Interpersonal perceptions and consequences of depressive-significant other relationships: a naturalistic study of college roommates

We investigated social perceptions and consequences of depression and anxiety in roommate relationships. Mildly depressed, anxious but nondepressed, and nondepressed-nonanxious students (targets) and normal, same-sex roommates (a) rated the interpersonal impact on themselves of typical associations with their roommates and (b) judged their own interpersonal impact. Only depressed men received negative evaluations and emotional reactions from their roommates. However, depressed women reported more negative reactions to their normal roommates than vice versa. Finally, depressed targets perceived their interpersonal impact negatively, whereas their normal roommates perceived their own interpersonal impact as overly positive. These findings suggest that negative relationships between depressives and nondepressed others may be attributable, at least in part, to both participants' misperceptions of their social behavior and its consequences.     

Depressed and Nondepressed Students: Differences in Interpersonal Perceptions

The purpose of this research was to study the patterns of interpersonal behavior of depressed students. Depressed participants rated themselves lower than non-depressed controls on assertiveness and initiation of interactions and significantly higher than controls on concern about what others think, introversion, and submissiveness. Depressed participants also scored significantly higher than controls on measures of dependency, self-criticism, and the need to please others. Ratings of depressed participants by their roommates were not correlated with depressed participants' self-reports. In contrast, most subscale self-ratings and roommate ratings for the nondepressed participants and their roommates were significantly correlated.     

Depressed and nondepressed students: differences in interpersonal perceptions

The purpose of this research was to study the patterns of interpersonal behavior of depressed students. Depressed participants rated themselves lower than non-depressed controls on assertiveness and initiation of interactions and significantly higher than controls on concern about what others think, introversion, and submissiveness. Depressed participants also scored significantly higher than controls on measures of dependency, self-criticism, and the need to please others. Ratings of depressed participants by their roommates were not correlated with depressed participants' self-reports. In contrast, most subscale self-ratings and roommate ratings for the nondepressed participants and their roommates were significantly correlated.     

When depression breeds contempt: reassurance seeking, self-esteem, and rejection of depressed college students by their roommates.

To test Coyne's (1976b) theory of depression, students' levels of depressive symptoms, reassurance seeking, and self-esteem were assessed at Time 1, and their same-gender roommates' appraisals of them were assessed 5 weeks later. Mildly depressed students engaged in the type of reassurance seeking described by Coyne. Among men, but not women, mildly depressed students were rejected if they strongly sought reassurance and had low self-esteem but not if they did not seek reassurance or had high self-esteem. Although induction of depressed symptoms in roommates did occur, this contagion effect did not account for the depression-rejection relationship. The prediction that unsupportive, intolerant, or unempathic others would be particularly likely to respond with rejection to reassurance-seeking depressed students with low self-esteem received partial support. Implications for future work on the interpersonal aspects of depression are discussed.     

Induction of depressive affect after prolonged exposure to a mildly depressed individual

Scores on the Beck Depression Inventory (BDI) were periodically obtained from the roommates of college students who exhibited a persistent mild depression over a 3-month period. For comparative purposes, BDI scores were also obtained from roommates of individuals who were transiently depressed and from subjects with nondepressed roommates. In comparison with control subjects, the roommates of persistently depressed persons displayed a progressive increase in BDI score over the course of the study.     

The impact of postpartum depressed mood on mother-infant interaction: Like mother like baby?

The purpose of this study was to test the hypothesis that mothers with depressed mood would exhibit less optimal interaction than their nondepressed counterparts and that their infants would show similar deficits in interactional behavior. Twenty-two mothers and their 2-month-old infants were videotaped in to-minute free-play segments in a laboratory playroom, and their interactions were coded using both time-sampling and global clinical ratings of behavior. Mothers with depressed mood were judged significantly lower on overall positive interaction, Expressivity/Affective Involvement, and Responsivity/ Sensitivity than were nondepressed mothers. Infants of mothers with depressed mood were rated significantly lower than infants of nondepressed mothers on corresponding interaction domains. Mothers with depressed mood were also rated as more variable than nondepressed mothers along a continuum of withdrawal to controlling/intrusive behavior. Contrary to prediction, level of maternal stimulation and infant activity did not differ as a function of depression in maternal mood. We conclude that mild to moderate symptoms of maternal depression may have salient but selective effects on mother-infant interaction.     

Infants of “depressed” mothers interact better with their nondepressed fathers

To determine whether infants of “depressed” mothers interact better with their nondepressed fathers, twenty-six 3- to 6-month-old infants were videotaped during face-to-face interactions with their parents. The “depressed” mother group consisted of twelve 3- to 6-month-old infants and their “depressed” mothers and nondepressed fathers. The control group was composed of 14 nondepressed mothers and nondepressed fathers and their 3- to 6-month-old infants. In the “depressed” mother group, the nondepressed fathers received better interaction ratings than the “depressed” mothers. In turn, the infants received better interaction ratings when they interacted with their nondepressed fathers than with their “depressed” mothers. In contrast, nondepressed fathers and mothers and their infants in the control group did not differ on any of their interaction ratings. These findings suggest that infants' difficult interaction behaviors noted during interactions with their “depressed” mothers may not extend to their nondepressed fathers. The data are discussed with respect to the notion that nondepressed fathers may “buffer” the effects of maternal depression on infant interaction behavior.     

Schema congruence and depression: loss of objectivity in self- and other-inferences

Depressed and nondepressed students judged the plausibility of positive and negative inferences ostensibly made either by themselves or by others. Negative self-inferences were judged by depressed students as more plausible, and positive other-inferences as less plausible. The results were in accord with Beck's (1967) theory of schema-based distortion in depression, which proposes that persons vulnerable to the development of depression are prone to make erroneous negative inferences and to then regard those inferences as plausible and correct. The results also suggested that depressed persons responded differentially depending on whether they were instructed to consider the inferences as their own or another's, whereas nondepressed persons did not.     

Social comparison and depression: company's effect on misery

In this article, two studies examine the social comparison processes of depressed and nondepressed college students. In the first study, subjects' preferences for information from others were assessed after they had received a manipulation intended to improve or worsen their mood states. The responses of the depressed subjects provided evidence of downward comparison, as they indicated a preference for information from people who were experiencing negative affect--but only when they themselves were also experiencing relatively negative affect, and not when their moods had been temporarily improved. In the second study, subjects' moods were assessed before and after they had received information indicating another person was currently experiencing very negative affect. This information had little effect on the nondepressed subjects, however, the mood states of the depressed persons improved after they read the information. In general, the results indicate that realizing that others are doing worse may help depressed persons to feel somewhat better.     

Depression, self-focused attention, and expectancies for positive and negative future life events for self and others

In two studies, we examined depressed and nondepressed persons' judgments of the probability of future positive and negative life events occurring to themselves and to others. Study 1 demonstrated that depressed subjects were generally less optimistic than their nondepressed counterparts: Although nondepressed subjects rated positive events as more likely to happen to themselves than negative events, depressed subjects did not. In addition, relative to nondepressed subjects, depressed subjects rated positive events as less likely to occur to themselves and more likely to occur to others and negative events as more likely to occur to both self and others. Study 2 investigated the role that differential levels of self-focused attention might play in mediating these differences. On the basis of prior findings that depressed persons generally engage in higher levels of self-focus than nondepressed persons do and the notion that self-focus activates one's self-schema, we hypothesized that inducing depressed subjects to focus externally would attenuate their pessimistic tendencies. Data from Study 2 supported the hypothesis that high levels of self-focus partially mediate depressive pessimism: Whereas self-focused depressed subjects were more pessimistic than nondepressed subjects, externally focused depressed subjects were not. The role of attentional focus in maintaining these and other depressive pessimistic tendencies was discussed.     

Depression,self-focus,and social interaction

The accuracy of depressed and nondepressed subjects' perceptions of their own and a social interactional partner's performance was investigated. Twenty depressed and twenty nondepressed college students participated in dyadic interactions and then rated their own and their partner's social behavior. The interactions were also rated by objective coders. Depressed subjects were differentiated from nondepressed subjects on several measures by both the coders and the subjects. Depressed subjects' self-ratings were correlated with the coders' ratings more often than were the nondepressives' ratings, suggesting depressives provided more accurate self-observations. Contrary to prediction, depressives were also more accurate in judging their partner's behavior. Depressives experienced heightened levels of self-focused attention, but this attentional focus did not mediate the relationship between depression level and self-accuracy. Finally, an analysis of the verbal statements suggests that performance differences between depressives and nondepressives may be a function of the quantity, rather than the quality, of the verbal production.     

Effect of depression level and length of acquaintance on reactions of others to a request for help

This experiment examined an interpersonal-process view of depression by assessing subjects' reactions to a request for help from a hypothetical depressed or nondepressed person with whom they had been acquainted for a relatively short (2 weeks) or long (1 year) period of time. Subjects responded to each of the four hypothetical persons by indicating their probable affective reactions to the request, the number of minutes they would be willing to help, their desire for future social contact with the hypothetical person, and their expectations of future requests for help. Requests from depressed persons elicited significantly more anger and social rejection but equal amounts of concern and willingness to help. This mixed response pattern was interpreted as providing partial support for an interpersonal-process view of depression. In addition a path analysis provided limited support for Coyne's (1976b) hypothesis that rejection of depressed persons results from the negative mood they induce in others.     

Examining the world of the depressed: do depressed people prefer others who are depressed?

Two studies were conducted to examine the interpersonal world of the depressed person. In Study 1, depression levels and perceptions of depressed and nondepressed people and their best friend were assessed to test the hypothesis that depressed Ss have best friends who are themselves more depressed than the best friends of nondepressed Ss. The hypothesis was confirmed, suggesting that depressed persons may prefer others who also tend toward depression. To examine this possibility, in Study 2 depressed and nondepressed college students spoke with one another in either depressed-depressed, nondepressed-depressed, or nondepressed-nondepressed pairs. It was found that depressed Ss felt worse than nondepressed Ss after speaking with nondepressed targets, but not after speaking with depressed targets. There were no differences in liking or in perceived similarity between the groups. Implications for the social world of the depressed person are discussed.     

Family interaction patterns and childhood depression

Rehm's (1977) self-control model for depression and Bandura's (1977) concept that children internalize external controls placed upon them are united to predict family interaction patterns that may contribute to the etiology or maintenance of depression in children. Families of depressed, nondepressed, and nonclinic children were compared on rates of, and criteria for, parental and self-reinforcement. Mother-father-child interactions were sequentially coded to reveal that mothers of depressed and nonclinic children both set very high criteria for rewarding their children, compared to mothers of clinic/nondepressed children. However, mothers of depressed children rewarded their depressed children at much lower rates than mothers of either clinic/nondepressed or nonclinic children. These factors, taken together, are discussed in terms of their possible etiological role in childhood depression.     

Interpersonal betrayal and cooperation: effects on self-evaluation in depression

Despite numerous studies demonstrating that depressed people are generally self-critical, little is known about interpersonal stressors that may activate or increase this negative self-evaluation. In this study, the effect of interpersonal betrayal and cooperative social interaction on self-evaluation processes in depressed and nondepressed women was assessed. Depressed subjects who experienced interpersonal betrayal were more critical of their performance on a subsequent task than were nondepressed subjects or depressed subjects who had experienced a cooperative interaction. Depressed subjects in the betrayal condition also behaved more aggressively toward their betraying partner than did nondepressed betrayed subjects. Depressed subjects were more critical of their own personality characteristics than were nondepressed subjects, regardless of condition. Results suggest that some negative cognitive schema among depressed persons may be altered by interpersonal factors, although it is not clear whether such effects are secondary to increases in self-criticism after conflict or to decreases in self-critical tendencies after positive interaction. Given the variability in results with different measures of self-evaluation, researchers are urged to use multiple, diverse measures of self-evaluation in future efforts to study variability in self-appraisal.     

Infants of depressed mothers show less “depressed” behavior with their nursery teachers

This study investigated whether infants' “depressed” behavior (i.e., less positive affect and lower activity levels) noted during their interactions with their depressed mothers generalized to their interactions with their nondepressed nursery teachers. Field et al. (1988) reported that infants of depressed mothers also show “depressed behavior” when interacting with nondepressed female adults, suggesting that the infants develop a generalized “depressed mood style” of interaction. However, in that study the adults were also strangers to the infants, confounding the results. In the present study, eighteen 3-month-old infants interacted with their depressed mothers and also with their nondepressed familiar teachers in 3-minute episodes. The infants' behavior ratings improved when they interacted with their familiar teachers compared to their interactions with their mothers. The infants' low activity levels and negative affect were specific to their interactions with their depressed mothers. Thus, the data suggest that the infants respond differentially to depressed and nondepressed adults who are familiar.     

Depressed mothers' and their infants' interactions with nondepressed partners

Twenty depressed adolescent mothers were videotaped interacting with their own infant and with the infant of a nondepressed mother. In addition, nondepressed mothers were videotaped with their own infant as well as with the infant of a depressed mother. Depressed mothers showed less facial expressivity than nondepressed mothers and received less optimal interaction rating scale scores (a summary score for state, physical activity, head orientation, gaze, silence during gaze aversion, facial expressions, vocalizations, infantized behavior, contingent responsivity, and gameplaying). This occurred independent of whether they were interacting with their own infant versus an infant of a nondepressed mother, suggesting that depressed mothers display less optimal behaviors to infants in general. The infants of both depressed and nondepressed mothers received better head orientation and summary ratings when they were interacting with another mother, perhaps because the other mother was more novel. Infants of nondepressed mothers, in particular, had better summary ratings (state, physical activity, head orientation, gaze, facial expressions, fussiness, and vocalizations) than the infants of depressed mothers when interacting with depressed mothers. Thus, it may be that infants of nondepressed mothers are generally better interaction partners than infants of depressed mothers. Another related possibility is that they persist longer in trying to elicit a response from mothers less responsive than their own, given that they have learned to expect a response to their behavior.     

Social context and perceived task difficulty as mediators of depressive self-evaluation

This study examined two separate, but potentially interactive, influences on depressive self-evaluation: social context and perceptions of task difficulty. First, it was hypothesized that, if negative self-evaluations of depressed individuals are motivated by a desire to elicit attention and sympathy from others, depressed subjects should evaluate themselves more negatively than nondepressed subjects in a public setting, but not when they make self-evaluative judgments in private. Second, it was hypothesized that negative self-evaluation results from a bias to perceive tasks as being intrinsically easy, i.e., if a task is easy, a given score would be evaluated more poorly than if the task were difficult. It was found that the self-evaluations of depressed subjects were influenced by the social context, but not always in a negative direction. Depressed subjects did not differ from nondepressed subjects when performance evaluations were made in private. In a public condition, depressed subjects evaluated themselves more negatively than nondepressed subjects following an easy task, but evaluated themselves more positively following a difficult task. Depressed subjects did not evidence a bias to perceive tasks as being intrinsically easy. Depressed subjects did rate the tasks to be more difficult for themselves than they thought they would be for others and this expectancy was predictive of negative self-evaluation. These results were discussed in terms of alternative self-presentation motives and theories of social cognition. Self-evaluation often involves social comparison and researchers need to attend to the potentially complex interactions among social and cognitive processes.I would like to thank Deborah Davis and Paul Westerholm for their help in data collection, Ruth Maki for her statistical expertise, and the anonymous reviewers for their thoughtful comments. Portions of this paper were presented at the Annual Convention of the American Psychological Association, New Orleans, 1989.     

Contingency judgments by depressed college students: sadder but not always wiser

We examined depressed and nondepressed college students' perceptions of control over outcomes in a task similar to the one introduced by Alloy and Abramson (1979, Experiment 2). In Experiment 1, when subjects completed a contingency learning task with no one else present, nondepressed subjects perceived themselves to have more control over frequently occurring response-independent outcomes than did depressed subjects, which replicated Alloy and Abramson's finding. When subjects completed the task in the presence of an observer, depressed students perceived themselves to have more control than did nondepressed students. In Experiment 2, the observer effects found in Experiment 1 were replicated, and we extended those results by showing that when response-independent outcomes occurred relatively infrequently, depressed and nondepressed subjects who completed the task in the presence of an observer did not reliably differ in their estimates of personal control. In Experiment 3, which included minor procedural variations from the other experiments, the pattern of results found in Experiments 1 and 2 was replicated under conditions in which observers were present while subjects received frequently occurring outcomes. Together, the results of the three experiments demonstrate that the consistently accurate personal control estimates of depressed subjects that have been found across a variety of situations break down when subjects complete a contingency learning task in the presence of an observer, and outcomes occur independently of response at a high frequency.