Interpersonal skill and depression in college students: An analysis of the timing of self-disclosures

Interpersonal skill and depression in college students: An analysis of the timing of self-disclosures

BEHAVIORTHERAPY13, 271-282 (1982) Interpersonal Skill and Depression in College Students: An Analysis of the Timing of Self-Disclosures NEIL S. JACOB...

692KB Sizes 0 Downloads 22 Views

BEHAVIORTHERAPY13, 271-282 (1982)

Interpersonal Skill and Depression in College Students: An Analysis of the Timing of Self-Disclosures NEIL S. JACOBSON

University of Washington ERLING A. ANDERSON

University of lowa The present study compared the timing of self-disclosures in depressed and nondepressed college students during conversations with confederates. It was predicted that depressed subjects would be more likely than nondepressed subjects to emit self-disclosures at times when such remarks were not directly solicited by confederate inquiries. To investigate the specificity of this tendency to depression, subjects expressing tendencies toward social avoidance and distress were also examined. The experimental situation consisted of a 10-rain waitingroom interaction where subjects were surreptitiously audiotaped while conversing with a confederate. Confederates were carefully trained according to a standardized script. Depressed and nondepressed subjects did not differ substantially in the frequency of various categories of verbal behavior, with the exception of negative self-statements, which were significantly more frequent for depressed subjects. Sequential analyses comparing depressed and nondepressed subjects in the timing of self-disclosures revealed that depressed subjects were more inclined to self-disclose following confederate remarks that did not directly solicit selfdisclosures, Thus, the major hypothesis was confirmed. A g r e a t deal o f a t t e n t i o n h a s b e e n p a i d in r e c e n t y e a r s to t h e identific a t i o n o f p s y c h o l o g i c a l or b e h a v i o r a l deficits a s s o c i a t e d w i t h d e p r e s s i o n ( B e c k , 1967; C o y n e , 1976a; L e w i n s o h n , 1974; S e l i g m a n , 1971). S o m e h a v e h y p o t h e s i z e d that d e p r e s s e d p e o p l e e x h i b i t i n t e r p e r s o n a l skill def i c i e n c i e s ( C o y n e , 1976a; L e w i n s o h n 1974; W e i s s m a n & P a y k e l , 1974). F o r Portions of this paper were presented at the annual Convention of the Association for Advancement of Behavior Therapy, November, 1980, New York City. Preparation of this manuscript was supported by NIMH Grant #5 R01 MH3383-02, awarded to the senior author. The authors wish to thank Jeffrey Leaman and Marilyn Hupfeld for acting as confederates, and Teresa Bena for serving as experimenter. Requests for the reprints should be sent to Neil S. Jacobson, Department of Psychology NI-25, University of Washington, Seattle, WA 98195. 271

0005-7894/82/027 i -028251.00/0

Copyright 1982 by Association for Advancement of Behavior Therapy All rights of reproduction in any form reserved.

272

JACOBSON AND ANDERSON

example, Lewinsohn and his associates (Lewinsohn 1974; Lewinsohn, Biglan, & Zeiss, 1976; Libet & Lewinsohn, 1973) have hypothesized that depressed people are deficient in their ability to emit responses that are socially reinforcing to others and fail to elicit sufficiently high rates of social reinforcers from others. Coyne (1976a) has suggested more specifically that depressed people elicit negative emotional responses in others, which prompts the latter to reject them. Research investigating the relationship between social skill and depression has produced contradictory findings. When compared to either normals or psychiatric controls, depressed people tend to be rated as less skilled by both observers and interacting partners (Coyne, 1976b; Lewinsohn, Mischel, Chaplin, & Barton, 1980; Youngren & Lewinsohn, 1980). Despite the ability to discriminate between depressed and nondepressed people on the basis of global ratings of social skill, investigators have been unsuccessful in identifying specific behavioral deficits unique to depression (Coyne, 1976b; Jacobson, 1981; Youngren & Lewinsohn). The discrepancy between data based on global skill ratings and that based on specific behavioral measures reflects, in our view, both conceptual and methodological limitations in previous research on the relationship between depression and social skill (cf. Jacobson). Behavioral measures of social skill tend to be derived unsystematically. Most typically, a number of behavioral indices of social skill are posited a priori rather than derived empirically. An alternate strategy for deriving behavioral measures of social skill would be to base such measures on a theoretical model which posits specific social skill deficiencies in depressed people. Although such theoretical models exist, they lack the degree of specificity from which to derive predictions about observable behavior. The resultant behavioral measures of social skill bear little logical connection to the theoretical models which ostensibly served as their foundation. Another limitation of previous research lies in the data reduction and statistical techniques used for data analysis. Previous studies have simply compared depressed and nondepressed subjects in the overall frequency with which certain behaviors are emitted. To arrive at these base rates or cumulative frequencies, behaviors are summed across an entire sequence of interaction, and the timing of particular behaviors is completely ignored. Not only is a great deal of information lost when analyses are confined to cumulative summation across time, but such analyses belie the complexity of social behavior in general and interpersonal skill in particular. Existing definitions of social skill implicate the timing of particular behaviors, and their immediate situational context, rather than simply the frequency with which particular behaviors are emitted (Curran & Monti, 1982; Fischetti, Curran, & Wessberg, 1977; Trower, 1980). The present study begins with the premise that depressed people tend to be self-preoccupied, and that this self-preoccupation is reflected in their social encounters. This is a common observation embedded in a variety of otherwise diverse theoretical formulations of depression (Co-

I N T E R P E R S O N A L SKILLS A N D DEPRESSION

273

hen, Baker, Cohen, Fromm-Reichmann, & Weigert, 1954; Coyne, 1976a; Hinchliffe, Hooper, & Roberts, 1978). By self-preoccupied, we refer to an interpersonal style that is fundamentally self-centered and self-oriented. We expected this self-centeredness to be reflected in a relative lack of interest and attention paid to others in interpersonal situations, as well as a tendency to refer to oneself even when the rules of social interaction dictate that self-references are inappropriate. Two years of extensive pilot work, looking at initial social encounters between opposite-sexed strangers, revealed that, when instructed to "get to know one another," normal college students maintain a relative balance between three types of remarks: self-disclosures, statements encouraging the interacting partner to talk about themselves (e.g., questions), and statements about the environment. Subjects who received high ratings of social skill tended to be those who were willing to talk about themselves when such self-disclosures were solicited from the partner but who avoided spontaneous or unsolicited references to self. The distinction between solicited and unsolicited self-disclosures has to do exclusively with the question of timing, and cannot be derived simply from the overall frequency of self-disclosures. Whereas the latter is a statement about the unconditional probability of self-disclosures, the former has to do with conditional probabilities. By comparing conditional (e.g., the probability of self-disclosure, given no direct solicitation) and unconditional (the overall base rate of self-disclosure) probabilities, one can directly answer experimental questions having to do with timing. The hypotheses in the present study were based on the assumption that depressed college students, given their self-preoccupation, would manifest a greater than normal tendency to deliver unsolicited self-disclosures. In order to establish whether or not this tendency was unique to depressives, rather ttian a nonspecific tendency characterizing anyone who found social encounters stressful, a control was included for generalized social avoidance and distress not specific to depression. Four groups of subjects were compared: depressed subjects who reported relatively high levels of social anxiety, depressed subjects who reported low levels of social anxiety, nondepressed subjects who reported high levels of social anxiety, and a normal control group consisting of subjects who were neither depressed nor reported high levels of social anxiety.

METHOD

Subjects Subjects were selected from a pool of introductory psychology students who were required to participate in a specified number of experiments in order to complete course requirements. Selection for depression was based on the Beck Depression Inventory (BDI; Beck, 1967). To be classified as depressed, subjects had to score at least 12 on two consecutive occasions, with a 30-day interval separating the two administrations. Sub-

274

JACOBSON A N D A N D E R S O N

jects whose score was 5 or lower on two repeated administrations were classified as nondepressed. The subject pool also completed the Social Anxiety and Distress Scale (SAD; Watson & Friend, 1969) and on that basis were classified into categories of high and low social anxiety. The final sample consisted of 64 subjects: 31 depressed subjects a n d 33 nondepressed subjects. More specifically, there were 16 depressedsocially anxious subjects (8 women, 8 men); 15 depressed-nonanxious subjects (11 women, 4 men); 17 nondepressed-socially anxious subjects (8 women, 9 men); and 16 normal control subjects (8 women, 8 men). Mean BDI scores for depressed and nondepressed subjects, respectively, were 18.31 (range 12-34) and 2.25 (range 0--4). Mean SAD scores for socially anxious and nonanxious subjects, respectively, were 21.2 (range 16-25) and 4.6 (range 0-9). Procedure The experiment lasted for approximately 30 min. Subjects were told that the experiment would consist of a problem-solving task to be performed in collaboration with a stranger. They were not informed that the experiment would be audiotaped, although they signed an informed consent statement which included the following passage: " . . . we frequently use video or audiotape to record subjects' behavior during various phases of the experiment . . . . Your signature indicates that you are aware that you may be taped during portions of the s t u d y . . . " Subjects were brought to a room where they completed the BDI for the second time. Then a confederate was brought into the room by an experimenter and introduced to the subject, although the former's confederate status was witheld. The experimenter then informed both subject and confederate that they would be partners in a subsequent problem solving task. But, prior to the task, they would each be given a questionnaire to complete. The questionnaire was presented to each of them and the experimenter left the room. The questionnaire was a list of artifacts which the subject was to rankorder in terms of their survival value on the moon. Subjects were under the impression that the problem-solving task would involve an effort to produce a consensus with the confederate on the correct rank-ordering. However, first they were to rank-order the items independently. Five min later the experimenter returned to the room, collected the questionnaires, and said, "I'll be back in a few minutes to give you your instructions." The experimenter then left the room. For the next 10 rain, the ensuing conversation between subject and confederate was taped. This conversation produced the raw data used to analyze the social interaction. After this 10-min period had elapsed, the experimenter returned to the room and presented the instructions for the problem-solving task and then left the room again so that subjects could complete it. Data from this task were not used in the present study, and thus will not be described further. After the task had been completed, the experimenter returned to the

INTERPERSONAL SKILLS AND DEPRESSION

275

room, and subjects were debriefed. The status o f the confederate was revealed and the subjects were informed that the preproblem solving conversation had been taped. T h e y were not told that the study involved an investigation of the relationship between social skill and depression. All subjects were then given the option of having their tapes erased. N o n e utilized this option.

Confederates One male and one female served as confederates in this experiment. Both were drama majors at the University o f Iowa. T h e y were selected because of their acting ability, high degree of physical attractiveness, and relative ease in interpersonal situations. The confederates were carefully trained to follow a standardized script in their conversations with subjects. The goal was for them to be responsive to subject initiations but at the same time to force the subject to initiate and maintain the conversation. Periodic checks on the behavior of confederates during their interactions with subjects indicated that they behaved in a c c o r d a n c e with experimental instructions. Moreover, there were no differences between the two confederates on any of the behavioral categories examined in the study, nor were there any differences in confederate behavior as a function of subject classification.

Response Measures Subject behavior was coded using an interaction coding system summarized in Table 1. Each speaking turn resulted in at least one verbal code (cf. Duncan & Fiske, 1977), and the behavioral categories were mutually exclusive and exhaustive with one exception: " t o p i c c h a n g e " was always double-coded with one additional category. The primary dimension of interest involved whether each remark was a self-disclosure, an inquiry designed to elicit information about the other (either " q u e s t i o n " or " s u m m a r y of o t h e r " ) , or a statement about the environment. Other codes reflected more specific content areas which were of interest and therefore studied in an exploratory manner. F o r example, "negative self-statements" were distinguished from other selfdisclosures because of the cognitive self-devaluation often attributed to depressed people (Beck, 1967). Three behavioral coders were carefully trained to an overall criterion of at least 70% agreement between each possible c o d e r pair. Subsequent to formal training, weekly calibration sessions were held to maintain accuracy. To assess inter-rater agreement, approximately ¼ of all interactions were coded by all raters. Raters were kept blind as to which interactions were coded by all. Criteria for agreement were very stringent in that coders were required not only to agree on the f r e q u e n c y with which particular behaviors were coded during specific interactions, but also on the sequence of codes throughout the entire interaction. Reliability statistics were calculated for subjects coded by all raters, based on Cohen's

276

JACOBSON AND ANDERSON TABLE 1 VERBAL CONTENT CODING SYSTEM (VCCS)

1. N S (negative self-statements): A n y r e m a r k relating to o n e s e l f that is clearly critical or self-minimizing. E x a m p l e s : ' T m not very good at m a t h . " " I was lucky to get a c c e p t e d . " ' T v e always had trouble with w o m e n . " This category takes p r e c e d e n c e over other categories that m a y overlap with it. T h u s , w h e n an utterance also fits other categories s u c h as SD, code N S instead o f SD. Example: " I u s e d to be a d r a m a major, b u t I w a s n ' t very good at acting so I switched to p s y c h o l o g y . " 2. QU (question): A n y question designed to elicit information from the o t h e r p e r s o n about themselves. Examples: " W h a t ' s your m a j o r ? " " D o you like I o w a C i t y ? " Questions w h i c h are not directed at discovering information a b o u t the other p e r s o n should not be coded QU. Rather, such r e m a r k s should be coded TA. Examples: "'What time is it?" "'How m u c h more time do we have to spend h e r e ? " 3. SO ( s u m m a r y of other): A c o m m e n t on the other p e r s o n ' s previous r e m a r k which is not a reference to the self. Questions which do not ask for further information on a r e m a r k j u s t m a d e by the other should also be coded SO. Conventional verbal a s s e n t s s u c h as " ' u h - h u h , " "'I s e e , " or " r e a l l y " are n o t coded SO. In fact, they are not coded at all. Make sure that the s t a t e m e n t is really a c o m m e n t on s o m e t h i n g the o t h e r p e r s o n said before coding SO. In other words, person A can not emit an SO immediately after an SO by p e r s o n B, unless another r e m a r k has intervened. 4. SD (Self-disclosure): A n y reference to the self, provided that the reference is not selfcritical (in which case you would code NS). Examples: ' T i n -I "I "I

a Psych. m a j o r . " like you."' live at 435 N. Main S t r e e t . " often feel like other people are out to get m e . "

5. SE ( s t a t e m e n t s about the environment): A n y s t a t e m e n t about the e n v i r o n m e n t , as long as there is no reference to either the self or the other. Opinions would be coded SE. 6. TT (Task-talk): A n y c o m m e n t related to the task that the subjects are engaged in. If a s t a t e m e n t fits both SO and TT, SO takes precedence.

7. TC (topic change): This is always double-coded with a n o t h e r category. It refers to an explicit abrupt a t t e m p t to change the subject, without reference to the previous remark. 8. TA (talk): A n y other verbal response.

(1960) Kappa statistic. A special procedure for calculating Kappa for several coders was used, a procedure described by Gottman (1979) and adapted from Fleiss (1971). Reliability figures were acceptable for all categories, with Kappa values ranging from .61 to .89. The mean Kappa value was .77.

INTERPERSONAL SKILLS AND DEPRESSION

277

RESULTS

Overall Frequencies The proportion of subject speaking turns that fell into each of the coding categories was calculated. There were only slight differences between depressed and nondepressed subjects in the overall frequencies with which particular behaviors were emitted; in fact, the similarities were considerably more striking than the differences. A multivariate analysis of variance, testing the overall significance of those differences, was conducted, using Wilks Limbda criterion. The result, F ( 7 , 4 2 ) = 2.191, p < .054, suggests an overall difference between depressed and nondepressed subjects of marginal significance? Univariate t tests revealed only one significant difference: Depressed subjects delivered a significantly higher rate of negative self-statements than nondepressed subjects, t(48) = 2.09, p < .05. Self-disclosures were 35.2% of the statements uttered by depressed subjects compared to 32.2% for nondepressed subjects. These differences were nonsignificant. In general, the evidence for differences between depressed and nondepressed subjects in the overall frequency of particular behavioral codes was meager.

Timing of Self-Disclosures It was predicted that depressed subjects would be more inclined to emit unsolicited self-disclosures than their nondepressed counterparts. A self-disclosure was defined as solicited whenever the confederate directed a "question" or a "summary of other" toward the S. Unsolicited selfdisclosures were those that followed either a self-disclosure or a "statement about the environment" emitted by the confederate. The likelihood of a S timing his/her self-disclosures immediately following a particular category of confederate remark is tested by a comparison of conditional and unconditional probabilities. For example, for a hypothetical subject (l), the overall probability of emitting a self-disclosure [P(SD0] can be contrasted with the probability of emitting a selfdisclosure immediately following a self-disclosure by the confederate (2) [P(SD~ISD2)]. To the extent that the conditional probability exceeds the overall or unconditional probability, one can argue that subject #1 is particularly likely to time his/her self-disclosures immediately following self-disclosures on the part of subject #2. This comparison can be described by calculating a Z score based on P ( S D 0 , P(SD2), and P(SDIlSDz) 2 (cf. Gottman, 1979). The greater the magnitude of the Z

1 This marginal main effect for depression was part of a four way (2 × 2 × 2 × 2) multivariate analysis of variance where the factors consisted of depression, social anxiety, sex of subject, and sex of confederate. Z = x - N P / ( N P Q ) ½ where x = observed joint frequency of SD~ and S D 2 , N P = predicted joint frequency, N P Q = variance of the difference between predicted and observed joint frequency.

278

JACOBSON AND ANDERSON TABLE 2 MEAN Z SCORES FOR SELF-DISClOSURES (SO) IN DEPRESSED AND NONDEPRESSED SUBJECTS

Depressed 1) 2) 3) 4) 5)

P(SDs[ SOc) 1 P(SDs [ SDc) P(SDs I QUc) P(SDs[ SEc) P(SDs ]TCe)

1.77 0.51 1.93 0.75 0.35

Nondepressed

Univariate F(1,48)

1.70 -0.27 1.50 0.19 0.19

ns

11.14"* ns

4.63* ns

1 Subscripts s and c refer to " s u b j e c t " and " c o n f e d e r a t e " respectively. * p < .05. ** p < .002.

score (assuming that the sign is positive), the greater the tendency on the part of subject # 1 to time his/her self-disclosures immediately following self-disclosures on the part of subject #2. 3 To compare depressed and nondepressed subjects on the timing of their self-disclosures, Z scores were computed for each subject, based on the conditional probability of SD1 given particular confederate codes. Since there were eight possible confederate codes, there were eight possible Z scores for each subject. The comparisons between depressed and nondepressed subjects were based on the mean Z scores for each group. A four-way (2 × 2 × 2 z 2) multivariate analysis of variance was conducted, with eight Z scores as criterion measures, and between-groups comparisons based on presence or absence of depression, presence or absence of social anxiety, sex of confederate, sex of subject. E x c e p t for a statistically significant interaction between sex of subject and sex of confederate, F(7,42) = 2.65, p < .03, 4 the only significant effect was a main effect for depression, F(7,42) = 2.97, p < .02. Table 2 lists mean Z scores for depressed and nondepressed subjects on the five conditional probabilities of primary interest. Inspection of the table reveals several commonalities among depressed and nondepressed subjects in the timing of their self-disclosures. Both groups of subjects tended to self-disclose immediately following confederate " q u e s t i o n s " 3 In the present study, Z scores were being used merely as descriptive statistics. U n d e r certain conditions (cf. Gottman, 1979), a Z score greater than 1.96 indicates a sequence effect, that is, the differences between conditional and unconditional probabilities are statistically significant. But we were not interested in the statistical significance o f these differences for any given subject; rather, our interest was in comparing the relative differences between conditional and unconditional probabilities for groups of subjects. 4 This interaction was examined and found to be unimportant. None o f the univariate tests proved to be statistically significant.

INTERPERSONAL SKILLS AND DEPRESSION

279

and "summaries of other." Similarly, neither group of subjects emitted a high frequency of unsolicited self-disclosures. However, there were also some significant differences between depressed and nondepressed subjects that are consistent with the hypotheses of the present study. First, despite the low frequency of unsolicited self-disclosures for both groups, depressed subjects were significantly more likely to self-disclose immediately following a confederate self-disclosure than were their nondepressed counterparts. These differences reflected the slightly greater than chance probabilities that depressed subjects would emit self-disclosures immediately following confederate selfdisclosures, compared to the slight disinclination on the part of nondepressed subjects to self-disclose at such times. Second, depressed subjects were significantly more likely to self-disclose following confederate statements about the environment than their nondepressed counterparts. Depressed subjects were more inclined to emit unsolicited self-disclosures than their nondepressed counterparts, and thus they could be distinguished from control groups on the basis of the timing of their selfdisclosures.

DISCUSSION Importantly, depressed subjects were found to be more inclined to refer to themselves when such self-references were not directly solicited than were their nondepressed counterparts. Thus, although depressed and nondepressed subjects did not differ in the proportion of their self-disclosures, they did differ in the timing of their self-disclosures. Since there were no significant differences between socially anxious and nonanxious subjects, and since there were no significant interactions between the depression and social anxiety factors, these effects were specific to depression, and not indicative of a general disposition in people tending toward social avoidance and distress. The predictions in this study were based on the hypothesis that depressed people are preoccupied with themselves, and that therefore they will focus discussion on themselves even at times when the normal flow of conversation suggests a more natural focus on the interacting partner. Other factors than heightened self-preoccupation could explain these results. For example, depressed subjects may be deficient in their knowledge of the rules of normal social interaction with strangers. Another possibility is that nondepressed people are more guarded during encounters with strangers, and therefore disinclined to self-disclose unless the partner directly solicits such disclosures. Moreover, the choice of operationalizing self-preoccupation in terms of unsolicited self-disclosures was made a priori, and one could argue that there are better or at least a/ternative ways of examining self-preoccupation behaviorally. We prefer to emphasize the empirical finding and the methodology for analyzing interpersonal aspects of depression, rather than the theoretical hypothesis.

280

JACOBSON AND ANDERSON

In interpreting these findings, several additional qualifications are in order, and these qualifications underscore the preliminary nature of the present investigation. First, the differences between depressed and nondepressed subjects were subtle, despite their having attained statistical significance. Neither group emitted an exceedingly high rate of unsolicited self-disclosures. Second, this study provides no direct evidence that these differences constitute a social skills deficit in depressed people. The findings merely demonstrate a difference between depressed and nondepressed subjects in their interactions with strangers. It remains to be demonstrated that self-disclosures occurring at times other than when they are directly solicited are, in fact, viewed as inappropriate by others. Third, the external validity of these findings is open to question, since the subjects were not seeking therapy for depression. The controversy regarding the value of a nonclinical population for research on depression will not be summarized here (cf. Depue & Monroe, 1978; Hammen, 1980). Obviously, it is important that these results be replicated with a clinical population of depressives. However, there is some evidence that college students scoring in the depressed range on the BDI are clinically depressed (Blumberry, Oliver, & McClure, 1978). Moreover, by requiring that depressed subjects score in the depressed range twice, with a month interval in between, we effectively eliminated those subjects suffering simply from transient mood disturbances. We predict that differences between poorly functioning depressives and others will be more pronounced than was the case with this sample of college students. It should be pointed out that the definition of self-disclosure used in the present study was content-free (see Table 1), and therefore it is possible that depressed and nondepressed subjects may differ in the quality of self-disclosures, despite the absence of differences between the two groups in overall frequency. For example, future investigations might examine self-disclosures on a "positive/negative" dimension to determine whether depressed self-disclosures tend to be more negative. In the present study, we did code one type of self-disclosure, negative selfstatements, as a separate category. Indeed, this was the one category where base rate differences were found between depressed and nondepressed subjects, with depressed subjects emitting a greater frequency of such remarks. Negative self-statements were almost nonexistent in the nondepressed sample, and infrequent even among depressed subjects. But the fact that such remarks did occur more frequently in depressed subjects provides empirical support for theories emphasizing the tendencies toward self-derogation in depressed people (Beck, 1967). A more comprehensive exploration of self-disclosure content may provide further insight into both the cognitive and interpersonal aspects of depression. The present findings justify the further use of sequential analyses in the investigation of interpersonal behavior and depression. Our findings underscore the importance of examining the timing of particular behaviors in lieu of the more common method of comparing overall frequencies

INTERPERSONAL SKILLS AND DEPRESSION

28]

summed across time. Indeed, had we confined our analyses to base rates, we would have found nothing. Future investigations can extend the inquiry by examining other behaviors, and by addressing the critical questions of generalization across subject populations and across interpersonal situations.

REFERENCES Beck, A.T. Depression: Clinical, experimental, and theoretical aspects. New York: Harper & Row, 1967. Blumberry, W., Oliver, J. M., & McClure, J . N . Validation of the Beck Depression Inventory in a university population using psychiatric estimate as the criterion. Journal of Consulting and Clinical Psychology, 1978, 46, 150-155. Cohen. J . A . A coefficient of agreement for nominal scales. Education and Psychological Measurement, 1960, 20, 37--46. Cohen, M. B., Baker, G., Cohen, R. A., Fromm-Reichmann, F., & Weigert, E . B . An intensive study of twelve cases of manic-depressive psychosis. Psychiatry, 1954, 17, 103-137. Coyne, J.C. Toward an interactional description of depression. Psychiatry, 1976, 39, 2840. (a) Coyne, J. C. Depression and the response of others. Journal of Abnormal Psychology, 1976, 85, 186--193. (b) Curran, J. P., & Monti, P . M . Social skills training: A practical handbook for assessment and treatment. New York: Guilford Press, 1982. Depue, R. A., & Monroe, S.M. Learned helplessness in the perspective of the depressive disorders: Conceptual and definitional issues. Journal of Abnormal Psychology, 1978, 87, 3-20. Duncan, S. I., & Fiske, D. W. Face to face interaction: Research, methods, and theory: New York: Wiley, 1977. Fischetti. M., Curran, J. P., & Wessberg, H . W . Sense of timing: A skill deficit in heterosexual-socially anxious males, Behavior Modification, 1977, 1, 17%194. Fleiss, J . L . Measuring nominal scale agreement among many raters. Psychological Bulletin, 1971, 76, 378-382, Gottman, J. M. Marital interaction: Experimental investigations. New York: Academic Press, 1979. Hammen, C . L . Depression in college students: Beyond the Beck Depression Inventory. Journal of Consulting and Clinical Psychology, 1980, 48, 126--128. Hinchliffe, M. K., Hooper, D., & Roberts, F . J . The melancholy marriage: Depression in marriage and psychosocial approaches to therapy. New York: Wiley, 1978. Jacobson, N . S . Assessment of overt behavior. In L. P. Rehm (Ed.), Behavior therapy for depression, New York: Academic Press, 1981. Lewinsohn, P . M . A behavioral approach to depression, In R. J. Friedman & M. M. Katz (Eds.), The psychology of depression: Contemporary theory and research. Washington, DC: W. H. Winston & Sons, 1974. Lewinsohn, P. M., Biglan, A., & Zeiss, A . M . Behavioral treatment of depression. In P. O. Davidson (Ed.), The behavioral management of anxiety, depression, and pain. New York: Brunner/Mazel, 1976. Lewinsohn, P. M., Mischel, W., Chaplin, W., & Barton, R. Social competence and depression: The role of illusory self-perceptions. Journal of Abnormal Psychology, 1980, 89, 194-202.

282

JACOBSON AND ANDERSON

Libet, J. M., & Lewinsohn, P . M . Concept of social skill with special reference to the behavior of depressed persons. Journal of Consulting and Clinical Psychology, 1973, 40, 304-312. Seligman, M. E . P . Helplessness: On depression, development, and death. San Francisco: W. H. Freeman, 1971. Trower, P. Situational analysis of the components and processes of behavior of socially skilled and unskilled patients. Journal of Consulting and Clinical Psychology, 1980, 48, 327-339. Watson, D., & Friend, R. Measurement of social-evaluative anxiety. Journal of Consulting and Clinical Psychology, 1969, 33, 448-457. Weissman, M. M., & Paykel, E . S . The depressed woman: A study of social relationships. Chicago: University of Chicago Press, 1974. Youngren, M. A., & Lewinsohn, P. M. The functional relation between depression and problematic interpersonal behavior. Journal of Abnormal Psychology, 1980.89, 331341. RECEIVED: July 7, 1981 FINAL ACCEPTANCE: January I l, 1982