Psicothema, 2002. Vol. Vol. 14 (nº 4). 783-787

Itziar Etxebarria, Xabier Isasi y Judith Pérez

This study analysed the nature of the events that cause guilt, as well as age and gender differences in this respect and in guilt intensity. The sample consisted of 202 subjects, aged from 15 to 19 and from 25 to 48. They reported their most recent experience of guilt and three things that usually made them feel guilty, and they rated the intensity of the feelings in each case on a 7-point scale. Most of the reported events were interpersonal, many of them involving close relationships. Both the proportion of interpersonal events and the guilt intensity were greater in females than in males in the adolescent group. In the adult group, males’ scores on these variables were closer to females’ scores, and gender differences disappeared. Results suggest that the interpersonal nature of events may be a fundamental variable for understanding not only the antecedents of guilt, but also individual differences in its intensity.

Naturaleza interpersonal de los eventos que provocan culpa. Diferencias de edad y género. Este estudio analizó el carácter de los eventos que provocan sentimientos de culpa, así como las diferencias de edad y género al respecto y en la intensidad de la culpa. Se pidió a 202 sujetos, adolescentes y adultos, que describieran su última experiencia de culpa y 3 cosas que habitualmente les hacían experimentar esta emoción, y que puntuaran la intensidad de la culpa en cada caso en una escala de 7 puntos. La mayor parte de los eventos citados fue de carácter interpersonal. La proporción de eventos interpersonales y la intensidad de la culpa fueron mayores en las mujeres que en los varones en el grupo de adolescentes. En el de adultos, las puntuaciones de los varones en ambas variables se acercaban a las de las mujeres. Los resultados sugieren que el carácter interpersonal de los eventos constituye una variable fundamental para la comprensión no sólo de los antecedentes de la culpa, sino también de las diferencias individuales en su intensidad.

Nowadays, empirical studies on guilt are quite numerous (for a review, see Baumeister, Stillwell, & Heatherton, 1994; Etxebarria, 2000; Tangney, 1995). However, studies directly aimed at analysing the types of events that elicit guilt are still scarce. The few studies referring to this aspect seem to support the interpersonal or relational approach to guilt (Baumeister et al., 1994; Jones, Kugler, & Adams, 1995), which posits that guilt basically appears in relation to actions or omissions that imply real or possible harm to others, especially those closest to us, and has a decisive role in controlling and repairing those actions. But most of these studies have analysed samples of North American adolescents or university students. It therefore seems expedient to analyse the antecedents of guilt in subjects of other ages and cultures, in order to see to what extent such conclusions can be generalised.

The present study is part of wider research aimed at analysing what types of events make people feel guilty in a Spanish sample with two age groups (adolescents and adults). Elsewhere we have presented results on the great variety of categories of events that can cause people to experience guilt both in general and in the different age and gender groups (Etxebarria & Pérez, 2002). The main aim of the study presented here was to analyse the interpersonal or relational nature of those events.

In relation to this issue, one interesting aspect that has hardly been studied is the relationship between the interpersonal nature of the antecedent events and the intensity of the ensuing guilt. Thus, as a second aim, we decided to analyse this. As a third objective, this study aimed to analyse age and gender differences in the nature of these events. In addition, taking into account research on age and gender differences in guilt intensity and the relevant implications of the supposed female tendency to experience more intense guilt feelings than males, the study aimed to analyse age and gender differences in guilt intensity. Let us look briefly at the research on these four issues.

Baumeister, Reis, and Delespaul (1995) asked a group of 74 North American university students to report their most recent experiences of 6 different negative emotions. They found that in reports of guilt, other people tended to be involved to a greater degree than in other emotional experiences. Moreover, more than half (56%) of the guilt reports involved partners with whom the reporters had a close relationship, whereas only 30% of the reports about other emotions did. As for the categories of guilt-producing events, the most common one was «neglecting someone», usually a relationship partner. The authors concluded that the sources of guilt tend to be mostly of an interpersonal nature, especially involving the people closest to us. Baumeister et al. draw attention to the fact that in their study, traditional moral issues such as lying or cheating were not the main sources of guilt. They suggest that from their results guilt emerges «less as a matter of making one’s actions conform to societal laws and moral principles than as a matter of making one’s actions conform to goals and standards that arise from one’s particular personal and interpersonal concerns» (p. 1266).

In another study, Baumeister, Stillwell, and Heatherton (1995) asked 47 psychology students to describe an incident in which they had made someone feel anger and had felt guilty afterwards, and a similar incident in which they had not felt guilty. Analysis showed that, in most of the cases in which the subject experienced guilt, the victim was someone with whom the subject had a communal relationship (defined as involving norms of mutual concern for each other’s welfare, such as in family or love relationships). The same authors, in a sample of subjects aged from 19 to 65, asked half of the subjects to report an incident in which they had induced guilt feelings in someone else, and asked the other half to report an incident in which someone had made them feel guilty. Results from the study showed that guilt inductions were linked to close relationships; these inductions happen between people who have quite close relationships with each other. With regard to the types of events that prompted subjects to make someone else feel guilty, by far the largest category involved interpersonal neglect.

Likewise, Jones et al. (1995), in a study involving college students in which they distinguished between relational transgressions (transgressions that involved inflicting harm on an important person for the subject, such as lying to a friend, betraying his or her confidence or hating a relative) and nonrelational ones (transgressions that basically involved the violation of a norm, without involving any harm to people, such as masturbation, smoking marijuana or looking at nude pictures), found that both state guilt and trait guilt significantly correlated with the frequency of the relational transgressions, but not with that of the nonrelational ones.

Taking these studies into account, we hypothesised that in our sample too, most of the guilt-inducing events would be interpersonal, and would involve close relationships. Moreover, on the basis of the results of the study by Jones et al. (1995), and the theoretical analyses by both Hoffman (1998) and Baumeister et al. (1994), we hypothesised that interpersonal events would probably elicit more intense guilt than non-interpersonal ones.

Let us now consider the research on age and gender differences in the interpersonal nature of guilt-producing events. With respect to gender differences, although studies specifically aiming to analyse this issue are scarce, there are nevertheless a number of studies which support the presence of more interpersonal concerns in females than in males (e.g., Eagly, 1987; Gilligan, 1982; Hoffman, 1975). This greater presence of interpersonal concerns in women suggests that the problems that can emerge in personal relationships are a more common source of guilt in women than in men. Thus we hypothesised a greater proportion of antecedent events of an interpersonal nature in women. As for age differences in the nature of guilt-producing events, there is a lack of research in this area. However, we supposed that in adulthood, due to the increase of interpersonal concerns related to partner, children, etc., there would be a greater proportion of antecedent events of an interpersonal nature than in adolescence.

Finally, let us consider age and gender differences in guilt intensity. Regarding gender differences, there is a large body of empirical evidence supporting the idea that females tend to experience more intense guilt than males (Bybee, 1998; Etxebarria, 1992, 1994; Harvey, Gore, Frank, & Batres, 1997; Hoffman, 1975; Lutwak, & Ferrari, 1996; Lutwak, Ferrari, & Cheek, 1998; Tangney, 1990). Most of the studies have been carried out with adolescent or university student samples; studies with adult samples are scarce and have not provided consistent results. However, in accordance with the above-mentioned hypothesis of a greater proportion of antecedent events of an interpersonal nature in women of both age groups, we hypothesised that females would show more intense guilt than males in both age groups. With regard to age differences in guilt intensity, some studies suggest that guilt feelings tend to increase during childhood and then decline during adolescence (Bybee, Merisca, & Zigler, 1995), especially among males (Williams & Bybee, 1994). On the other hand, various other studies suggest a later increase in guilt feelings. For example, Bybee and Williams (1997, cited in Bybee, 1998) found an increase in the intensity of guilt for urban females from the 8th to the 11th grades, and Bybee, Zielonka, and Mayne (1997) found a positive correlation between age and guilt among men in their twenties and thirties. Taking these data into account, and the above cited hypothesis about a greater proportion of antecedent events of an interpersonal nature in adulthood, we hypothesised that adults would show more intense guilt than adolescents.

Method

202 adolescents and adults participated in the study. The adolescent group consisted of 129 individuals, ranging in age from 15 to 19 years (64 males and 65 females, M= 16.8, SD= 1.27), all of whom were students from various public and private schools. The adults were contacted individually through university students and in small groups through different cultural and leisure associations. This, along with the reluctance shown towards the test by some of them, especially males, made it quite difficult to get an adult group which was numerous enough and with a balance of sexes. In the end, the adult group consisted of 73 individuals, ranging in age from 25 to 48 years (34 males and 39 females, M= 35.1, SD= 7.3).

A questionnaire was used in which: first, participants were asked to think about the last time they experienced guilt feelings, to describe the situation to the best of their ability, and to rate the intensity of those feelings on a 7-point scale (1= no guilt, 7= very intense guilt); second, they were asked to describe what kinds of events usually made them experience guilt feelings and to mention three of them, rating the intensity of those feelings in each case on a similar 7-point scale.

Two collaborators applied the questionnaire to the students collectively in the classroom. It took about half an hour. The adults answered it individually at home.

We coded the reported events as either interpersonal or non-interpersonal depending on whether or not there was explicit mention of the act affecting someone else. For example, an event such as «betraying my best friend’s confidence» was coded as interpersonal; an event such as «drinking too much» as non-interpersonal. Then all interpersonal events were given a score of 1 to 5 according to the degree of intimacy of the relationship between the subject and the main other person involved in the account: 5 for a close relationship partner, including family member, romantic partner, or close friend; 4 for an acquaintance; 3 for a more impersonal, role-based relationship (e.g., professor, client) or impersonal interaction partner such as an audience for a speech or performance; 2 for a stranger with whom further interactions were expected; 1 for a stranger with whom no interactions were expected. In this scoring system, the first example cited above would be given a score of 5.

In order to analyse the reliability of the latter two variables, two students from upper-level psychology courses independently coded the answers of 60 subjects. The agreement between them in coding the events as interpersonal or non-interpersonal was very high, but somewhat lower in the case of the last experience of guilt: kappa indices were .73 in this case and .99, .90, and .96 in that of the accounts of usual guilt mentioned first, second and third, respectively. The agreement level between the two judges in coding the intimacy degree was r= .76 for the case of last-experienced guilt and r= .99, .95, and .77 for that of the accounts of usual guilt mentioned first, second and third, respectively.

Results

The hypothesis that most of the guilt-inducing events would be of an interpersonal nature was supported. 86.5% of the reports of the last experience of guilt made reference to interpersonal events; only 13.5% of the events were non-interpersonal, t (184)= 14.53, p<.001. We found a similar pattern for the three events that usually made our participants feel guilty. In this case, the proportion of interpersonal events was somewhat smaller, but continued to be significantly higher than that of non-interpersonal events: 70.6% of the events cited first, t (179)= 6.07, p<.001; 63.3% of those cited second, t (179)= 3.7, p<.001; and 60.1% of those cited third, t (162)= 2.63, p<.05.

It should be pointed out here that in the present study, answers such as «lying» and «stealing», although they could be regarded as interpersonal events –given that such behaviours have implications for others– were not coded as such, unless there was some clear reference to the act affecting another person. This kind of short answer was more frequent in the second part of the questionnaire, when answering the question about usual guilt, and this is why percentages of interpersonal events were lower in this case.

Not only were most of the events interpersonal, but as had also been hypothesised, a large proportion of them involved close relationships. Indeed, 52.7% of the total number of reports of the last experience of guilt, that is, 61.8% of the reports of an interpersonal nature, involved people with whom subjects had a close relationship (score 5, e.g., a member of their family, their romantic partner, a close friend), t (143)= 2.91, p<.05. In their report of the three events that usually made them feel guilty, the percentage of interpersonal events that involved a close relationship was: 83.1% in the first case, t (70)= 7.44, p<.001; 80.6% in the second, t (61)= 6.09, p<.001; and 69.5% in the third, t (58)= 3.25, p<.001.

In order to compare guilt intensity elicited by interpersonal events and non-interpersonal ones in each of the four guilt accounts, t tests were performed. These tests showed a significant difference only in the case of the usual guilt mentioned in first place, with interpersonal events eliciting more intense guilt than non-interpersonal ones, t (166)= 2.44, p<.05. Thus, the hypothesis that interpersonal events would elicit more intense guilt than non-interpersonal ones found only weak support. Descriptive statistics are presented in Table 1.

A χ² test, carried out in order to analyse to what extent there were gender and age differences in the interpersonal nature of the antecedent events of the last experience of guilt, showed that, in the adolescent group, the females’ percentage of interpersonal events (91.7%) was significantly higher than the males’ (75.4%), χ² (1, 202)= 5.66, p<.05. In the adult group, males drew closer to women in their percentages (90% and 92.1%, respectively). A second χ² test, carried out to analyse to what extent there were gender and age differences in the interpersonal nature of the antecedent events of usual guilt, did not show significant differences.

As can be seen, the hypothesis regarding a greater proportion of interpersonal events for females versus males found only partial support: in the case of the last experience of guilt, the difference was significant in the adolescent group, but not in the adult group. With respect to the hypothesis of a greater proportion of interpersonal events in the adult age group, the data showed some tendency in this sense in males, but the difference between adult and adolescent males was not significant.

To analyse whether females presented more intense guilt than males both in adolescence and adulthood, as well as possible differences in guilt intensity between the two age groups, a 2 x 2 (Age x Gender) ANOVA was performed. The cell means are listed in Table 2.

In the case of the last experience of guilt, this analysis showed a significant interaction, F(1, 167)= 4.76, p<.05. Adolescent males showed very low scores in guilt intensity. Adolescent females showed significantly higher scores than them, t (104)= 3.57, p<.001; adult females scored marginally higher than them too, t (87)= 1.96, p<.06. Adult males showed scores close to those of females, marginally higher than those of adolescent males, t (78)= 1.79, p<.08. Thus, in the adult group, gender differences practically disappeared. Similarly, in the case of the usual experiences of guilt, the average intensity of guilt experiences was higher in females than in males in the adolescent group, t (110)= 2.27, p < .05 and adult females scores were higher than those of adolescent males, t (87)= 2.49, p<.05. Also, adult males presented higher scores than adolescent males, t (82)= 1.85, p<.07.

As can be seen, the hypothesis regarding more intense guilt feelings in females found support in the adolescent group, but not in the adult group. The hypothesis regarding more intense guilt feelings in adults found support in males, but not in females.

Discussion

The main aim of the present study was to analyse the nature of the events that elicit guilt in a Spanish sample of adults and adolescents, in order to see to what extent previous conclusions about the interpersonal sources of guilt could be generalised. As we have pointed out in the introduction, this study is part of wider research about the types of events that make people feel guilty. Analyses reported elsewhere (Etxebarria & Pérez, 2002) showed that the specific categories of events mentioned by Spanish participants, and their corresponding frequencies, were remarkably similar to those found by Baumeister, Reis, and Delespaul (1995) in a sample of North American university students. Although such results were of great interest, what is even more so from a theoretical point of view are results from the present study which show that most of the events mentioned by the participants as antecedents of their experiences of guilt were of an interpersonal nature, many of them involving close relationships. These results provide new support to Baumeister´s interpersonal view of guilt as an emotion that is basically elicited by actions or omissions that have negative implications in interpersonal relationships.

However, it is important to point out that this should not lead us to conclude that guilt is a simple «social» emotion that has nothing to do with morality. If we take Turiel´s distinction between conventional and moral events (Turiel, 1983) and his definition of the latter as those events that involve harming other people (e.g., physical harm, deceit, violation of rights), then the intrinsically moral nature of such experiences of guilt is clear.

Two other basic aims of the present study were to analyse age and gender differences in the nature of the guilt-producing events and in guilt intensity. Results regarding these two issues show a strong parallelism. Let us first discuss differences in guilt intensity.

Analyses of age and gender differences in guilt intensity revealed that, in adolescence, females showed more intense guilt than males, who presented very low levels of intensity. However, in adulthood males tended to present intensity scores on a par with those of females. These results deserve special attention. Firstly, they mean that perhaps females’ tendency to experience more intense guilt feelings, although clear in adolescence, cannot be generalised to other age groups. Secondly, these results mean that the low intensity of guilt in adolescence and its tendency to increase later on may only be generalised to males. Thirdly, they raise a twofold question: what are the reasons on the one hand for the differences between females and males in adolescence, and on the other hand for the increase in the intensity of guilt between adolescence and adulthood in males?

As for the first question, various studies suggest that gender differences in guilt intensity respond largely to the different use of disciplinary practices with daughters and sons: with female children and adolescents, parents make greater use of induction, the type of discipline associated with a greater intensity of empathy and guilt (Etxebarria, 1992, 1994; Harvey et al., 1997; Zahn-Waxler & Robinson, 1995). Together with this factor, the results of the present study suggest an additional one that would also provide at least a partial answer to the second question: the different proportion of antecedent events of an interpersonal nature in the different subgroups. The greater presence of interpersonal events among girls than among boys in the adolescent group and the increase in these types of concerns between adolescence and adulthood in males would explain the corresponding differences in guilt intensity.

In conclusion, our data suggest that age and gender differences observed in guilt intensity may be mediated by the different presence of interpersonal concerns in the different groups analysed. It would be interesting to analyse this issue more directly in the future.

Not only is it necessary to test this hypothesis more directly, but also, before taking for granted the differences found in the present study, to analyse whether these differences are replicated in similar samples. The intensity of guilt found in adult males –significantly higher than that of adolescent males and almost as high as women– can of course be explained by the high presence of interpersonal concerns in this group. However, some of the data suggest that the adult male group in this study may have been slightly biased. As indicated in the methodological section above, some of the adult males clearly showed a reluctance towards the questionnaire and did not answer, in some cases claiming that they «never» felt guilty. Whether or not this is true, it may be that the real interpersonal concerns and guilt intensity in adult males are somewhat less than those found in our study, in which case perhaps one could not speak of an increase in guilt from adolescence to adulthood in males, but rather of a lower guilt intensity in males relative to women also in adulthood, as posited by the original hypothesis. In subsequent studies this issue will have to be taken into account, perhaps resorting to tests of a more clinical nature such as in-depth interviews instead of questionnaires, so as to avoid the loss of subjects from this group and be better able to explore its characteristic experiences of guilt.

What seems clear is that the original hypothesis, with regard to greater guilt intensity in women (both adolescent and adult), needs further study before being rejected. A closer look at this issue would be of great interest, because the outcome might substantially alter our view on differences in guilt: we might maintain the hitherto prevailing view that women are especially prone to such feelings, or on the other hand conclude that it is not so much a question of gender difference as of difference between the adolescent male group –with very weak guilt feelings– and all the rest, including adult males.

As things currently stand, research into this area has not yielded results that clearly support either view. Some reviews conclude that from adolescence onward females tend to experience more intense guilt than males (Bybee, 1998), and on this basis we expected more intense guilt in females of both the age groups analysed. However, most of the comparative studies beyond adolescence have been carried out with samples of young people (most of them university students), and the few studies with adult samples do not provide consistent results: some of them have found that women score higher than men (Hoffman, 1975; Tangney, 1990), but others have not found similar differences (Kugler, & Jones, 1992; Müller, 1991).

Nevertheless, on the whole the results of the present study suggest that the interpersonal nature of events may be a fundamental variable for understanding not only the antecedents of guilt, but also individual differences in its intensity.

Baumeister, R.F., Reis, H.T. and Delespaul, P. (1995). Subjective and experiential correlates of guilt in daily life. Personality and Social Psychology Bulletin, 21(12), 1.256-1.268.

Baumeister, R.F., Stillwell, A.M. and Heatherton, T.F. (1994). Guilt: An interpersonal approach. Psychological Bulletin, 115(2), 243-267.

Baumeister, R.F., Stillwell, A.M. and Heatherton, T.F. (1995). Interpersonal aspects of guilt: Evidence from narrative studies. In J.P. Tangney and K. W. Fischer (Eds.), Self-conscious emotions: The psychology of shame, guilt, embarrassment, and pride (pp. 255-273). New York: Guilford Press.

Bybee, J. (1998). The emergence of gender differences in guilt during adolescence. In J. Bybee (Ed.), Guilt and children (pp. 113-125). San Diego, CA.

Bybee, J.A., Merisca, R. and Zigler, E. (1995, August). Developmental, situational, and gender differences in the intensity of guilt reactions. Poster presented at the annual meeting of the American Psychological Society, New York, NY.

Bybee, J. A., Zielonka, E. and Mayne, T. (1997). Guilt and shame among heterosexual and homosexual men: Relationships to sexual behavior and mental health. Unpublished manuscript, Northeastern University, Boston, MA.

Eagly, A.H. (1987). Sex differences in social behavior: A social-role interpretation. Hillsdale, NJ: Erlbaum.

Etxebarria, I. (1992). Sentimientos de culpa y problemática del cambio de valores en la mujer. Revista de Psicología General y Aplicada, 45, 91-101.

Etxebarria, I. (1994). «Non-rational guilt»: Predictors of its appearance in processes of change in moral values. Journal of Moral Education, 23(2), 145-164.

Etxebarria, I. (2000). Guilt: An emotion under suspicion. Psicothema, 12 (Suppl. 1), 101-108.

Etxebarria, I. and Pérez, J. (2002). ¿Qué cosas nos hacen sentir culpa? Categorías de eventos en adolescentes y adultos de uno y otro sexo. Manuscript submitted for publication.

Gilligan, C. (1982). In a different voice: Psychological theory and women’s development. Cambridge: Harvard University Press.

Harvey, O.J., Gore, E.J., Frank, H. and Batres, A.R. (1997). Relationship of shame and guilt to gender and parenting practices. Personality and Individual Differences, 23(1), 135-146.

Hoffman, M.L. (1975). Sex differences in moral internalization and values. Journal of Personality and Social Psychology, 32, 720-729.

Hoffman, M.L. (1998). Varieties of empathy-based guilt. In J. Bybee (Ed.), Guilt and children (pp. 91-112). San Diego, CA: Academic Press.

Jones, W.H., Kugler, K. and Adams, P. (1995). You always hurt the one you love: Guilt and transgressions against relationship partners. In J.P. Tangney and K.W. Fischer (Eds.), Self-conscious emotions: The psychology of shame, guilt, embarrassment, and pride (pp. 301-321). New York: The Guilford Press.

Kugler, K. and Jones, W.H. (1992). On conceptualizing and assessing guilt. Journal of Personality and Social Psychology, 62(2), 318-327.

Lutwak, N. and Ferrari, J.R. (1996). Moral affect and cognitive processes: Differentiating shame from guilt among men and women. Personality and Individual Differences, 21(6), 891-896.

Lutwak, N., Ferrari, J.R. and Cheek, J.M. (1998). Shame, guilt, and identity in men and women: The role of identity orientation and processing style in moral affects. Personality and Individual Differences, 25 (6), 1.027-1.036.

Müller, M.M. (1991). The stability of anger across age and sex in German cohorts born between 1930 and 1972. Personality and Individual Differences, 12(5), 417-425.

Tangney, J.P. (1990). Assessing individual differences in proneness to shame and guilt: Development of the Self-Conscious Affect and Attribution Inventory. Journal of Personality and Social Psychology, 59, 102-111.

Tangney, J.P. (1995). Recent advances in the empirical study of shame and guilt. American Behavioral Scientist, 38(8), 1.132-1.145.

Turiel, E. (1983). The development of social knowledge: Morality and convention. Cambridge: Cambridge University Press.

Williams, C. and Bybee, J. (1994). Do grade level, peer delinquency, emotionality, and gender affect ratings of own and other’s guilt? Paper presented at the annual meeting of the Eastern Psychological Association, Providence, RI.

Zahn-Waxler, C. and Robinson, J. (1995). Empathy and guilt: Early origins of feelings of responsibility. In J.P. Tangney and K.W. Fischer (Eds.), Self-conscious emotions: The psychology of shame, guilt, embarrassment, and pride (pp. 143-173). New York: Guilford Press.