Further investigation of the relationship between expressed social fear and assertiveness

Recently, Morgan (1974) investigated the relationship between expressed social fear and assertiveness. As hypothesized, Morgan found a statistically significant relationship between social fear as measured by ten items from the Fear Survey Schedule-III (FSS-III) (Wolpe and Lang, 1964) and assertiveness as measured by the Rathus Assertiveness Schedule (RAS) (Rathus, 1973b). The observed relationship, however, was of little predictive importance, as noted by Morgan, in that the Pearson product-moment correlation coefficients accounted at most for less than 6 per cent of the variance. This finding raises questions concerning traditional views of assertiveness (e.g. Salter, 1949; Wolpe 1958; Wolpe and Lazarus, 1966) which have hypothesized that an individual is “inhibited from the performance of ‘normal’ behavior because of neurotic fear” (Wolpe, 1973, p. 81).The literature concerning the relationship between assertive behavior and social fear is meager. Goldstein, Serber and Piaget (1970) have reported two illustrative case studies in which anger expressiveness induced through assertiveness training resulted in the clinically assessed reduction of social fear. Rathus (1973a) has reported the results of an empirical investigation of assertiveness training which might, as suggested by Morgan, lend indirect support for the hypothesized relationship. Using undergraduate female subjects, Rathus found that an assertive training group not only scored higher on a measure of assertiveness—the RAS—at post-testing than either the placebo or the no-treatment groups, but the training group also reported a nonsignificant tendency of being less fearful than the other two groups of social criticism and social incompetence, as measured by the Temple Fear Survey (Braun and Reynolds, 1969).Weinman et at. (1972) administered the Fear Survey Schedule-II (FSS-II) (Geer, 1965) to a sample of hospitalized patients diagnosed as schizophrenic. The FSS-II consists of 51 fear items, 13 or 25.9% of which can be considered to be interpersonal in nature. All three treatment conditions—socioenvironmental, desensitization, and relaxation therapy—resulted in a significant decrease in reported fear (p < 0.001).Bates and Zimmennan (1971) during the initial stages of constructing their Constriction Scale (CS) administered the CS and the FSS-II. The term ‘constriction’, incidentally, was used by the authors to “denote non-assertion” (p. 100). Unlike the RAS, a high score on the CS denotes low assertiveness. Correlation coefficients for the CS with the FSS-II of 0.32 for the male sample and 0.13 for the female sample were obtained. The male coefficient was significant at the 0.05 level. These findings may be confounded, as in the case of Weinman et al. (1972), by the inclusion of fear items on the FSS-II unrelated to interpersonal behavior, such as Thunderstorms' and ‘Hypodermic needles’.Morgan (1974) attacked the problem more directly. Drawing on several factor analytic studies of various fear schedules, Morgan constructed a 10-item Social Fear Scale (SFS) from the FSS-III. Using a large, undergraduate sample (n = 261), Morgan reported correlation coefficients for the RAS with SFS of-0.17 for the male sample, ?0.20 for the female sample, and ?0.24 for the sample as a whole.Whereas the magnitude of the correlation coefficients reported by Bates and Zimmerman (1971) may have been attenuated by the inclusion of items unrelated to social situations, Morgan's results may have been attenuated by the tapping of only limited aspects of social fear. Morgan reported that in the construction of the SFS, “The items selected were those which most consistently loaded heavily on ‘social fear’ factors and did not load heavily on other factors” (1974, p. 255). Inspection of FSS-III items not included in the SFS, however, indicated that several had been excluded which may be conceptualized as relevant to assertive behavior, such as ‘Speaking in public’, ‘Feeling angry’, and ‘People in authority’.It is suggested that an expanded SFS which would include all items from the FSS-III conceptually compatible with assertiveness might result in findings of greater predictive value. One purpose of this study is the replication of the Morgan (1974) study. In addition, however, a more comprehensive SFS and another measure of assertiveness will be used to further investigate the relationship between assertiveness and social fear. It is hypothesized that the expanded SFS will correlate inversely with both self-report measures of assertiveness to a degree significantly greater than the correlations of the two measures with the original SFS.