The development of a measure of parental stress, the Parental Stress Scale, is presented. All participants (total N = 1276) completed the Parental Stress Scale and some also completed the Parenting Stress Index, a generic measure of stress called the Perceived Stress Scale, relevant measures of emotions and role satisfaction (e.g. loneliness, marital satisfaction, guilt, etc.), and one group completed the Parental Stress Scale twice. Analyses suggested that the Parental Stress Scale is highly reliable, both internally and over time, and related to the general measure of stress. Also, results were consistent across parents of differing parental characteristics, suggesting the stability of scale characteristics. The validity of Parental Stress Scale scores was supported by predicted correlations with measures of relevant emotions and role satisfaction and significant discrimination between mothers of children in treatment for emotional/behavioral problems and developmental disabilities vs mothers of children not receiving treatment. Finally, a factor analysis suggested that a 4-factor structure underlies responses to the Parental Stress Scale, despite its high internal reliability.

The psychological construct of guilt and its measurement were investigated in a series of studies (N = 1,041). Data collection and analyses involved 3 broad issues: (a) the further development and validation of the Guilt Inventory, designed to measure trait guilt, state guilt, and moral standards; (b) comparisons involving extant measures of guilt, shame, moral standards, and other emotions for the purpose of exploring the distinctions between trait guilt and state guilt, affective guilt and moral standards, and guilt and other emotions including both shame and the full range of emotional traits; and (c) an exploration of the factor structure of the trait guilt scales. Results supported the validity and reliability of the Guilt Inventory and the validity of most of the other guilt measures. However, results also raised questions regarding commonly articulated distinctions between trait and state guilt, guilt and shame, and, for some measures, affective guilt and moral standards. Discussion focuses on issues of measurement and conceptualization.

for their helpful comments on drafts of this article. We also wish to acknowledge the contributions of Connie J. Dyke and Dorothy F. Marsil, whose technical expertise facilitated the completion of this research.

The relationship between social skill deficits and the psychological state of loneliness was examined in two studies. Study 1 compared conversational behaviors of high-lonely and low-lonely college students during brief heterosexual interactions. Results indicated that the two loneliness groups differed significantly in their use of a specific class of conversational behaviors termed partner attention, with high-lonely as compared to low-lonely subjects giving less attention to their partners. Study 2 examined the casual relationship between social skill and loneliness by directly manipulating the use of partner attention in a group of high-lonely males. For that group, increased use of partner attention during dyadic interactions resulted in significantly greater change in loneliness and related variables relative to interaction only and no-contact control groups. The utility of conceptualizing loneliness as a social skills problem is discussed.

Four studies (total n = 469) examined correlates of loneliness in order to explore explanations for the persistence of loneliness among college students. Self-report and attitude scales, ratings of others following dyadic interactions, and self and other ratings at two points during an extended period of group interactions indicated that lonely students (a) rated themselves more negatively and reported deficits in social skills and self-concept, (b) rated specific others and people-in-general more negatively and were more alienated and externalized, (c) expected others to rate them negatively, but (d) in general were not differentially rated by others except in the initial phase of group interactions and by lonely others following dyadic interactions. Results suggested that loneliness may be perpetuated by its cognitive and affective concomitants, with some evidence for gender differences, whereas inconclusive evidence was found regarding responses of others to the lonely person.Loneliness is a distressing and common problem in which a person's network of interpersonal relationships is smaller or less satisfying than desired (Peplau & Perlman, 1979). One national survey indicated that 26% of the sample reported feeling lonely, suggesting the pervasiveness of the problem (Bradburn, 1969). The determinants of loneliness have not been identified. However, they are usually considered (e.g., Weiss, 1973) to be a combination of personal vulnerabilities (e.g,, ineffective social skills) and circumstances (e.g., divorce, social mobility).Despite its importance, until recently, systematic research in this area was relatively uncommon. In particular, one question which 1, Preliminary reports of this paper were presented at the