: After the incompatible response induction, the experimenter returned with the questionnaires containing the dependent variables. These included self-report measures of current mood, personal impressions of the accomplice (liking, reasonableness, and pleasantness), and job-related impressions of the accomplice. Participants were also asked to predict how they would behave in (p.293) future conflicts with the accomplice, rating the likelihood of engaging in the conflict behaviors of accommodation, competition, compromise, and collaboration.

Main results: Looking at the effects relevant to gratitude (the gift condition), participants had a marginally more positive mood in the gift condition than in the control condition. Participants in the gift incompatible response condition reported liking the accomplice more, and rated him as more pleasant in comparison to participants in the control condition. In addition, none of the incompatible response conditions had a significant effect on participant’s job-related ratings of the accomplice. Regarding future conflict behaviors, participants in the gift condition were significantly more likely to engage in the strategy of collaboration, in comparison to participants in the control condition. Effects of the gift condition were similar to effects of the other incompatible response conditions, with a few small differences.

Conclusions: The author concluded that responses incompatible with the anger can be successfully used to reduce negative affect and increase constructive responses to conflict. The author noted that the incompatible response inductions used in this study were minimal, and that stronger levels of incompatible responses may lead to even more constructive responses to conflict.

Commentary: One limitation of this experiment, from the standpoint of gratitude research, is that the specific emotion of gratitude was not measured in the gift condition. However, since the gift condition produced similar effects to the other incompatible response conditions, it is likely at least a small amount of gratitude, which is theorized to be positively valenced, was experienced by participants. This experiment underscores the prosocial nature of grateful feelings, which caused these participants to rate the accomplice more positively as well as to endorse a more constructive conflict strategy in the future, in comparison to participants who did not experience positive emotions. More research is needed to differentiate between any specific effects of gratitude on impressions and behaviors, independent of other positive emotions.

Correspondence: Author was at Purdue University when this study was published.

Baumgarten-Tramer, F. (1938). “Gratefulness” in children and young people. Journal of Genetic Psychology, 53, 53–66.

Objective: In the context of a theoretical analysis of gratitude, to categorize different types of gratitude and to examine whether children and adolescents recognize that certain situations are relevant to gratitude.

Design: Qualitative.

Setting: Elementary and secondary schools in the city of Bern, Switzerland.

Participants: In the first study mentioned, participants included 1,059 school children between the ages of 7 and 15. The second study mentioned included 530 children between the ages of 10 and 15 years.

(p.294) Assessment of outcome variables: Participants in the first study responded to questionnaires that asked, “What is your greatest wish?” and “What would you do for the person who granted you this wish?” Participants in the second study were presented with a scenario in which an individual saved a rich person’s life, but the rich person was injured in the process and took the other individual to court. Participants were asked what they thought the court’s decision should be in the case.

Main results: The author found that participants’ responses in the first study could be grouped into four categories: verbal gratefulness, concrete gratefulness, connective gratitude, and finalistic gratefulness. These different forms of gratitude differed in frequency depending on the age of the participant. In the second study, 14.71% of participants directly mentioned the concept of ungratefulness in reaction to the scenario, whereas the remainder of the participants described the scenario in terms of unfairness.

Conclusions: There are individual differences between children in their expressions of gratitude. Some of these differences are related to age. Whereas results from the second study revealed that children are not always aware of the relevance of gratitude to certain situations, the first study demonstrated that children express gratitude when a situation clearly calls for gratefulness.

Commentary: Although not methodologically or statistically rigorous, this article presents good theoretical material that will be of interest to gratitude researchers.

Correspondence: The author was at the University of Bern, Switzerland, when these studies were published.

Bar-Tal, D., Bar-Zohar, Y., Greenberg, M. S., & Hermon, M. (1977). Reciprocity behavior in the relationship between donor and recipient and between harm-doer and victim. Sociometry, 40, 293–298.

Objective: To investigate the effect of relationship closeness on the reciprocation of favors and harm.

Design: Between-subjects experiment.

Setting: The setting of the experiment is not explicitly stated, but it was most likely conducted in a laboratory or classroom setting.

Participants: Participants were 50 male and 50 female undergraduate students from a course in introductory social psychology.

Manipulated variables: The two manipulated variables were the nature of the relationship between the potential helper and the person in need, and the sex of the participant. Each participant read two similar scenarios about a protagonist who missed a bus to an important event and did not have a car. He/she then called someone to ask for a ride. The scenarios varied in outcome: In one scenario, the person gave the protagonist a ride (help-giving), and in the other scenario, the person refused to give the protagonist a ride (harm-doing). The nature of the relationship was varied by (p.295) manipulating the identity of the potential helper: parent, sibling, friend at dormitory, acquaintance at dormitory, or stranger at dormitory (parent = closest relationship, stranger = least close).

Assessment of outcome variables: Items at the end of each scenario asked participants to rate the extent to which the called person was under an obligation to help. For the help-giving scenarios, participants also rated how grateful they felt toward the person who helped them, and for the harm-doing scenarios, they rated how resentful they felt toward the person who refused to help. Participants were also asked open-ended questions about why they thought the person they called did/did not help them, and how they felt toward the person who helped/didn’t help them.

Main results: The authors first examined participant ratings of the help-giving scenarios. Male participants thought close friends and acquaintances were the most obligated to help, whereas female participants felt that parents had the most obligation to help. Across both sexes, closer relationships were related to increased obligation to help. Conversely, less close relationships were related to increased gratitude. In other words, parents were seen as the most obligated to help, but participants said they would feel most grateful when receiving help from a stranger. Responses to the open-ended questions were content-analyzed and coded. Using the author’s coding system, participants were more likely to state that strangers and acquaintances helped for social responsibility reasons, whereas participants who read about calling a parent, sibling, or close friend were more likely to mention role obligation as the main reason for help. Participants who read about being helped by a stranger or acquaintance were more likely to mention they felt indebtedness and attraction for the helper, and participants reading about the three closer relationships were also most likely to report feeling indebted.

With the harm-doing scenarios, participants also rated that parents and siblings were under the most obligation to help, with close friends under an intermediate obligation, and acquaintances and strangers under the least obligation. Participants felt the most resentment toward parents and siblings who had refused to help and the least resentment toward nonhelping strangers. Regarding open-ended questions, participants who read about acquaintances and strangers were most likely to mention fear as an explanation for their refusal to help.

Conclusions: The nature of the relationship between the potential benefactor and recipient affects individuals’ reactions when they are helped or refused helped. Closer relationships, such as those with parents and siblings, bring with them increased obligation to help, decreased gratitude in response to help, and increased resentment in response to the refusal to help.

Commentary: This experiment demonstrates that role obligation is an important component of gratitude. Individuals feel most grateful for help when they do not feel entitled to that help out of relationship obligations. Researchers would be wise to take into account the relationship between benefactor and recipient when investigating grateful responses to help.

(p.296) Correspondence: Daniel Bar-Tal was at Tel-Aviv University when this study was published.

Baumeister, R. F., & Ilko, S.A. (1995). Shallow gratitude: Public and private acknowledgement of external help in accounts of success. Basic and Applied Social Psychology, 16, 191–209.

Objective: To test the hypothesis of “shallow gratitude,” that people privately take credit for their successes, but publicly express gratitude by sharing the credit with others.

Design: Between-subjects experiment.

Setting: Laboratory.

Participants: Participants were 67 undergraduate students.

Manipulated variables: Participants were asked to write two stories, one concerning an important success experience that occurred in the past 2 years, and the other concerning a major failure experience. Participants in the private condition were asked not to put any identifying information in the stories. Participants in the public condition were asked to write their names on every sheet of paper they filled out, and were told that they would discuss their stories with a small group after completing the two stories.

Assessment of outcome variables: Stories were coded for mentions of direct help, emotional support, direct hindrance, and emotional hindrance from others. They were also coded for fairness and contradictory material (self-enhancing statements in failure stories or self-deprecating statements in success stories).

Main results: Mentions of help from others occurred mainly in success stories, whereas mentions of hindrance from others occurred mainly in failure stories. For success stories, mentions of direct help were significantly more frequent in the public condition, compared to the private condition. A marginal effect occurred for emotional support in the same direction. In other words, participants seemed more likely to express gratitude to others for their successes when their accounts were to be made public, rather than kept private. Regarding failure stories, participants were equally likely to mention direct hindrance and emotional hindrance from others in both the public and private stories. The authors also present supplementary findings regarding their fairness and contradictory statements variables.

Conclusions: These results support the “shallow gratitude” hypothesis, that individuals exhibit less self-serving bias when their success stories are public versus private. Because failure accounts did not show a public-private difference in the mention of the actions of other people, these results cannot be explained by the nonmotivational cue of the presence of others in the public condition. This implies that public expressions of gratitude are not always genuine, but part of a motivated self-presentation strategy.

(p.297) Commentary: Although this study does not directly assess grateful emotions, it has important ramifications for gratitude research. Expressions of gratitude may vary depending on the self-presentational motivations of the participants, especially in retrospective accounts.

Correspondence: Roy M. Baumeister, Department of Psychology, Case Western Reserve University, Cleveland, OH 44106.

Becker, J. A., & Smenner, P. C. (1986). The spontaneous use of thank you by preschoolers as a function of sex, socioeconomic status, and listener status. Language in Society, 15, 537–546.

Objective: Researchers investigated whether preschoolers would spontaneously say “thank you” in a familiar social setting when their parents were not present, and whether certain demographic variables were related to spontaneously saying “thank you.”

Design: Between-subjects experiment.

Setting: Day care centers in a southeastern metropolitan area.

Participants: Participants were 250 children (129 girls, 121 boys) between the ages of 3.5 and 4.5 years. Of the participants, 146 children were from low-income families, whereas 104 children were from middle-income families.

Assessment of predictor variables: Children from low-income families were drawn from day care centers that did not charge a fee for services. Children from middle-income families were drawn from day care centers that charged the highest rates in the area, and these children had parents employed in professional occupations. Experimenters also noted the sex and ethnicity of children. Children played a color-naming game with teachers, and after correctly guessing the color on a card, each child went to another room, where he or she was given a sticker by a child model or an adult model.

Assessment of outcome variable: Responses to the receipt of the sticker were recorded on audiotape and noted by the adult model. Researchers noted whether participants said “thank you” or related phrases.

Main results: Thirty-seven percent of children said “thank you” in response to receiving a sticker. Girls were more likely to respond with “thank you” than boys. Children from lower income families said “thank you” more often than children from middle-income families, but no effects were found for ethnicity. Children were more likely to say “thank you” to the adult model than to the child model.

Conclusions: Previous studies cited by the authors revealed a lower frequency of children responding “thank you” in the presence of parents. The authors argued that young children may see parental prompting to say “thank you” as part of the politeness routine, and that they then wait for the parent to prompt them before they say “thank you.” However, in the absence of parents, the authors posited, children can (p.298) recognize when a situation calls for saying “thank you.” Demographic variables such as sex, socioeconomic status, and benefactor status all affected children’s use of “thank you.” The authors suggested that these differences may be caused by differences in socialization.

Commentary: Although these researchers did not measure children’s perceptions or experiences of gratitude, it is possible that saying “thank you” in response to a gift is indicative of some amount of gratitude, even in children. The authors mentioned this when they discussed the relationship between socioeconomic status and frequency of saying “thank you.” Although it is possible that children from lower income families are socialized to say “thank you” more frequently, the authors noted that these children also seemed to be more excited about receiving the stickers than did children from middle-income levels. Therefore, the fact that children from lower income families said “thank you” more often may indicate that they also felt more gratitude. At the least, these studies show that some children as young as 3.5 years old recognize that expressions of gratitude are relevant in certain situations, without being prompted by their parents.

Correspondence: Both authors were affiliated with the Department of Psychology at the University of South Florida at the time of their research.

Biner, P. M., & Kidd, H. J. (1994). The interactive effects of monetary incentive justification and questionnaire length on mail survey response rates. Psychology and Marketing, 11, 483–492.

Objective: To test equity theory in the context of responding to a mail survey. These researchers believed that feelings of obligation would lead to an increase in the number of surveys sent back by respondents. Specifically, they wished to explore the effects of under- and overcompensation on response rates.

Design: Between-subjects experiment.

Setting: A midwestern city with a population of 120,500.

Participants: Participants were 200 people whose names were selected from a telephone directory using a systematic random sampling technique. Approximately 90% of the residents of the city were listed in the directory.

Manipulated variables: Participants were assigned to one of four conditions in a 2 x 2 design using a randomized blocks procedure. Participants in all conditions were sent a questionnaire ostensibly designed to assess community needs, along with a cover letter and a $1 bill. The first manipulation was presented in the cover letter. Half of the participants read that the $1 was included to make individuals obligated to return the questionnaires (equity-salient condition). The other half of the participants read that the $1 was a token of appreciation in advance for filling out the questionnaire (standard condition). The other manipulation consisted of the length of the questionnaire. Half of the participants received a one-page questionnaire with 10 questions (short questionnaire condition), and the other half received a five-page questionnaire with (p.299) 50 questions (long questionnaire condition). Researchers hypothesized that participants who received $1 with the short questionnaire would feel overcompensated, and that participants who received $1 for the long questionnaire would feel undercompensated.

Assessment of outcome variables: Questionnaires were coded to determine participants’ treatment conditions. The dependent measure consisted of the number of completed questionnaires returned within a 3-week period.

Main results: The equity-salient letter produced a significantly higher response rate than the standard letter, but only with the short questionnaire. With the long questionnaire, the effect was reversed but nonsignificant.

Conclusions: These results support an equity interpretation of monetary compensation’s effects on response rates. Overcompensation—in this case, receiving payment for a short questionnaire—along with a statement of obligation caused participants to increase their rates of response.

Commentary: On the surface, these results seem to contradict other research reviewed in this bibliography. Other studies suggest that expression of gratitude, as in the standard condition of the present study, should increase response rates. However, there are a number of differences between the present study and other research. One difference is that the present study did not have a control condition in which neither equity nor gratitude was made salient. It may be that the standard cover letter expressing gratitude led to a higher response rate than a letter that had no gratitude or equity at all. Additionally, this study can be seen as consistent with Carey, Clicque, Leighton, and Milton’s (1976) results. (See the next entry in this bibliography.) Carey et al. compared the effects of saying “thank you” to customers with the effects of thanking customers and telling them about a new sale. The “thank you” plus information about a sale did not increase sales as much as the simple “thank you” did. This may have occurred because the information about the sale made customers think that the person calling them had ulterior motives. In the same way, the inclusion of $1 with all questionnaires in the present study could have been perceived by participants as coercive, nullifying any effects the expression of gratitude might have had on response rates. More directly, this present study also demonstrates that the expression of gratitude is different from (or at least weaker than) a direct statement of obligation or equity.

Correspondence: Paul M. Biner, Department of Psychological Science, Ball State University, Muncie, Indiana 47306

Carey, J.R., Clicque, S. H., Leighton, B.A., & Milton, F.(1976). A test of positive reinforcement of customers. Journal of Marketing, 40, 98–100.

Objective: To explore whether thanking customers for their business would serve as positive reinforcement, thereby increasing sales.

Design: Controlled field experiment.

(p.300) Setting: A central Texas city of 22,000.

Participants: Participants were 440 customers of M&M Jewelers.

Manipulated variables: Customers were randomly assigned to one of three conditions. For the first condition, customers were called and thanked for their business. In the second condition, customers were thanked for their business and told of a special sale. In the control group, no call was made.

Assessment of outcome variables: Overall sales for the retail establishment were measured. Additionally, individual sales from customers from the different conditions were also recorded.

Main results: The store experienced a 27% increase in sales during the test month as compared with the previous month. This increase in sales was due to participants in the two experimental groups. The largest increase in sales came from customers who were simply thanked (70%), whereas there was a small increase for those who were thanked and told about the sale (30%), and no increase for customers in the control group. The increase in sales seemed due to the renewed interest of dormant rather than regular customers. In addition to increased sales, there was also an increase in payment of delinquent accounts by customers.

Conclusions: Thanking customers can serve as a form of positive reinforcement, stimulating increased spending, especially among customers who make purchases less regularly.

Commentary: It is interesting to note that the thank-you condition had a much bigger effect than the condition in which customers were also told about a sale. It may be that news about a sale cheapened the expressed gratitude in the second condition, making it seem as though store employees were calling customers expressly to increase sales. The differences between conditions suggest that the increase in sales was caused, not so much by general positive reinforcement, but by the reinforcing nature of gratitude.

Correspondence: No designated mailing address, but the first author listed his affiliation at the University of Texas at San Antonio at the time of the study.

Clark, H. B., Northrop, J.T., & Barkshire, C.T. (1988). The effects of contingent thank you notes on case managers’ visiting residential clients. Education and Treatment of Children, 11, 45–51.

Objective: To explore whether thank-you letters would increase the frequency with which case managers visited their adolescent clients in a residential program.

Design: Controlled field experiment.

Setting: Adolescent residential unit at a community mental health center.

Participants: Case managers from the welfare department or juvenile court who were assigned to youth in the residential unit.

(p.301) Intervention: After a baseline period of 20 weeks during which no special interactions occurred between staff and case managers, staff implemented the intervention phase. During the intervention phase, thank-you letters were sent to case managers and their supervisors, contingent upon each case manager’s visiting the residential unit. Letters thanked the case manager for his or her visits, underscored the importance of the visits, and indicated that the supervisor was sent a copy of the letter. The intervention phase lasted 20 weeks. After this intervention, there was a third period of 10 weeks in which no letters were sent, approximating the previous baseline phase.

Assessment of outcome variables: Throughout the three phases of the study, case manager visits were recorded in two ways. First, case managers were required to sign a log book during each visit and to list the client or staff members they were visiting. Second, case managers were required to meet with one of the staff before or after meeting with their clients. At the end of each week, each staff member checked off which case managers they had seen that week, and each staff member indicated whether he or she had met with his or her client. Reports by staff members produced over 98% agreement with the log book.

Main results: Results were reported separately for two different social service agencies by which the case managers were employed. Case managers from the different agencies showed a similar pattern: Case managers’ visits to their clients increased during the intervention period (74%, 81%) when compared with the baseline (45%, 40%) and postintervention (44%, 55%) periods.

Conclusions: Thank-you letters served as effective reinforcers for case managers’ visiting their clients. There were certain qualifications to the data: Some case managers and clients changed during the 50 weeks of the study, and the increase in visits began to drop off during the end of the intervention phase. This graded reduction in the potency of the intervention may reveal the existence of habituation effects. Suggestions by the authors for further research include investigation of the specific content of thank-you letters and the effect of increased visits on clients.

Commentary: This study provides strong support for the reinforcing nature of gratitude. The advantages of this study include its location in the field, as well as its use of behavioral measures. A small weakness of this study is that perceptions of gratitude were not actually measured, but this would have been difficult, given the naturalistic nature of the study.

Correspondence: Hewitt B. Clark, Associate Chairperson, Department of Child and Family Studies, University of South Florida, Tampa, Florida, 33612

Emmons, R.A., & McCullough, M.E. (2003). Counting blessings versus burdens: An experimental investigation of gratitude and subjective well-being in daily life. Journal of Personality and Social Psychology, 84, 377–389.

Objective: To test the causal effect of gratitude on physical and psychological well-being.

(p.302) Design: Between-subjects experiment.

Setting: Participants completed questionnaires in their homes and returned questionnaires either directly to the experimenter in a classroom setting (Studies 1 and 2) or by mail (Study 3).

Participants: Participants for Studies 1 and 2 consisted of college students. Study 1 included 146 female and 54 male undergraduate students. Study 2 participants consisted of 125 female and 41 male undergraduate students. Participants in Study 3 were 44 female and 21 male adults with congenital and adult-onset neuromuscular disorders.

Manipulated variables: In Study 1, participants were asked to complete 10 weekly reports of either gratitude, hassles, or meaningful events. In Study 2, participants were asked to complete 14 daily reports of either gratitude, hassles, or downward social comparisons. In Study 3, half of the participants completed 21 daily reports of gratitude, and the other half of the participants were assigned to a control condition and only completed the dependent measures for 21 days.

Assessment of predictor variables: In Study 1, dependent variables included self-reports of the following: mood (including grateful mood, as well as positive and negative affect), physical symptoms, reactions to aid received, time spent exercising, and global life appraisal.

In Study 2, dependent variables included self-reports of the following: mood, physical symptoms, health behaviors (including amount of strenuous and moderate exercise, number of caffeinated and alcoholic beverages consumed, number of pain relievers taken, and amount and quality of sleep), and prosocial behaviors.

In Study 3, dependent variables included self-reports of the following: mood, global life appraisals and feelings of connection with others, health behaviors (including amount and quality of sleep, ratings of physical pain, and amount of exercise), and difficulties with daily activities. In addition, the participants’ significant others completed observer reports of the participants’ positive and negative affect and life satisfaction.

Main results: In Study 1, participants who completed weekly gratitude reports rated their lives more favorably than those participants who wrote about hassles or meaningful events. Gratitude participants also reported fewer symptoms of physical illness than participants in the other two groups and spent more time exercising than participants in the hassles group. Across all three groups, feelings of gratitude in response to aid received was positively related to feelings of joy and happiness, favorable life appraisals, and optimism about the coming week.

In Study 2, participants who completed daily gratitude forms experienced an increase in positive affect over the 13 days of the study, when compared with participants who completed daily hassles forms. The effect that the daily gratitude versus daily hassles interventions had on positive affect was mediated by self-reported feelings of daily grateful mood. Writing daily about gratitude did not have an effect on physical symptoms or health behaviors, in contrast to Study 1. Participants in the daily gratitude condition reported offering more emotional support, in comparison (p.303) with participants in the hassles and downward comparison groups, and gratitude participants also showed a marginal effect of self-reported helping of others, in comparison with the hassles participants.

In Study 3, writing about daily gratitude increased daily positive affect and decreased daily negative affect in participants, and these effects were mediated by daily felt gratitude. Participants in the gratitude condition also felt more satisfied with their lives, felt more optimism about the coming week, and felt more connected with others, relative to participants in the control condition. Gratitude participants also reported getting more sleep at night than control participants. Observer reports also rated participants in the gratitude condition as having higher positive affect and higher life satisfaction, in comparison with observer reports of participants in the control condition.

Conclusions: Focusing on gratitude rather than hassles, life events, downward social comparison, or a control group increases components of both psychological and physical well-being. These positive effects of writing about gratitude are mediated by the experience of gratitude and are not simply a general increase in positive affect.

Commentary: These studies provide good support for the positive potential of a grateful outlook on life. Researchers provided a number of different comparison conditions and dependent measures, and results generalized beyond college students to adults coping with neuromuscular disorders. These studies made good use of experimental design to test the causal relationship between a grateful outlook and well-being.

Correspondence: Robert A. Emmons, Department of Psychology, University of California, One Shields Avenue, Davis, California 95616, e-mail:raemmons@ucdavis.edu

Farwell, L., & Wohlwend-Lloyd, R. (1998). Narcissistic processes: Optimistic expectations, favorable self-evaluations, and self-enhancing attributions. Journal of Personality, 66, 65–83.

Objective: The study described here is the third in a series of three. Studies 1 and 2 demonstrated that individuals high in narcissism tended to have more optimistic expectations for the future and to possess enhanced self-evaluations. Study 3 investigated the effect of narcissism on perceptions of a collaborator in the context of a collective group project.

Design: Correlational.

Setting: This study was conducted in a laboratory setting.

Participants: Participants were 37 female and 30 male undergraduate students enrolled in an introductory psychology class.

Assessment of predictor variables: Participants’ narcissism scores were collected using the Narcissistic Personality Inventory (Raskin & Hall, 1979).

(p.304) Assessment of outcome variables: Individuals were paired with partners. Participants were asked to predict how successful they and their partners would be on a joint creativity test. Each participant was asked to predict his or her own performance on the test and his or her partner’s individual performance, as well as their joint performance. After being given bogus positive feedback about the outcome of the task, participants were asked to rate the extent to which they thought their own performance was due to luck, task ease, own ability, own effort, as well as partner ability and effort. Participants also rated items measuring personal affect (e.g., pride) and interpersonal affect (e.g., gratitude).

Main results: Narcissism was positively related to prediction of a participant’s own performance on the upcoming test and negatively related to predictions of a partner’s performance. Narcissism was not related to predictions of joint performance. Narcissism was related positively to individuals attributing test success to their own ability and efforts, but it was unrelated to ascribing the success to external factors such as luck, task ease, or a partner’s ability and effort. Interpersonal affect (gratitude) was negatively related to narcissism.

Conclusions: The trait of narcissism is comprised of optimistic expectations about the future, as well as positive overassessments of one’s current outcomes. Narcissism is related to attributing success to one’s own efforts, but is not related to a disinclination to attribute success to a collaborator’s effort or ability. This suggests a self-aggrandizement component to narcissism, rather than other-derogation. Narcissism was also related to less liking and gratitude toward one’s collaborator. These results raise the possibility that narcissism can be disruptive in interpersonal contexts.

Commentary: Although not direct measures of gratitude, increased attributions to one’s own ability and effort can be related to the inhibition of gratitude, in that the experience of gratitude requires that one recognize the contributions of others. If one is too caught up in self-aggrandizement (as narcissists appear to be), this may take one’s attentional focus off other-oriented attributions necessary for gratitude.

Correspondence: Lisa Farwell, Department of Behavioral Studies, Santa Monica College, 1900 Pico Boulevard, Santa Monica, California 90405

Fredrickson, B. L., Tugade, M. M., Waugh, C. E., & Larkin, G. R. (2003). What good are positive emotions in crises? A prospective study of resilience and emotions following the terrorist attacks on the United States on September 11th, 2001. Journal of Personality and Social Psychology, 83, 365–376.

Objective: To examine the occurrence and function of positive emotions following a crisis. Specifically, researchers tested the hypotheses that positive emotions such as gratitude, interest, and love would have calming physiological effects, broaden people’s ways of thinking, and build resources for future coping.

Design: One-group pretest-posttest.

Setting: Laboratory.

(p.305) Participants: Participants were 18 male and 29 female undergraduates and recent graduates of the University of Michigan.

Assessment of predictor variables: Participants completed a number of questionnaires before the September 11, 2001 crisis, including measures of ego-resiliency (Block & Kremen, 1996), a selection of three of the Big Five personality measures (Neuroticism, Extraversion, and Openness; Costa & McCrae, 1992), the Satisfaction with Life Scale (SWLS; Diener, Emmons, Larsen, & Griffin, 1985), the Life Orientation Test (LOT; Scheier & Carver, 1985), and a measure of tranquility constructed from two tranquility-related filler items from the LOT.

Assessment of outcome variables: Participants returned to the lab for the post-crisis measures from September 23, to November 6, 2001, during which a number of world events related to the September 11th crisis continued to unfold. Participants rated their current mood, and wrote about the most important problem or stressor they had experienced since September 11, 2001. They completed a questionnaire about finding positive meaning in their current problems, and rated the extent to which they had experienced numerous positive and negative emotions since the September 11th attacks. Participants also noted their depressive symptoms on the Center for Epidemiological Studies-Depression Measure (CES-D; Radloff, 1977). Finally, participants completed post-crisis questionnaires of the SWLS, the LOT, and tranquility, using the same questionnaires that were used precrisis.

Main results: None of the participants reported losing a loved one in the September 11th attacks. The most frequently experienced emotions after the attacks were sympathy/compassion, gratitude, interest, love, and anger. Higher levels of trait resiliency were related to lower levels of sadness, and higher levels of interest, joy, hope, sexual desire, pride, and contentment. Individuals with high trait resiliency also reported less depression and increased psychological resources, and these effects were mediated by the experience of positive emotions (including gratitude).

Conclusions: After the September 11th crisis, gratitude was one of the positive emotions experienced by individuals with both high and low trait resiliency. Participants felt grateful for their own safety and the safety of those close to them. Although all participants experienced negative emotions in the aftermath of this crisis, the experience of positive emotions such as gratitude, love and interest helped to buffer resilient individuals from the negative emotions and psychological symptoms that a crisis normally brings. The authors concluded that the experience of positive emotions is one of the key components of trait resilience.

Commentary: Although not centered specifically on gratitude, this study demonstrates some adaptive functions that grateful feelings might share with other positive emotions in the context of a crisis. It also suggests that gratitude can occur not only in response to positive outcomes, but also in the midst of negative world events. Teigen (1997), reviewed below, discusses the role of counterfactual thinking in the experience of gratitude, and counterfactual thinking (“It could have been worse”) may be an important factor in the experience of gratitude in response to negative outcomes such as the September 11th attacks.

(p.306) Correspondence: Barbara L. Fredrickson, Department of Psychology, University of Michigan, 525 East University Avenue, Ann Arbor, MI, 48109–1109. E-mail: blf@umich.edu.

Gallup, G. H., Jr. (1998, May). Thankfulness: America’s saving grace. Paper presented at the National Day of Prayer Breakfast, Thanks-Giving Square, Dallas.

Objective: To explore the nature of gratitude in adults and teenagers in the United States.

Design: Two telephone interview polls.

Setting: The adults were surveyed from U.S. homes in mid-April, 1998. Teenagers were surveyed during the period between January and March, 1998.

Participants: There were 482 adults, aged 18 and older, and 500 teenagers between the ages of 13 and 17.

Assessment of predictor variables: Participants were from one of two samples: an adult sample or a sample of teenagers.

Assessment of outcome variables: Participants answered self-report rating measures on the following topics: frequency of expressing gratitude to others, frequency of expressing gratitude toward God, different ways they might express gratitude, whether expressing gratitude makes them happy, and how many people they know who seem to be grateful all the time for no reason.

Main results: Most adults in the United States say that they express gratitude to God (54%) and to others (67%) “all the time.” The rank order of endorsed ways of expressing gratitude were as follows: telling family and friends that one is grateful to them (96%), worshipping and praying (85%), giving money to charity (81%), community service (70%), and saying grace at meals (64%). The majority of adults said that expressing gratitude made them “extremely happy” (60%). Only 25% of adults said that they know a lot of people who are dispositionally grateful, whereas 68% said that they know a few people who are grateful all the time. Results for teenagers paralleled those for adults, although teens tended to express less gratitude.

Conclusions: The author contrasts gratitude with the negative view that many have of society. The fact that so many people report gratitude, and endorse so many different forms of gratitude, paints a more optimistic picture of humanity.

Commentary: This study is useful in showing the importance of gratitude in the everyday lives of individuals. People seem to perceive an abundance of gratitude in society, and to see gratitude as a positive emotion, the expression of which brings happiness into their lives.

Correspondence: No information listed.

(p.307) Gillani, N. B., & Smith, J. C. (2001). Zen meditation and ABC relaxation theory: An exploration of relaxed states, beliefs, dispositions, and motivations. Journal of Clinical Psychology, 57, 839–846.

Objective: To investigate the psychological effects of Zen meditation on experienced practitioners.

Design: Between-subjects quasi-experiment.

Setting: All participants were tested in the Chicago area. Participants in the Zen meditation group were tested during regularly scheduled weekly group meditation in their Buddhist temples. Participants in the control group were tested in a general psychology class at a junior college.

Participants: Participants included 59 Zen meditators from local Japanese Zen temples, and 24 students from a local junior college.

Assessment of predictor variables: Participants in the meditation group engaged in Zen meditation for about an hour during a regularly scheduled group meditation session. Participants in the control group engaged in a silent 60-minute relaxation activity that consisted of reading leisure material during class. Participants were not randomly assigned to experimental conditions.

Assessment of outcome variables: Outcome variables consisted of self-report responses to items from the Smith Relaxation Inventory Series (Smith, 2001), which assesses variables associated with successful relaxation. In constructing the Smith Relaxation Inventory Series, Smith and his colleagues used a lexicographical technique to identify words that people used to describe experiences that they had while engaging in methods of relaxation, such as yoga and meditation. Through factor analysis, Smith and his colleagues constructed 15 relaxation state factors (R-States), including the relaxation state most relevant to gratitude: Love and Thankfulness. Participants filled out reports about their experiences of relaxation states before and after the experimental intervention. In addition, they filled out questionnaires about their relaxation beliefs (R-Beliefs), their propensities to experience a particular relaxation state over a 2-week period of time (R-Dispositions), their motivations to experience more of a given relaxation state (R-Motivations), and attitudes they might have that are not conducive to relaxation (R-Attitudes). The factor Love and Thankfulness took the form of an R-State, an R-Belief, and R-Disposition, and an R-Motivation. All scales were given before the intervention, and the R-State items were given again after the intervention.

Main results: Meditators scored higher than control participants on the pretest of the R-Motivation of Love and Thankfulness. In other words, experienced meditators were more motivated to experience the relaxation state of Love and Thankfulness than were control participants. There were no group differences on presession measures of the R-State Love and Thankfulness, but after the intervention, participants who had meditated reported higher levels of love and thankfulness, whereas control participants did not report change in this variable.

(p.308) Conclusions: Because of the quasi-experimental nature of this experiment, it was not possible to know whether differences between mediators and controls were due to meditation or to other demographic variables. It is also possible that individuals who are more prone to focus on love and thankfulness are also more likely to engage in Zen meditation. However, the fact that meditators experienced a change in presession to postsession love and thankfulness suggests that thankfulness is an important component of Zen meditation.

Commentary: The results of this research suggest that gratitude is an important outcome of Zen meditation. In light of the possible positive effects of gratitude on psychological and physical well-being (see Emmons & McCullough, 2003, listed previously), the exploration of different techniques of inducing gratitude is important for practitioners and researchers alike.

Correspondence: Jonathan C. Smith, Director, Roosevelt University Stress Institute, Roosevelt University, 430 South Michigan Avenue, Chicago, Illinois 60605

Graham, S. (1988). Children’s developing understanding of the motivational role of affect: An attributional analysis. Cognitive Development, 3, 71–88.

Objective: These researchers used two studies to explore the development of the understanding of attributional processes behind complex emotions.

Design: Laboratory experiment.

Setting: Research rooms at a university-affiliated elementary school.

Participants: Study 1 participants consisted of 125 children from three age groups: 5- to 6-year-olds (n = 34), 7- to 8-year-olds (n = 46), and 10- to 11-year-olds (n = 39), divided approximately evenly by sex within each age group. Only children who had some knowledge about pride, gratitude, and guilt were included in the study. The experimenter asked children to tell her what would make them proud/grateful/guilty. Children were retained as participants if they mentioned an accomplishment as a source of pride, another’s actions as a source of gratitude, and a wrongdoing as a source of guilt.

Study 2 participants were recruited from the same elementary school as those from Study 1, and consisted of 105 children from three different age groups: 5- to 6-year-olds (n = 38), 7- to 8-year-olds (n = 32), and 10- to 11-year-olds (n = 35).

Manipulated variables: In Study 1, children were presented three stories, each involving a different emotion: pride, gratitude, or guilt. In the pride scenarios, the situation was either caused by the protagonist of the story, or by other people/outside forces (locus). In the gratitude and guilt scenarios, the protagonist either did or did not have control over his or her actions (controllability). In Study 2, the same three scenarios were presented, with similar variations. In addition, researchers manipulated how much of the emotion the protagonist in each scenario might feel. Target children in the scenario were either said to have felt a lot or none of the emotion in question (pride, gratitude, and guilt).

(p.309) Assessment of outcome variables: In Study 1, after being told each scenario, children were asked to rate the causes of the outcome of each scenario in terms of locus for the pride scenario and controllability in the gratitude and guilt scenarios. Next, they were asked to rate the affective responses of the target child—how proud, grateful, and guilty the target child would be for the respective scenarios. They were also asked to rate the intensity of an irrelevant affect for each scenario: angry (pride scenario), scared (gratitude scenario) and glad (guilt scenario). Last, children rated a series of behavioral intentions. In the pride scenario, they rated how many gold stars the child would give himself or herself for getting an A on a test; in the gratitude scenario, they rated how likely the target child would be to give a gift to the team captain who picked him or her for a team; and in the guilt scenario, they indicated how much money the target child should give to another child whose bike he or she had damaged.

In Study 2, children were asked to rate the same behavioral intentions of the target children presented in Study 1. In this way, the researchers investigated in Study 2 whether attributions and affect were direct causes of behavioral intentions.

Main results: In Study 1, irrelevant affect was not rated highly by children in any age group. Looking specifically at gratitude, all age groups perceived that controllable causes elicited more gratitude, but this was perceived significantly more distinctly by the 8- and 10-year-olds. Additionally, all children were more likely to say that the target child would reciprocate with a gift when the cause was controllable, but again this effect was significantly more prominent for the 8- and 10-year-olds. Correlational analyses showed that, as children get older, their evaluations of characters’ likely attributions, affects, and behavioral intentions become more highly interrelated.

In Study 2, a significant main effect was present for affect: The amount of pride, gratitude, or guilt that the target child was said to have felt influenced participants’ judgments of behavioral intentions. In relation to gratitude, feelings of gratitude in the target child caused participants to rate that the child was more likely to reciprocate a positive outcome with a gift. There was also a main effect for controllability, with participants inferring gift giving when the cause was controllable, but this accounted for much less variance than felt gratitude. Although there were age by affect interactions with pride and guilt, there was no significant interaction with gratitude.

Conclusions: As children age, their understanding of complex emotions, including gratitude, increases. Children as young as 5 understand that a controllable positive event elicits more gratitude than one that was uncontrollable, and that gratitude leads to a greater probability of reciprocating a favor. As children grow older, these associations between controllability, gratitude, and reciprocation become more developed and interrelated.

Commentary: This set of studies shows that even young children have some knowledge about the events that elicit gratitude, as well as the behavioral intentions paired with this emotion. Future developmental studies could move beyond the sole use of scenarios, to behavioral measures of gratitude.

Correspondence: Sandra Graham, Graduate School of Education, University of California, Los Angeles, California 90024

(p.310) Graham, S., & Barker, G. P. (1990). The down side of help: An attributional-developmental analysis of helping behavior as a low-ability cue. Journal of Educational Psychology, 82, 7–14.

Objective: To investigate whether the receipt of help can serve as a cue for low ability, and whether this effect differs for children of various age groups.

Design: Between-subjects experiment.

Setting: An elementary school room outside of participants’ classrooms.

Participants: Participants were 90 elementary school children. Children were selected to vary by age, with 15 female and 15 male participants from each of the following age categories: ages 4–5, ages 7–8, and ages 11–12.

Assessment of predictor variables: Experiment 2 contained the gratitude-relevant measures; therefore only information from Experiment 2 will be reported. The predictor variable (help/not help) was manipulated by having children watch a video where a teacher is supervising a math test. She glances at the paper of one student and walks by, then glances at the paper of another student and gives him unsolicited help. Both students are then seen turning in their papers, and are told that they did well on the test (scored 8 out of 10 problems).

Assessment of outcome variables: After watching the video, participants filled out questionnaires about the effort and ability of each of the two students in the video. Additionally, they rated the extent to which each of the students would feel happiness, pride, gratitude, sadness, and worry. Participants were also asked which of the two students they would choose to work with in a group math task.

Main results: Children age 7 and older rated the helped students as lower in ability than the nonhelped student. Children of all age groups rated the nonhelped student higher on effort. Ratings of happiness differed by age, with the oldest children rating the helped and nonhelped students as equally happy. Older children rated the nonhelped student as higher in pride, whereas younger children did not make a distinction between students in ratings of pride. All children rated the helped child as feeling more gratitude than the nonhelped child, demonstrating a link between attributing success to an external source, and feelings of gratitude. Though this effect was significant, it is interesting to note that children still rated grateful feelings of nonhelped children between 2 and 3 on a 1–7 scale. Children did not rate the students as overly sad or worried. Lastly, older children were more likely than younger children to pick the nonhelped student as a future group member.

Conclusions: Receiving unsolicited help in an achievement context serves as a low-ability cue, especially after ages 5–6.

Commentary: The gratitude-relevant dependent measure provides further support for the hypothesis that gratitude is elicited by positive outcomes due to external agents. The low but nonzero ratings of gratitude of the nonhelped student also point to the possibility that people may perceive a least a small amount of gratitude for positive outcomes, regardless of agent. This study is limited by its scenario methodology (p.311) (the children were not actually experiencing gratitude themselves, but inferring it in others), but also contains a strong point in its developmental perspective.

Correspondence: Sandra Graham, Graduate School of Education, University of California, Los Angeles, CA 90024.

Graham, S., Hudley, C. & Williams, E. (1992). Attributional and emotional determinants of aggression among African-American and Latino young adolescents. Developmental Psychology, 28, 731–740.

Objective: To determine whether aggressive children differed in the intentionality attributions they made for others’ ambiguous behavior with negative outcomes, and whether this increase in intentionality attribution led to increases in anger and aggressive action tendencies.

Design: Mixed-model experiment.

Setting: Classrooms and resource rooms in a junior high school that primarily served students from families with low socioeconomic status.

Participants: Participants were 88 students (74 males and 14 females) from the 7th and 8th grades.

Assessment of predictor variables: An aggressive group of 44 students was created through peer and teacher nominations. A control group was created of 44 nonaggressive students that matched the aggressive group by gender and ethnicity. Children were presented with 8 stories in which they were to imagine themselves as the protagonist. In the stories, the protagonist experiences a negative outcome, caused by a hypothetical peer. Each story had 4 different versions which differed by intention: prosocial, accidental, ambiguous, and hostile. Children read one story paired with a prosocial intent, one with a hostile intent, two with accidental intent, and four with ambiguous intent. The authors presented an example of one of the 8 stories, which asks children to imagine that one of their homework papers blows away, and another child steps on the paper and leaves a muddy footprint on it. In the prosocial intent condition (which is the condition most relevant to feelings of gratitude), the story continues, “The other kid turns to you and says, ‘I could see that your paper was going to blow in the gutter. I’ll help you copy it over’” (p. 734). In comparison, the hostile version of the story reads “The other kid laughs at you, says, ‘That was your tough luck,’ and then turns and runs into the school.”

Assessment of outcome variables: Children responded to dependent measures on questionnaires. They were asked about the intentionality in each of the stories, as well as how mad, angry, and thankful they would feel for each story. Lastly, children were asked to rate their behavioral intentions for each story, with behavioral intentions ranging from prosocial to indirectly and directly aggressive.

Main results: Ratings of intentionality were lowest for the accidental and prosocial versions of the stories, and highest for the hostility versions of the stories, with (p.312) ambiguous stories rated as intermediate in intentionality. Aggressive children rated ambiguous stories as more intentional than did nonaggressive children. Participants also reported lower levels of anger for the prosocial stories, and higher levels of anger for hostile stories. There was also a main effect for aggression, with aggressive children feeling more anger than nonaggressive children. All children reported feeling more gratitude for the prosocial story, compared to all the other story versions. Regarding action tendencies, aggressive children were more likely to endorse the behavioral options of “get even” and “have it out right then and there,” compared to nonagressive children. The authors then used EQS to test an attributional explanation of aggression, which states that attributions cause emotions, which in turn cause action tendencies. The attributional model was the best fit to the data.

Conclusions: The attributional model provides a good explanation for childhood aggression. Aggressive children were more likely to perceive intention in ambiguous negative situations, and reported more feelings of anger in reaction to negative situations.

Commentary: Both aggressive and nonaggressive children inferred gratitude when the story stated that the other child was trying to help the protagonist. This provides further support that gratitude serves as an indicator that one has received a benefit from an external agent.

Correspondence: Sandra Graham, Graduate School of Education, University of California, Los Angeles, CA 90024.

Harris, M. B. (1972). The effects of performing one altruistic act on the likelihood of performing another. Journal of Social Psychology, 88, 65–73.

Objective: To determine whether engaging in a prosocial action in the absence of reinforcement will make one more likely to act prosocially in the future, and to compare the effects of reward, punishment, and no reinforcement on later prosocial behavior.

Design: Study 1 was a controlled field experiment. Study 2 was a laboratory experiment.

Setting: Study 1 was conducted on a street in the Albuquerque, New Mexico, area. Study 2 was conducted in classrooms from 12 classes at the University of New Mexico. These classes ranged from freshman to graduate level and were in four different departments.

Participants: Participants in Study 1 were 54 people walking down the street. Locations varied from campus to different shopping centers to busy streets. Participants in Study 2 were 276 students from the University of New Mexico.

Manipulated variables: In Study 1, participants were randomly assigned to one of three conditions. The experimenter stopped participants on the street. In the time condition, the experimenter asked the participant what time it was. In the directions (p.313) condition, the experimenter asked participants for directions to another street. When the time or directions were given, the experimenter repeated the information in a neutral tone and then asked participants for a dime. In the dime-only control condition, the experimenter said only, “Excuse me” and asked participants for a dime.

In Study 2, participants were asked to volunteer to write a letter to a high school student and offer to answer any questions the student might have about the university. After writing the letters, participants were randomly assigned to one of three response conditions. Participants in the no-response control condition did not receive responses to their letters. Participants in the positive-response condition and the negative-response condition received letters, ostensibly from the high school students to whom they had written. Letters in the positive-response condition expressed gratitude at participants’ offer to help them out. Letters in the negative-response condition contained negative comments in response to the participants’ offer of help.

Assessment of outcome variables: In Study 1, after the experimenter made his or her first request (if any), he or she then asked participants for a dime. The dependent measure was whether the participant gave the experimenter a dime.

In Study 2, participants received a second, unrelated request for help 2 weeks after the first request and 10 days after receiving a response, if any. A student made an announcement to all 12 classes asking for volunteers to help with a publicity campaign to inform people about the work the university does with the community. The dependent measure was whether participants signed up to help with this project.

Main results: In Study 1, more participants in the Time condition (44.4%) and the Directions condition (38.9%) gave the experimenter a dime, as compared with those in the control condition (11.11%). In Study 2, more students who had heard the original request to write letters (17.6%) signed up to help with the publicity campaign, as compared with students who were absent or in a control class and did not hear the first request (8.9%). However, there were no significant differences in subsequent helping between the positive-, negative-, and no-response conditions.

Conclusions: These studies suggest that helping someone makes a person more likely agree to subsequent requests for help, at least if the subsequent request occurs relatively close in time to the first request. This effect is independent of the type of feedback received, as participants who were both positively and negatively reinforced showed increased rates of subsequent helping. Nor is actual helping necessary to increase later helping: Participants in Study 2 who heard the first request but did not help (16.7%) still showed increased subsequent helping when compared with the students who were absent when the request was made (9.5%) or enrolled in another section (8.8%). The author explains these results using a social norm interpretation: Being asked for help makes a norm of social responsibility salient, causing an increase in later help.

Commentary: The results of this experiment contrast with other studies summarized in this bibliography that showed increased helping following a positive reinforcement (Clark et al., 1988; Moss & Page, 1972; Rind & Bordia, 1995). However, Harris found that this effect does not necessarily occur if the second request occurs a (p.314) relatively long time after the initial helping behavior. One aspect of the present study that is different from the others is the time lag between the positive and negative reinforcement and the second request (10 days). In contrast, the time lags for the Clark et al. and the Moss and Page studies were only a few minutes, whereas the time lag in the Rind and Bordia study was undetermined, but potentially only a day or two. These differences suggest that, if there is a reinforcing effect of gratitude, it may last only a short while. Further studies need to explore the effects over time that factors such as gratitude and the salience of social norms have on future prosocial behavior. Alternatively, it may be that the positive reinforcement of gratitude may only increase prosocial behavior toward the person who originally expressed the gratitude. Of the previously cited studies, the two that found effects for gratitude (Clark et al., Rind & Bordia) showed that participants were more likely to help the same person who expressed gratitude to them. Study 2 of the present set of experiments, which compared the effects of positive, negative, and no reinforcement, presented participants with opportunities to help two different sets of individuals. If, instead, participants were given a second opportunity to help the same individuals who gave them reinforcement, the effects of gratitude might be seen more clearly.

Correspondence: Mary B. Harris, Department of Educational Foundations, University of New Mexico, Albuquerque, New Mexico 87106

Harris, P. L., Olthof, T., Meerum Terwogt, M., & Hardman, C. E. (1987). Children’s knowledge of the situations that provoke emotion. International Journal of Behavioral Development, 10, 319–343.

Objective: To investigate children’s understanding of the situational determinants of complex emotions.

Design: Mixed model experiment. Researchers presented two studies; however, only Study 1 contained measures relevant to gratitude.

Setting: The setting is not explicitly stated; however it is assumed that children were tested at their respective schools.

Participants: Participants were 80 children from Oxford or Amsterdam. Although not directly noted, it is assumed that researchers recruited equal numbers of English and Dutch children.

Assessment of predictor variables: Twenty children each were recruited from each age group: 5 years, 7 years, 10 years, and 14 years. Children were presented with 20 emotion terms, including grateful. English-speaking children were presented stimuli in English, and Dutch children in Dutch.

Assessment of outcome variables: For each emotion term, children were asked to describe a situation that would evoke that particular emotion. Children ages 5 and 7 years were interviewed individually, whereas children ages 10 and 14 completed questionnaires. Two judges read children’s responses and rated the emotion that was most likely to have been the stimulus. Responses were given 1 accuracy point if one (p.315) judge picked the actual stimulus emotion, and 2 points if both judges picked the correct term.

Main results: Older children were more accurate than younger children in describing situations that elicited particular emotions. For English children, very few 5 year olds gave accurate descriptions of gratitude situations, whereas children 7 years and older were more accurate in their gratitude descriptions. Effects for gratitude in Dutch children were seen at age 10 and older. Cluster analyses revealed an “all-or-nothing” process of understanding emotion, with complex emotions (including gratitude) being acquired by children abruptly at later ages.

Conclusions: Results demonstrated that children were able to describe situations for more complex emotions that did not have distinctive facial expressions.

Commentary: These data speak to the complexity of gratitude. Unlike simple emotions such as happy or angry, gratitude takes longer for children to acquire. This study suggests that children understand the situations that elicit gratitude at around age 7 or older.

Correspondence: First author was at the University of Oxford at the time of this study.

Hegtvedt, K.A. (1990). The effects of relationship structure on emotional responses to inequity. Social Psychology Quarterly, 53, 214–228.

Objective: To examine the determinants of emotional responses to inequity. Specific to gratitude, this researcher predicts that overreward will be related to feelings of deservingness and gratitude, in addition to guilt. High status should be associated with feelings of deservingness, whereas low status should be associated with gratitude.

Design: Between–subjects experiment.

Setting: Classroom.

Participants: Participants were 118 female and 97 male undergraduate sociology students.

Manipulated variables: Participants were each given a vignette which described a student who needed a paper typed by a student typist. Participants imagined themselves in the role of the typist. Half of the vignettes portrayed the student seeking typing services as male, and the other half portrayed the student as female. Gender of the student was varied in order to manipulated status: Male participants typing for a female student were inferred to have higher status, whereas female participants typing for a male student were inferred to have lower status. Participants typing for same-sexed students were assumed to be equal in status. Vignettes were also written to vary in power: in the low power condition, the typist needed the job but few people needed typing services, and the student casually sought services among the many typists available. In the high power condition, the typist did not need the job and many students needed typing services, and the student desperately sought services from the (p.316) few available typists. Finally, equity was manipulated by stating that the going rate for the job that the student needed was $30, and that the student either paid the typist $30 (equity), $20 (underreward), or $40 (overreward).

Assessment of outcome variables: Emotional responses were measured via questionnaire, and included ratings of satisfaction, deservingness, gratefulness, anger, resentfulness, helplessness, and guilt.

Main results: Underrewarded participants reported more distress than equitably rewarded participants, and overrewarded participants showed the least amount of distress. Higher power participants who had been equitably or overrewarded rated themselves as more deserving than did low power participants who were similarly rewarded. In contrast, low power participants who were equitably or overrewarded rated themselves as more grateful than high power participants receiving the same pay. There were no effects of status on deservingness or gratitude.

Conclusions: Rather than a simple relationship between inequity and emotional responses, this researcher found that reactions to inequity depended not only on the outcome/input ratio, but on the individual’s power in the relationship.

Commentary: These results underscore the importance of expectations and deservingness in the experience of gratitude. Individuals who feel that they deserve beneficial treatment will be less likely to feel grateful for a benefit, whereas individuals who do not feel that they deserve a benefit will be grateful for it. One factor that determines feelings of deservingness and gratitude is relationship power, which in this study incorporated the need of the recipient and the value of the benefit.

Correspondence: Karen A. Hegtvedt, Department of Sociology, Tarbutton Hall at Emory University, 1555 Pierce Dr.,Atlanta, GA, 30322.

Jackson, L. A., Lewandowski, D.A., Fleury, R. E., & Chin, P. P. (2001). Effects of affect, stereotype consistency, and valence of behavior on causal attributions. Journal of Social Psychology, 141, 31–48.

Objective: To compare the effects of anger and gratitude with sadness and happiness on causal attributions when target behavior varied in stereotype consistency and valence.

Design: Randomized experiment.

Setting: Laboratory.

Participants: Participants were 229 Anglo-American undergraduate psychology students (133 women, 96 men).

Manipulated variables: Participants were induced to experience an angry, sad, grateful, happy, or neutral mood. Participants in the four affect conditions were asked to think of an event that evoked the target affect in vivid detail and write about the event on a blank piece of paper. Participants in the neutral condition were asked to (p.317) recall places they had been the previous day and write about those places and routes on a blank sheet of paper. Participants were then presented with a scenario about an African American man. This scenario manipulated stereotype consistency and valence. The stereotype-consistent positively valenced behavior presented the target as receiving a full athletic scholarship to play basketball at Duke University. Stereotype-consistent negatively valenced behavior presented the target as having been convicted of armed robbery. Stereotype-inconsistent positively valenced behavior presented the target as having received a full academic scholarship to Harvard University to study business administration. Stereotype-inconsistent negatively valenced behavior presented the target as having been convicted of computer theft and the diversion of corporate assets into a personal account.

Assessment of outcome variables: Causal attributions for the target’s behavior were measured using the revised Causal Dimension Scale (CDSII) (McAuley, Duncan, & Russell, 1992), which assesses locus of causality, stability, personal control, and external control.

Main results: Positive affect, including gratitude, caused participants to attribute causality more strongly to the target than to the situation, to make more stable attributions, and to view the behavior as more controllable by the target and slightly less controllable by others. When compared with happy participants, participants induced to feel gratitude viewed the cause of positive behavior as more stable and slightly more controllable by the target. In contrast, happiness was related to more stereotypic thinking when the behavior was negative.

Conclusions: Cognitive appraisal dimension is an important factor to consider when looking at the effects of affect on social judgments. Additionally, it is important to consider both positively and negatively valenced behaviors when looking at attributions of behavior.

Commentary: Although gratitude is an affect or emotion, it has important cognitive components that may have consequences for other cognitions. The differential effects that gratitude and happiness had on attributions of behavior also underscores the idea that gratitude is not reducible to general positive affect.

Correspondence: Linda A. Jackson, Department of Psychology, Michigan State University, East Lansing, Michigan 48824, e-mail: jackso67@msu.edu

Lane, J., & Anderson, N. H. (1976). Integration of intention and outcome in moral judgment. Memory and Cognition, 4, 1–5.

Objective: To explore whether moral judgments follow the same cognitive algebra as nonmoral decision making. Specifically, these researchers were interested in whether a benefactor’s intention in providing a benefit and the value of the benefit affect gratitude in a multiplicative manner.

Design: Randomized experiment.

Setting: Participants were tested individually in the laboratory. Sessions lasted about 30 minutes.

(p.318) Participants: Participants were 20 students at the University of California, San Diego.

Manipulated variables: Participants were given paragraph vignettes as well as simple assertions that varied in intentionality and value. Each story contained a description of either a high, medium, or low intention, as well as a high, medium, or low benefit value. In addition, paragraph vignettes contained an additional condition in which value was not specified. Thus, the paragraph variables were arranged in a 3 x 4 design, and simple assertions were arranged in a 3 x 3 design. Participants read 12 randomly ordered paragraph vignettes, followed by 15 randomly ordered simple assertions.

Assessment of outcome variables: Participants were asked to rate how grateful the average person would be for each description.

Main results: Results for both paragraph and simple statement stimuli showed evidence for an averaging rule rather than a multiplicative one. Information about intentionality and value were averaged to determine level of gratitude, with higher levels of intentionality and value leading to greater anticipated experiences of gratitude.

Conclusions: The authors concluded that individuals weigh information about intentionality and value similarly. The averaging rule found in these studies is not intuitive, because the two pieces of information were not superficially similar. The authors cited other studies in which an additive-type rule was found when a multiplicative rule was instead expected. They speculated that deservingness might mediate the effect that intentionality and value have on gratitude—that when a well-meaning person tries hard to help us but fails, we still feel that the person deserves our gratitude.

Commentary: This experiment shows that, although intentionality and value are sufficient to elicit gratitude, both may not be necessary. Specifically, high value without intention may still bring forth some feelings of gratitude, and vice versa. This study has the advantage of an experimental design; however, this advantage is somewhat offset by the fact that hypothetical vignettes were used rather than actual gratitude situations that participants experienced. A good next step would be to look at participants’ reactions to gratitude-eliciting stimuli in the laboratory.

Correspondence: Jeneva Lane, University of Oklahoma, Norman, Oklahoma 73069

McCullough, M. E., Emmons, R. A., & Tsang, J. (2002). The grateful disposition: A conceptual and empirical topography. Journal of Personality and Social Psychology, 82, 112–127.

Objective: To develop a measure of the disposition to experience gratitude, and to examine the correlates of gratitude as a personality trait.

Design: Correlational, using both self-reports and peer reports.

Setting: Studies 1 and 3 were conducted in undergraduate classrooms. Study 2 was based on a survey conducted on the Internet. Study 4 examined the extent to which (p.319) the relationships found in Studies 1, 2, and 3 remained when extraversion/positive affect, neuroticism/negative affect, agreeableness, and social desirability were controlled.

Participants: Participants in Study 1 were 238 undergraduate psychology students (174 women, 57 men, 6 unrecorded). Study 1 also included 639 informants who were peers or family of 168 of the original participants. Participants in Study 2 consisted of 1,228 visitors (80% women, 15% men, 5% unrecorded; mean age = 44.6, SD = 12.0, range = 18–75) to the Web site for the magazine Spirituality and Health (http://www.spiritualityhealth.com) or other Web sites linked to the Spirituality and Health Web site. Participants in Study 3 were 156 undergraduate psychology students. Study 4 was based on a reanalysis of the data collected for Studies 1, 2, and 3.

Assessment of predictor variables: The grateful disposition was measured using the newly developed Gratitude Questionnaire-6 (GQ-6). In Study 1, this six-item questionnaire was created from a pool of 39 self-report items. The resultant GQ-6 was again administered in Studies 2 and 3. Study 2 also included an additional three-item adjective measure of the disposition to experience gratitude (using the adjectives grateful, thankful, and appreciative).

Assessment of outcome variables: Outcome variables in Study 1 included self-report questionnaire assessments of the following variables: life satisfaction (SWLS; Diener, Emmons, Larsen, & Griffin, 1985), vitality (Ryan & Frederick, 1997), subjective happiness (Lyubomirsky & Lepper, 1999), optimism (LOT; Scheier, Carver, & Bridges, 1994), hope (Adult Hope Trait Scale; Snyder et al., 1991), positive and negative affect (PANAS; Watson, Clark, & Tellegen, 1988), anxiety and depressive symptoms (BFI; Derogatis & Spencer, 1982), dispositional empathy (empathic concern and perspective taking subscales of the Interpersonal Reactivity Index; Davis & Oathout, 1987), social desirability (BIDR; Paulhus, 1998), spiritual transcendence (STS; Piedmont, 1999), self-transcendence (Kirk, Eaves, & Martin, 1999) several single-item measures of religiousness, and the Big Five Inventory (John, Donahue, & Kentle, 1991). Study 1 also included peer reports of participants’ gratitude (using a 12-item gratitude scale drawn from the same item-pool as the GQ-6), frequency of participants’ prosocial action, participants’ general prosocial tendencies, and participants’ Big Five traits.

Study 2 also included identical self-report measures of positive and negative affect, life satisfaction, and spiritual transcendence. In addition, scores on the Big Five were assessed using Saucier’s (1994) Big Five Mini-Markers scale, and the disposition to forgive was measured using 10 items based on McCullough, Worthington, and Rachal’s (1997) theory of forgiveness.

Study 3 included identical measures of the Big Five Inventory (John, Donahue, & Kentle, 1991), anxiety and depressive symptoms, dispositional empathy, optimism, spiritual transcendence, and other religious variables. In addition, Study 3 also assessed materialism (Values-Oriented Materialism Scale—Richins & Dawson, 1992; and Belk Materialism Scale—Ger & Belk, 1990) and envy (Dispositional Envy Scale; Smith, Parrott, Diener, Hoyle, & Kim, 1999).

(p.320) Main results: In Study 1, factor analyses revealed that the 39-item pool of gratitude items loaded strongly on a single factor. From this item-pool, six items were chosen that loaded strongly on the first factor and assessed unique aspects of the grateful disposition. This GQ-6 scale demonstrated good internal consistency reliability (α=. 82). Discriminant validity was established by distinguishing the GQ-6 from measures of happiness, vitality, satisfaction with life, optimism, and hope. Self-reports and observer reports of grateful disposition were moderately positively correlated. Grateful disposition was moderately positively related to self- and peer reports of positive affect and well-being, prosociality, religiousness and spirituality, social desirability, extraversion, and agreeableness, and moderately negatively related to negative affect and neuroticism. When the GQ-6 was regressed on the Big Five, agreeableness predicted unique variance in GQ-6 scores.

Study 2 involved a cross-validation of the single-factor measurement model of the GQ-6. Both the GQ-6 and the adjective measure of grateful disposition were positively and moderately related to positive affect, life satisfaction, spiritual transcendence, forgiveness, agreeableness, conscientiousness, extraversion, and openness. Both gratitude measures were also moderately negatively related to negative affect and neuroticism. Agreeableness and neuroticism uniquely predicted scores on the GQ-6, whereas agreeableness and openness predicted unique variance in the adjective measure of the grateful disposition.

Study 3 provided additional support for the single-factor measurement model of the GQ-6. Dispositional gratitude was again positively associated with measures of positive affect and well-being, prosociality, spirituality/religiousness, and social desirability. Additionally, dispositional gratitude was negatively related to materialism and envy. Gratitude was positively correlated with agreeableness, extraversion, and conscientiousness, and negatively correlated with neuroticism. Agreeableness, neuroticism, and extraversion predicted unique variance in grateful disposition.

Results of Study 4 showed that relationships between the grateful disposition and other variables remained significant, though substantially reduced, when extraversion and positive affect were statistically controlled. The only exception was self-transcendence, whose correlation was not significant after controlling for extraversion and positive affect. Likewise, the relationships between grateful disposition and most other variables were reduced but still significant after controlling for neuroticism and negative affect. The only correlations that did not maintain their valence and statistical significance were correlations with anxiety and the possessiveness subscale of the Belk Materialism Scale. Similarly, correlations between grateful disposition and other variables, though reduced, maintained their valence and statistical significance after controlling for agreeableness, with the exception of the possessiveness and nongenerativity subscales of the Belk Materialism Scale. Last, correlations between the grateful disposition and other variables were reduced but still statistically significant after controlling for social desirability, with the exception of the success subscale of the Richins Materialism Scale.

Conclusions: Measures of the grateful disposition are related to other variables such as positive and negative affect, well-being, spirituality/religiousness, prosociality, materialism and envy. Though gratitude is related to different factors of the Big Five, (p.321) the Big Five do not account for all of the variance in the grateful disposition. The relationship between the grateful disposition and other variables remains even after controlling for extraversion, agreeableness, positive and negative affect, and social desirability.

Commentary: This set of studies provides a gratitude questionnaire with good internal consistency reliability and discriminant validity. The grateful disposition was related in theoretically expected ways to other variables of interest. These studies provide a tool for examining the grateful disposition and its relationship to outcomes such as health and well-being.

Correspondence: Michael E. McCullough, Department of Psychology, University of Miami, 248185 Coral Gables, Florida, 33124–2070; e-mail: mikem@miami.edu

Robert A. Emmons, Department of Psychology, University of California, 1 Shields Avenue, Davis, California 95616; e-mail: raemmons@ucdavis.edu

Moore, D. W. (1996). Americans most thankful for family and health: Youth also thankful for career/job. Lincoln, NB: The Gallup Poll Monthly.

Objective: To explore what U.S. citizens list as the things they are most thankful for as the Thanksgiving season approaches.

Design: Telephone interview polls.

Setting: Polls were taken from U.S. homes from November 21 to November 24, 1996.

Participants: A randomly selected national sample of 1,003 adults, aged 18 years and older.

Assessment of predictor variables: Demographic information collected included participant age, ethnicity (black or white), and income.

Assessment of outcome variables: Interviewers asked participants, “As the Thanksgiving holiday approaches, we’d like to know what two or three things are you most thankful for in your life right now?” Participants were asked to provide up to three responses to this open-ended question.

Main results: The most common response was to mention family (61%), followed by own health (50%), job/career (21%), child/children (20%), spouse (12%), my life/just being alive (12%), freedom/living in the U.S. (10%), friends (8%), income/financial security (7%), home (7%), and God (5%). Older people were more likely to mention being thankful for their health, whereas younger people were more likely to mention their jobs or careers. More blacks mentioned being thankful for being alive (22%) than did whites (9%). This effect was also related to income, with lower income families more likely to be thankful for being alive.

Conclusions: U.S. citizens are most thankful for their families and their health, with certain demographic differences.

(p.322) Commentary: This survey helped to flesh out the objects of many people’s gratitude. It included individuals of many different ages, and compared ethnicities. A good first step in the study of gratitude is to find out what people are grateful for.

Moss, M. K., & Page, R.A. (1972). Reinforcement and helping behavior. Journal of Applied Social Psychology, 2, 360–371.

Objective: To determine the effects of positive versus negative reinforcement, as well as attractiveness of the reinforcer, on subsequent helping behavior.

Design: Controlled field experiment.

Setting: A main street in Dayton, Ohio, similar to many other metropolitan shopping areas.

Participants: Participants were 140 white individuals between the ages of 18 and 60, walking individually along the street. There were equal numbers of men and women.

Manipulated variables: Two female confederates administered the manipulation and dependent measures. The first confederate walked up to a participant and asked for directions to a popular local department store. After the participant gave directions, the confederate gave either a positive, neutral, or negative statement in response to the help. The positive response consisted of saying “thank you” and smiling. In the negative condition, the confederate cut the participant off just before he or she finished giving directions and said “I can’t understand what you’re saying, never mind, I’ll ask someone else” (Moss & Page, 1972, p. 363). In the neutral condition, the confederate said “Okay” after receiving directions. There was also a control condition in which participants did not interact with the first confederate before encountering the second confederate. Attractiveness of the first confederate was also varied, so that she appeared attractive for half of the participants and unattractive for the other half.

Assessment of outcome variables: A second confederate was stationed about 75 feet from the first confederate. The second confederate began to walk toward the participant, and when she was within 6 feet of the participant, she dropped a small bag and continued walking down the street. The dependent measure consisted of whether the participant helped the second confederate by either picking up the bag or calling out to her.

Main results: Participants who received a negative response from the first confederate were less likely to offer subsequent help (43% helped) than participants in other conditions. There were no significant differences in helping between participants in the positive (93%), neutral (88%), and control (90%) conditions. There was an effect for attractiveness of the first confederate, in that participants were more likely to give physical help to the second confederate after interacting with an attractive first confederate, and they were more likely to give verbal help after interacting with an unattractive first confederate.

(p.323) Conclusions: Individuals who are negatively reinforced for helping are less likely to give help in the future. The authors explained the lack of results for positive reinforcement by pointing to a possible ceiling effect in helping.

Commentary: We concur with the authors’ explanation for the lack of effects of positive reinforcement. The prosocial behavior in this experiment involved little cost to participants, and therefore many participants in the control condition ended up helping. The effects of gratitude might be more clearly seen with helping behaviors that occur with less naturalistic frequency. Additionally, it may be that gratitude may only be effective as a positive reinforcer for prosocial actions toward the same reinforcing agent. In other words, expressed gratitude might have increased helping toward the first confederate but left behaviors toward the second confederate unchanged.

Correspondence: Martin K. Moss, Department of Psychology, Wright State University, Dayton, Ohio 45431

Okamoto, S., & Robinson, W. P. (1997). Determinants of gratitude expressions in England. Journal of Language and Social Psychology, 16, 411–433.

Objective: To explore factors affecting expressions of gratitude in England. These researchers hypothesized that expressions of gratitude would increase in politeness as the benefactor experienced more imposition when helping the receiver.

Design: Study 1 was a between-subjects field design. Study 2 was a questionnaire study.

Setting: Study 1 took place at the doors of the main library of the University of Bristol. Study 2 was administered during an introductory psychology class at the University of Bristol.

Participants: In Study 1, participants were 228 people (108 males, 120 females), mostly students, passing through doors of the main library of the University of Bristol. In Study 2, participants consisted of 120 undergraduates (39 males, 81 females) from the same university.

Manipulated variables: In Study 1, the experimenter opened doors for people passing singly through the doors. The experimenter held the door open in one of four ways, each way increasing in imposition to himself: 1) Experimenter going in same direction as participant, without looking back after opening the door (least imposition), 2) Experimenter going in same direction as the participant and making eye contact after opening the door; 3) Experimenter going in the opposite direction as participant, with the experimenter going through the door first, then holding open the door and making eye contact with the participant; and 4) Experimenter going in the opposite direction from the participant, letting the participant in first and making eye contact (most imposition).

For Study 2, participants were presented with one of two questionnaires that contained different versions of six randomly ordered vignettes. In the vignettes, a giver does a favor for the receiver. The vignettes were written to correspond to three (p.324) responsibility conditions (neither person responsible for giver’s behavior, receiver responsible, and giver responsible) along with two imposition conditions (large and small). For example, the neither/large vignette consisted of a giver lending valuable photos for the receiver’s exhibition; for the receiver/small condition, the vignette described the giver’s picking up a pen that the receiver had just dropped; for the giver/large condition, the vignette described the giver’s returning money that was borrowed from the receiver, long after he or she had agreed to repay it. Participants were asked to imagine that they were the receiver in these scenarios under both of two status conditions (equal-status giver and high-status giver).

Assessment of outcome variables: In Study 1, the experimenter wrote down what participants said after the door was held open. In the condition in which the experimenter did not make eye contact with the participant, an observer noted what the participant said. These verbal expressions were coded for politeness.

In Study 2, participants wrote down what they would say in response to each vignette, depending on the status of the giver. Responses were coded for politeness. Participants also rated the imposition present in each vignette if a friend were the giver.

Main results: In Study 1, gratitude was expressed least frequently in the minimal imposition condition, where eye contact not made and the experimenter was the least imposed on. Accordingly, gratitude was expressed most frequently in the condition that posed the greatest imposition on the experimenter. When comparing the three conditions with the greater imposition, the condition with the most imposition elicited the most polite expressions of gratitude, and the condition that contained the least imposition of the three elicited more colloquial forms of gratitude.

Greater imposition led to greater politeness in Study 2 as well. For both equal-status and higher status givers, the greater the imposition, the longer the gratitude expressions and the more modifiers they contained. However, politeness was lower in response to giver-responsible vignettes. Along with gratitude expressions, researchers also looked for apology expressions in response to the vignettes. Expressions of apology were used most often in receiver-responsible vignettes but least often in the neither-responsible vignettes.

Conclusions: These studies reveal that more polite forms of gratitude expressions are used for favors of greater imposition. This effect occurred across different measurement designs and manipulations, with both equal-status and high-status benefactors. However, this effect was not present when the receiver was responsible for the imposition.

Commentary: Benefits that cause larger imposition to the benefactor call for more politeness and more expressed gratitude than smaller impositions. Future research can explore whether individuals who do not express the requisite level of gratitude in a situation with large imposition are seen as ungrateful, or whether there is a corresponding increase of actual experienced gratitude with increased imposition.

Correspondence: Shinichiro Okamoto, Department of Psychology, Faculty of Letters, Aichi Gakuin University, Nisshin-shi, Aichi 470–01, Japan, e-mail: okamoto@ dpc.aichi-gakuin.ac.jp

(p.325) Pyke, K., & Coltrane, S. (1996). Entitlement, obligation, and gratitude in family work. Journal of Family Issues, 17, 60–82.

Objective: To investigate how experience in a previous marriage affects feelings of gratitude and obligation in response to the division of household labor in second marriages.

Design: Correlational.

Setting: Questionnaires were filled out by participants at home and mailed back to the researchers. Interview setting was not specified.

Participants: Participants consisted of 97 husbands and 96 wives who had remarried. Spouses of participants were not allowed to participate in the study. Further qualifications included having at least one child in the home half the time, having a first marriage that had lasted one year or longer, and a second marriage that occurred at least one year ago. These 193 participants represented the survey sample. From this sample, 70 participants were selected for a follow-up interview.

Assessment of predictor variables: Predictor variables were assessed via questionnaire and interview questions. Predictor variables from the interviews included extramarital affairs during the first marriage, and the use of social comparison to previous marriages by participants. Predictor variables assessed in the questionnaires included family/gender ideology, employment hours, total household income, extramarital affairs during the first marriage, wife’s earnings, wife’s age, and presence of preschool children in the household.

Assessment of outcome variables: Dependent variables in the interview section included assessment of the proportion of household labor the husband contributed in a relationship (as rated by husbands or wives), and feelings of entitlement, obligation, and gratitude experienced by participants in their marriages. The dependent variable in the survey data was also the proportion of household labor contributed by the husband.

Main results: Interview data revealed that women tended to compare their second marriage with their first marriage, and this mitigated feelings of entitlement and bolstered feelings of gratitude. The authors also provided examples of men who compared their second marriage to their first. In some instances, men engaged in social comparison in order to justify a more traditional division of labor in their second marriage, and other times as an explanation for the high importance they now placed on egalitarianism. Interview data also revealed an effect for first marriage extramarital affairs by men. Women whose first husbands had engaged in extramarital affairs tended to experience gratitude at their current husband’s fidelity. Similarly, men who had cheated in their first marriage felt entitled to gratitude from their wives for their current fidelity.

Questionnaire data revealed that remarried wives spent twice as much time on household labor as husbands. Husbands with more egalitarian family ideologies, and husbands who worked fewer hours contributed more to household labor. Wife’s earnings and husband’s family ideology were positively related to husband’s household (p.326) labor, whereas wife’s family ideology and husband’s first-marriage extramarital affair were negatively related to husband’s proportion of household labor.

Conclusions: Experiences from prior marriages, such as experiences of husband’s extramarital infidelity, affected division of household labor in second marriages, and feelings of entitlement and gratitude for the spouses’ efforts in the home.

Commentary: Although this study did not contain an actual quantitative measure of gratitude, it still made some important theoretical and empirical points about the determinants of gratitude. Whereas results from Teigen (1997) and Fredrickson et al. (2003) focus on the role of “what could have been” in eliciting gratitude, this study shows that social comparison to “what was before” can also affect gratitude and related emotions. The results suggest that social comparison to past relationships might be an important determinant of gratitude in current relationships. Additionally, this study highlights the importance of individual construal in gratitude: the cost and value of various benefits are not static and objective—different benefits are worth more to some people than others, and some people feel entitled to more benefits than others. These differences in construals can be affected by factors such as previous relationship experience or sex role ideology, and they determine whether someone responds with gratitude to the action of a benefactor.

Correspondence: Contact information not provided. The first author was associated with the University of Southern California at the time of this study.

Rind, B., & Bordia, P. (1995). Effect of server’s “thank you” and personalization on restaurant tipping. Journal of Applied Social Psychology, 25, 745–751.

Objective: To determine whether a server’s expression of gratitude through writing “thank you” on the back of the check would increase the server’s tip amount. Researchers also tested if adding a personal signature after writing “thank you” increased tips.

Design: Controlled field experiment.

Setting: Lunch hours (11:00 a.m. to 3:00 p.m.) at an upscale restaurant in Philadelphia, located on the University of Pennsylvania campus. The experiment was conducted in late spring over a 5-day period, from Monday through Friday.

Participants: Participants were 51 dining parties having lunch. Many of the participants were faculty and other university staff. Total participants were 137 customers, with a mean of 2.69 customers per party.

Manipulated variables: Dining parties were randomly assigned to one of three conditions. In the control condition, the server delivered the check without writing anything on the back of it. In the thank-you condition, the server wrote “thank you” on the back of the check. In the thank-you-plus-name condition, the server wrote “thank you” on the back of the check and signed her name directly under it. After writing the (p.327) message, if any, on the check, the server placed the check on the party’s table, face down, and had no further interactions with the party.

Assessment of outcome variables: The server recorded the tip amount, bill before taxes, size of the dining party, and method of payment for each party.

Main results: Tip amounts were larger for the thank-you condition (18.10%) and the thank-you-plus-name condition (18.01%) when compared with tip amounts in the control condition (16.28%). The two experimental conditions did not differ significantly. Tip amounts did not differ by party size or method of payment.

Conclusions: The authors framed the results in terms of impression-management theory, positing that expressing gratitude increased the server’s perceived likability and friendliness, leading to increased influence and greater tip percentages. They listed several alternative explanations, such as increased perception of servers’ expectations for tips; self-perception; or reciprocity

Commentary: This study supports the theory of gratitude as a moral reinforcer (McCullough, Kilpatrick, Emmons, & Larson, 2001). Benefactors—dining parties—who became the targets of gratitude from the server later increased their prosocial behavior by leaving higher tips. The strong point of this study is that it looks at behavioral outcomes of gratitude in a field setting. However, because perceived gratitude was not actually measured, the exact mechanisms driving the results of this study are not clear-cut. This shortcoming is small, because this study speaks to expressed rather than perceived gratitude.

Correspondence: Bruce Rind, Department of Psychology, Temple University, Philadelphia, Pennsylvania 19122, e-mail: rind@templevm

Russell, J. A., & Paris, F. A. (1994). Do children acquire the concepts for complex emotions abruptly? International Journal of Behavioral Development, 17, 349–365.

Objective: This set of two studies looked at the course of development of more complex emotions, such as gratitude or pride. Study 1 sought to determine whether children acquire concepts for complex emotions abruptly or in a gradual manner. Study 2 investigated the hypothesis that children initially understand complex emotions in terms of the bipolar dimensions of pleasure and arousal.

Design: Study 1 was a laboratory experiment employing a within-subjects, cross-sectional design. Study 2 was also a laboratory experiment, using a mixed-model, cross-sectional design with participants randomly assigned to rate a number of different emotions on one of two bipolar scales.

Setting: In Study 1, children were tested individually by an experimenter. Although the location was not explicitly stated, this study most likely was conducted in a (p.328) laboratory. In Study 2, children were tested in a laboratory setting, whereas adults were presented the rating scales individually in public places such as shopping malls.

Participants: Study 1 participants were 96 children—12 girls and 12 boys at each of four ages: 4, 5, 6, and 7 years. Study 2 participants were 20 boys and 20 girls at each of two ages (4 and 5 years), as well as 20 women and 20 men older than 16 years.

Assessment of predictor variables: The predictor variables in Study 1 were the child’s age and the specific emotion the child was asked to describe. The complex emotions examined were proud, grateful, jealous, ashamed, and worried. Happy was also included as a noncomplex control emotion. In Study 2, the predictor variables were the participants’ age, the scale on which participants were rating emotions (pleasure or arousal), and the specific emotion participants were rating: complex emotions—grateful, proud, ashamed, jealous, and worried; and simple emotions—happy, calm, and sad.

Assessment of outcome variables: In Study 1, the experimenter asked children to tell a story about a fictional child who was feeling one of the tested emotions. The child was instructed to explain why the target child was feeling that emotion, and whether the target child was feeling good or bad. Children were asked to tell a separate story regarding each emotion. Their responses were coded first on the good/bad dimension (good = happy, proud, grateful; bad = ashamed, jealous, worried). Judges then rated whether the story the child gave for each emotion was appropriate to the given emotion, and a content analysis was done for each emotion. A correct response for grateful appeared when a story included “a positive event (statement, action, affection) from someone” to the target child (p. 353, Table 1). A modified version of the Harris, Olthof, Meerum Terwogt, and Hardman (1987) best guess procedure was also employed to differentiate emotion stories.

In Study 2, children were told that a stick figure was feeling a particular emotion and were asked to place this figure into one of five boxes corresponding to different ratings of either pleasure or arousal. Adults filled out a comparable rating questionnaire of either pleasure or arousal. Participants rated all eight emotions on the particular scale they were given.

Main results: In Study 1, results did not support either a strong or weak abruptness hypothesis; rather, they supported a gradual acquisition hypothesis. Children in the youngest age groups told better stories for some complex emotions than for others, with jealous producing the worst stories at the youngest ages, and gratitude producing slightly better stories. Additionally, the oldest children (7-year-olds), though they exhibited greater knowledge of the complex emotions, did not show complete knowledge when their stories about complex emotions were compared with their stories about the noncomplex emotion, happy. Many children showed a partial knowledge of complex emotions, their stories demonstrating knowledge of the emotion’s positive or negative valence.

In Study 2, patterns of pleasure and arousal ratings given by children were similar to those given by adults, though the average pleasure and arousal scores did differ by age. For all age groups, the emotion grateful was rated above the mean on both pleasure and arousal.

(p.329) Conclusions: The authors concluded that complex emotions are learned gradually, rather than abruptly. Younger children have some knowledge about the nature of complex emotions, such as their dimensions of arousal or pleasure. Older children, such as 7-year-olds, have a greater knowledge of complex emotions but do not have complete knowledge.

Commentary: These studies are an important step in determining the acquisition patterns of more complex emotions such as gratitude. They provide evidence that knowledge about gratitude is not acquired abruptly; rather, there is at least one intermediate stage in which children know general information about gratitude, such as its arousal and pleasure level. They demonstrate that children as young as 4 can have at least a rudimentary idea of what gratitude is like as an emotion.

Correspondence: James A. Russell, Department of Psychology, University of British Columbia, Vancouver, British Columbia, Canada V6T 1Y7, fax (604) 822 6923, email: jrussell@cortex.psych.ubc.ca

Saucier, G., & Goldberg, L. R.(1998). What is beyond the Big Five? Journal of Personality, 66, 495–523.

Objective: To find clusters of adjectives for describing people that are independent of the Big Five personality factors.

Design: Factor-analysis.

Setting: Not explicitly stated, but it is assumed that participants completed ratings in a classroom or related setting.

Participants: Participants were distributed among four original samples and one cross-validation sample. Of the original samples, Sample 1 consisted of 320 participants and 316 of their peers. Sample 2 included 187 participants. Sample 3 consisted of 360 participants and 329 of their peers. Sample 4 consisted of 201 peer ratings. The cross-validation sample consisted of 694 participants, about 57% of whom were women, with an average age of approximately 50.

Derivation of clusters: The authors used a combination of factor analysis and intuitive methods to derive 53 clusters to identify sources of variance peripheral to the Big Five.

Cross-validation: These 53 clusters of adjectives were rated by the cross-validation sample in two iterations using a 7-point scale.

Main results: A “minimax” criterion was used to determine whether a given cluster was peripheral to the Big Five: minimum multiple correlation with the Big Five factors, and maximum reliability. Using this criterion, the six clusters that stood out as being independent were short-tall, busy-overworked, employed-unemployed, religious-nonreligious, young-youthful, and slim-slender.

(p.330) In contrast, the cluster grateful-thankful was not independent of the Big Five. Its multiple correlation with the Big Five was. 40, and it was especially positively correlated with agreeableness (r =. 31) and negatively correlated with openness (r = -.24).

Conclusions: “Gratefulness” (gratitude) does not seem to be an independent personality trait. Instead, the tendency to feel grateful is related to other personality characteristics, such as agreeableness and openness.

Commentary: Though not an independent personality dimension in its own right, gratitude might be thought of as the emotional and behavioral outcome of a combination of Big Five traits.

Correspondence: Gerard Saucier, Department of Psychology, 127 University of Oregon, Eugene, Oregon 97403, email: gsaucier@oregon.uoregon.edu

Teigen, K. H. (1997). Luck, envy, and gratitude: It could have been different. Scandinavian Journal of Psychology, 38, 313–323.

Objective: To investigate the relationship between good or bad luck, feelings of envy and gratitude, and the role of counterfactual thinking. Study 1 explored the meanings behind people’s statements of luck in comparison to statements of goodness. Study 2 investigated the extent to which luck implied comparison with others, envy, gratitude, sympathy, and positive or negative impressions of the speaker. Study 3 tested whether the experience of gratitude makes people feel lucky, and whether the experience of envy causes people to feel unlucky and that someone else has been lucky.

Design: All three studies were within-subjects laboratory experiments.

Setting: Though not explicitly stated, it is assumed that all studies took place in a classroom or laboratory setting.

Participants: Study 1 participants consisted of 60 first-year psychology students at the University of Bergen, Norway. Study 2 participants were 262 first-year psychology students at the same university. Study 3 participants were 60 students from the same university.

Manipulated variables: Participants in Study 1 were presented with two pairs of statements. For each pair, one statement included the term lucky, and the other statement included the word good. For example, one pair of statements was as follows: “It is lucky that you have a job” and “It is good that you have a job.” One pair of statements was phrased in first person: “I am lucky that…” whereas the second pair was phrased in second person “You are lucky that…”

Participants in Study 2 were presented with two out of 16 statement pairs. There were four sets of phrases, with each set having permutations: lucky/good second per-son—“It is lucky that you have a job” and “It is good that you have a job,” and lucky/good first person—“It is lucky that I have a job” and “It is good that I have a job,” and so on (unlucky/bad second person and unlucky/bad first person).

(p.331) Participants evaluated one second-person and one first-person pair, with one statement in each pair being negative and the other positive.

Participants in Study 3 were asked to write two brief descriptions of situations in which they felt grateful for something. One description was to be of a situation in which participants had felt grateful to a specific person, and the other description was to be of a situation in which they felt a more abstract gratitude, perhaps toward life, or fate. After writing these descriptions of gratitude, participants were asked to write about a situation in which they felt envious of someone.

Assessment of outcome variables: Study 1 participants were asked to give a brief explanation of the meaning of both statements in each of the two pairs presented.

Study 2 participants rated each statement (using a 3-point scale) on whether it expressed sympathy, envy, comparison with others, and expressing/requesting gratitude. They also rated their impressions of the person issuing the statement, on a 3-point scale.

After completing the descriptions of gratitude and envy, participants in Study 3 completed questionnaires about each description. For the gratitude descriptions, participants rated the extent to which they felt unlucky and lucky, the extent to which the situation was pleasant or unpleasant, and whether something else could have easily happened. Participants also answered an open-ended question regarding other possible outcomes of the situation and rated this possibility in terms of pleasantness and unpleasantness. For the envy description, participants answered the same question, except that they rated the extent to which they felt someone else was lucky, rather than themselves.

Main results: In Study 1, several respondents perceived second-person luck statements to be more judgmental and demanding that the lucky person feel grateful, whereas similar goodness statements were perceived to be more neutral. First-person luck statements were also more likely than first-person goodness statements to elicit comments implying gratitude. In Study 2, positive first-person luck statements were seen as more expressive of gratitude than positive first-person goodness statements.

In Study 3, when asked to write descriptions of gratitude, most participants wrote about being grateful to specific people (personal gratitude) but seemed to have no problem generating descriptions of being grateful to abstract entities (impersonal gratitude). Of the 60 personal gratitude stories, 27 were about close friends, 12 were about parents or other relatives, and 21 were about strangers or more distant acquaintances. For strangers, the main elicitor of gratitude was lack of expectation, as when a stranger helps beyond the call of duty. Other objects of gratitude were concrete assistance, services, pleasant surprises, or emotional support. Of the impersonal gratitude stories, 27 cases were about particular episodes, such as dramatic situations that turned out better than expected, and 32 were more permanent descriptions of affairs, such as having healthy children or being alive in general. Both types of gratitude situations were described as lucky, and luck was independent of perceived pleasantness of the situation. Gratitude situations also elicited high ratings of “it could easily have been otherwise,” and counterfactual alternatives were always rated as worse than (p.332) what actually happened. In contrast, when people rated envy situations, they rated themselves as unlucky, whereas the objects of their envy were thought of as very lucky. Envy situations also elicited high counterfactual ratings, but these alternatives were always thought to be more positive than reality.

Conclusions: This study demonstrated the importance of counterfactual thinking in the relationship between attributions of luck, and envy and gratitude. Envy is felt if “things could have been better,” whereas gratitude is often felt if “things could have been worse.” In instances of envy, the self is seen as unlucky, whereas the self is seen as lucky in cases of gratitude. Although attributions of positive outcomes to external agents are an important source of feelings of gratitude, gratitude may be elicited by counterfactual thinking and feelings of luck.

Commentary: This set of studies is important for a number of reasons. First, Study 3 gives a sketch of the content of grateful descriptions. More important, it shows how different factors—attributions, counterfactual thinking, comparisons—play complex roles in the arousal of gratitude. Although external, intentional attributions for success are sufficient to elicit gratitude, they may not be the only possible antecedents of grateful emotion.

Correspondence: Karl Halvor Teigen, Department of Psychology, University of Tromsø, N-9037 Tromsø, Norway, email: karlht@psyk.uit.no

Tesser, A., Gatewood, R., & Driver, M. (1968). Some determinants of gratitude. Journal of Personality and Social Psychology, 9, 233–236.

Objective: To examine the effects of intention of benefactor, cost to benefactor, and value of benefit on receiver’s feelings of gratitude.

Design: Randomized experiment.

Setting: Classroom of undergraduates at Purdue University.

Participants: Participants were 126 undergraduate men.

Manipulated variables: Three scenarios were written, with each scenario reflecting every possible combination of intention, cost, and value at three intensity levels. This led to 27 different vignettes. Each participant was asked to read one randomly assigned vignette from each of the three scenarios, imagining himself or herself as the receiver of the benefit in the scenario.

Assessment of outcome variables: Following each scenario were a number of questions. Participants were instructed to answer the questions from the point of view of the receiver. Manipulation checks asked about intentionality, cost, and value. Two 6-point scales were used to assess gratitude: Participants rated how grateful they felt toward the benefactor, and how indebted they felt.

(p.333) Main results: The two items asking about felt gratitude and indebtedness were highly correlated, and the authors combined these items as a measure of gratitude. There was a significant main effect on felt gratitude for intention, cost, and value, and there were no interactions. Further analysis revealed that felt gratitude was a linear function of each of the independent variables. Regressions run separately on each scenario showed that intentionality and value significantly predicted gratitude with all three scenarios, but that cost was significant for only two scenarios.

Conclusions: These data support the hypothesis that intentionality, cost, and value are significant in determining feelings of gratitude.

Commentary: This study’s strong point is the use of a randomized experiment to directly test hypotheses about gratitude. Its results show the importance of the factors of intentionality, cost, and value. However, a weakness in this study is its use of scenarios instead of actual gratitude situations. Regardless, this experiment is an excellent starting point in the study of gratitude.

Correspondence: No mailing address listed. The first author was affiliated with the University of Georgia at the time of this publication.

Van Overwalle, F., Mervielde, I., & De Schuyter, J. (1995). Structural modelling of the relationships between attributional dimensions, emotions, and performance of college freshman. Cognition and Emotion, 9, 59–85.

Objective: To use structural equation modeling to test Weiner’s (1986) attribution model of emotions.

Design: Correlational.

Setting: Classroom.

Participants: Study 1 participants were 585 undergraduate students at two universities in Belgium. Study 2 participants were 621 undergraduate students from the same universities.

Assessment of predictor variables: In Study 1, participants were given questionnaires during class a few days after receiving scores on their midterm exams. Participants were asked for their midterm score, and rated their expectations about final exams, emotional responses to their midterm exams, and causal attributions about their midterms. The emotional dimension of gratitude was assessed using the items “gratefulness,” “trust,” and “appreciation” (p. 71). The causal attribution assessed that was relevant to gratitude was an external attribution of success to others.

Study 2 measures were similar to those in Study 1, with the exception of the following: Participants were asked the extent to which they experienced gratitude and anger “toward the teachers who examined” (p. 76), and participants were asked attributional questions about external control that referred to “teachers who examined” rather than “other people” (i.e., “The cause of something is something I/teachers who examined can(not) do something about”) (p. 76)

(p.334) Assessment of outcome variables: In Study 1, academic performance was assessed through participants’ scores on final exams. Study 2 outcome variables were identical to those in Study 1.

Main results: In Study 1, gratitude was significantly positively correlated with happiness, pride, hope, and expectation. It was not significantly correlated with other, negative emotions such as shame and anger. Structural equation models showed that gratitude was significantly related to a positive outcome on the midterm, but counter to predictions it was not significantly correlated with external attributions to others. (Another hypothesized effect between external attributions to others and anger also failed to materialize.)

In Study 2 the relationship between gratitude and midterm grade failed to reach significance. Like Study 1, there was no relationship between external control and gratitude and anger. However, when participants were divided into groups based on positive or negative outcome on their midterm exams, effects for gratitude and anger emerged: For participants who did well on their midterms, the path from gratitude to external control was significant and positive.

Conclusions: The authors explained the lack of significant effect between external control and gratitude and anger in Study 1 by suggesting that external attribution questions asking about “other persons” was too general. This led to the use of “the teacher” as a referent in Study 2. The results of Study 2 emphasized the importance of taking outcome into account when investigating social emotions such as anger and gratitude. In general, the authors found support for the relationship between attributions and emotion and for Weiner’s (1986) idea of outcome-related and attribution-related emotions.

Commentary: These studies provide support for the proposal that gratitude is elicited by positive outcomes that are attributed to others, and underscore the positive emotional valence of gratitude.

Correspondence: Frank Van Overwalle, Pleinlaan 2, B-1050 Brussels, Belgium.

Ventimiglia, J. C. (1982). Sex roles and chivalry: Some conditions of gratitude to altruism. Sex Roles, 8, 1107–1122.

Objective: To investigate the effects of sex role on grateful responses of recipients of chivalrous action.

Design: Between-subjects experiment.

Setting: This study is a field experiment conducted in the entrances of a city library and a university library.

Participants: Participants were 479 individuals, 52% men and 48% women.

Assessment of predictor variables: The experimenter approached library doors shortly before the participant arrived, and held open the door for the participant. An observer (p.335) recorded relevant predictor variables, including the participant’s physical attractiveness, age, race, and need state. Setting was manipulated by gathering half of the data outside of a city library, and half of the data outside a university library. Sex of the benefactor was manipulated by using one male experimenter and one female experimenter.

Assessment of outcome variables: The observers coded participants’ responses to the favor.

Main results: The two most frequent positive responses to door opening were thanking the benefactor, and smiling, whereas the two most frequent negative responses were obliviousness to the favor and hesitation. Examining these four responses, female participants were more likely to have higher percentages of positive responses and lower percentages of negative responses, when compared to male participants. Male experimenters were more likely to elicit negative responses from participants than were female experimenters. Setting (city vs. university library) had no interpretable effect on these positive or negative responses, aside from an effect of increased disapproval in city versus university libraries.

Coders used four categories to further classify participant responses: gratitude, confusion, disapproval, and avoidance. Gratitude responses included participant expressions of thanks, reciprocation, and smiles and nods directed at the experimenter; confusion responses consisted of puzzled looks, disagreements, and blushing; disapproving responses included frowns, laughter, and obliviousness toward the benefactor; and avoidance responses consisted of avoidance of eye contact with the experimenter, and hesitation. Participants demonstrated more gratitude in response to the favor when they had a higher need state, were older, more attractive, and female. More disapproval was elicited when the participant was entering the university library, when the person holding open the door was male, and when the participant was more attractive. Confusion occurred most often when the participant was male, or when the experimenter was female. Avoidance was more likely when the participant was physically attractive or male. Whereas female participants showed more gratitude, male participants tended to emit more confusion and avoidance.

Regarding the effect of sex roles on responses to a favor, male participants helped by a woman showed more confusion than did female participants helped by a man. Women helped by another woman exhibited more gratitude, less disapproval, and less avoidance than did men helped by another man.

Looking at the impact of physical attraction on the sex role effect on responses, unattractive female recipients helped by male benefactors showed the most gratitude, whereas unattractive male recipients helped by a female benefactor exhibited the least gratitude. With confusion, attractive female recipients helped by a male benefactor were least confused, whereas attractive male recipients helped by a female benefactor were most confused. Looking at same-sex recipient-benefactor pairs, attractive male and female recipients both showed more disapproval at a benefactor of the same sex, when compared with less-attractive participants.

Conclusions: The most frequent response to having a door held open was the expression of gratitude, either by saying “Thanks” or smiling or nodding. The effect of sex (p.336) roles on gratitude—the most gratitude occurred when a man held open a door for a women—demonstrates that traditional norms regarding sex roles still affect individual’s responses.

Commentary: Although these results are to a certain extent tied to the context of the “door-opening ceremony” (p. 1122), they underscore the importance of normative context in the study of gratitude. A benefit or favor that has negative ramifications for one’s self-image, whether it be in the realm of sex roles, self-efficacy, or some other domain, may not be interpreted as a benefit, effectively tempering any gratitude effects.

Correspondence: The author was affiliated with Memphis State University at the time of this study.

Weiner, B., Russell, D., & Lerman, D. (1978). Affective consequences of causal ascriptions. In J. H.Harvey, W. J. Ickes, & R.F. Kidd (Eds.), New directions in attribution research (Vol. 2). Hillsdale, N. J.: Erlbaum.

Objective: To investigate the emotions elicited in an achievement context, and to examine the effects of internal and external attributions on these emotional responses.

Design: Between–subjects experiment.

Setting: Laboratory.

Participants: 90 undergraduate students.

Manipulated variables: Participants were given two achievement scenarios: one with a positive outcome, and one with a negative outcome. The scenario relevant to gratitude was the positive outcome scenario. Scenarios varied by the attributions presented for the outcome. Positive outcome scenarios contained one of following attributions: ability, personality, unstable effort, stable effort, other’s effort, other’s motivation and personality, task difficulty, luck, mood, and intrinsic motivation.

Assessment of outcome variables: After each scenario, participants rated 85 emotions that the researchers predicted would be related to success or failure. Participants only rated success-relevant emotions for the positive outcome scenario, and failure-relevant emotions for the negative outcome scenario.

Main results: Researchers listed the 10 most frequent emotions experienced for each attribution. Looking at the positive outcome scenario, the emotion appreciative was tied for second place for the attribution “other’s effort” and tied for 6th for the attribution “other’s motivation and personality.” The emotion thankful was tied for first place for the attribution of “luck.” Next, researchers determined “discriminating affects,” or emotions that did not appear across all attributions. An emotion was considered discriminating if it’s mean for one attribution was higher than the mean across the other attributions. The emotions appreciative and grateful appeared as (p.337) discriminating affects for attributions to “other’s effort.” The emotions grateful and appreciative appeared for attributions to “other’s motivation and personality.” The emotion of thankful appeared for attributions to “luck.” Other discriminating emotions that were associated with attributions to others effort and other’s motivations were composed, relaxed, proud, modest, thoughtful, and charmed.

Conclusions: Gratitude is listed as a “dominant discriminating affect” for successes attributed to “other’s effort and personality” (p. 76). The authors conclude that different attributions for success lead to specific emotional responses.

Commentary: These results support the prediction that feelings of gratitude are related to external attributions of positive outcomes to help from other people or from abstract external agents such as luck.

Correspondence: No contact information is listed; however the authors were affiliated with the University of California at Los Angeles at the time of this research.

Weiner, B., Russell, D., & Lerman, D. (1979). The cognition-emotion process in achievement related contexts. Journal of Personality and Social Psychology, 37, 1211–1220.

Objective: To investigate the effect of attributions on emotions in an achievement context. Researchers were elaborating and improving upon their previous work in Weiner, Russel, & Lerman (1978).


Setting: Laboratory.

Participants: In Study 1, participants were 79 undergraduate students enrolled in general psychology. In Study 2, participants were 48 undergraduate students.

Assessment of predictor variables: In Study 1, participants were asked to remember 12 instances where they found out the outcome of a test. For six of these times, participants were asked to remember a time when they did well; for the other six instances, participants were asked to remember a time when they did poorly. Each scenario asked participants to think of a time when they made a different attribution for their failure or success: ability, unstable effort, stable effort, personality, others, and luck.

In Study 2, participants were presented with 12 scenarios that randomly varied by outcome (positive or negative) and six possible emotion descriptions (the descriptions relevant to gratitude included the emotion terms appreciative, grateful, and modest, and surprised, astonished, and thankful). A sample scenario described a student who just received a high score on an important example and felt surprised, astonished, and thankful.

Assessment of outcome variables: In Study 1, after writing about the details of an outcome, participants were asked to list three emotions that they felt in that situation, and then to rate the intensity of different emotions listed on the questionnaire.

(p.338) In Study 2, after each scenario, participants were asked to rate the attribution the person in the scenario would have made for his/her success or failure. Attributions rated for the success stories (which are the stories relevant to gratitude) included ability, unstable effort, stable effort, task ease, luck, and others.

Main results: In Study 1, gratitude and thankfulness were among the emotions that participants listed in the free-response portion of the questionnaire. Gratitude was most often mentioned for successful outcomes that were attributed to others, as was thankfulness. Both emotions were also mentioned a number of times when attributing success to luck. Gratitude arose as a discriminating emotion for attributions to others. Discriminating emotions were emotions that were mentioned significantly more for one attribution than for all others. Gratitude and thankfulness were not mentioned by participants in relation to failure outcomes.

In Study 2, the emotion cluster of appreciative, grateful, and modest had the highest rating for the attribution of success to others, and the emotion cluster of surprised, astonished, and thankful had the highest rating for the attribution of success to luck. The differences between the attribution ratings for these two clusters of emotion were significantly different from each other.

Conclusions: Attributions of success to an external agent lead to the emotions of gratitude and thankfulness. The authors found support for a three-step cognitive process of emotions, where the individual first evaluates performance based on success or failure (leading to outcome-dependent emotions such as happiness), then makes an attribution for that outcome (leading to attribution-dependent emotions such as gratitude), and finally the individual makes causal judgments relevant to the self-concept (leading to low or high self-esteem). The authors generalized past research using hypothetical achievement scenarios, to methodologies employing recollections of achievement, and explored the use of emotions as cues for attributions.

Commentary: These experiments provide further support for the hypothesis that gratitude and associated emotions such as appreciation and thankfulness stem from attributions of success to external agents, whether those external agents be other people, or abstract agents such as luck.

Correspondence: Bernard Weiner, Department of Psychology, University of California, Los Angeles, Los Angeles, California, 90095.

Zaleski, Z. (1988). Attributions and emotions related to future goal attainment. Journal of Educational Psychology, 80, 563–568.

Objective: To investigate attributions for the attainment of anticipated goals, and their concomitant emotional reactions.

Design: Study 1 was a between-subjects experiment, whereas Study 2 was a correlational study.

(p.339) Setting: Although not explicitly stated, it is assumed that Study 1 took place in a laboratory setting. Study 2 was conducted in students’ classrooms.

Participants: In Study 1, participants were 166 male and 165 female graduate and undergraduate students. In Study 2, participants were 392 undergraduate students from a university in the United States, and from a university in Canada.

Assessment of predictor variables: In Study 1, participants were asked to write down goals that they had set for one of five time-periods: 1 week, 1 month, 1 year, 10 years, or a life goal. In Study 2, participants noted their year in college (freshman, sophomore, etc.) on a questionnaire.

Assessment of outcome variables: In Study 1, after participants wrote about their goal, they filled out a questionnaire with the dependent measures, which included value of the goal, expectancies, effort and persistence, and attributions about attainment of the goal. Attributions that were relevant to the study of gratitude involved attributions of success to luck, and to “external conditions.”

In Study 2, dependent measures were collected on a questionnaire, and included ratings of internal and external attributions for current academic performance as well as future graduation. Participants were also asked to rate the emotional reactions they would have upon graduation, which included proud, surprised, and grateful. Participants answered similar questions regarding their attributions and emotions if they were to fail to graduate.

Main results: In Study 1, the most frequently cited attribution for success was effort and ability, followed by attributions to external circumstances. Participants were more likely to attribute successes to internal factors, and less to external factors. The longer the time range of the goal, the less participants attributed external factors to success, and the more they attributed external factors for failure. This effect was particularly driven by external attributions to task difficulty.

Like Study 1, in Study 2, successes were attributed more to internal factors than external factors. When asked about emotional reactions to successful graduation, participants rated that they would feel a high amount of pride, and intermediate amount of gratitude, and only a small amount of surprise. Emotions of surprise and gratitude were positively correlated with external attributions of success. There was no effect of year in school on these variables.

Conclusions: Attributions of anticipated outcomes coincided with research on attributions for past outcomes. The author found mixed results for the time variable between the two studies.

Commentary: These studies provide further research for gratitude’s link with attributions of success to others. Although participants probably were not experiencing the emotions in question during the studies, these studies have the advantage of examining participants’ own goals and anticipated outcomes, rather than presenting participants with hypothetical scenarios.

Correspondence: Zbigniew Zaleski, Department of Psychology, Catholic University of Lublin, Al. Ralawickie 14, Poland, 20–950.