Ronald N. Bone

Extraversion, Field-Dependence, and the Stroop Test

Scores on extraversion, neuroticism, and psychoticism were correlated with each other, and with scores on the rod-and-frame test and the Stroop test, for 97 male and 97 female Ss, and the resulting tables of correlations factor-analyzed for the sexes separately. Of interest for the study of personality was a factor containing loadings on extraversion, field dependence, speed on the Stroop colour and word naming tasks, and Stroop interference. Loadings were much higher and more clearly marked for men than for women. Women were significantly more field-dependent than men. The results are in line with expectation.

EXTRAVERSION, NEUROTICISM, AND SENSATION SEEKING

Zuckerman, Kolin, Price, and Zoob ( 1964) developed the Sensation-seeking Scale (SSS) to measure the construct of optimal stirnulacion or arousal level, with high scores reflecting an impulsive individual that needs change and independence from others. Using Eysenck and Eysencks (1963) measure of extraversion ( E ) , Farley and Farley (1967) predicted and obtained a significant SSS-E relationship. However, Zuckerman and Link ( 1968) found no relationship between E and the SSS. They did find, however, chat the SSS did not correlate (size of rs not mentioned) with neuroticism ( N ) , suggesting sensationseeking is independent of self-admitted pathology. The present smdy investigated relations among the SSS, E and N as measured by Eysencks (1959) Maudsley Personality Inventory (MPI). In addition, the inclusion of males and females allowed for an inspection of possible sex differences, as previous research employed all male samples. A total of 260 undergraduate psychology students (123 women and 137 men) were given the SSS and the MPI in that order. Product-moment correlations were computed, yielding a significant 7 of .23 berween SSS-E ( p < .01) and lending support to Farley, et al. (1967). The correlation of .03 berween SSS and N was nonsignificant, which replicates Zuckerman, et al. ( 1968). SSSE sex differences were minimal for men ( r = .24, p < .01) and women ( r = .23, p < .05); and SSS-N for men ( r = . 0 l ) and women ( r = .03). In light of these findings, it is surprising that Zuckerman, et al. (1968) failed to obtain significance with SSS-E, although their sample might have been biased in some not so obvious manner.

Relationship of Rod-and-Frame Scores to Dream Recall

Equivocal evidence exists as to the relationship of dream recall to field independence. Although previous work (Witkin, Dyk, Faterson, Goodenough, & Karp, 1962) showed field independents recalled more, Montgomery and Bone (1970) employing an Identical-figures Task obcained no such relationship. Considering these findings the present scudy was designed to clarify this area both by employing a large sample and two measures of dream recall. Rod-and-frame scores, average deviacion from vertical over 8 trials, were obtained from 86 male and 90 female introductory psychology students. Ss were asked to race their questionnaire dream recall on a 7-poinc scale, high scores reflecting greater recall. Dream diary recall was assessed by the number of dreams recalled upon awakening over a 24-day period. Product-moment correlations were computed. Rod-and-frame correlations were -.06, -.I8 and -.I1 for the questionnaire measure and .21, -.I4 and .03 for the diary measure for male, female and total data respectively. Field independence was not significantly related to either index for the total group and women, although it was in the predicted direction (Witkin, et al., 1962). While data for men follow the same pattern for the questionnaire index, significance was actained for the diary measure (p < .05) but in the opposite direction (e.g., field-dependenc men recall more dreams). This is difficult co reconcile with Witkins (1970) hypothesis that field-dependents rely on repression and should recall fewer dreams. Repression, however, has recently been questioned as an explanatory mechanism (Bone, Nelson, & McAllister, 1970) for failure to recall dreams. No explanation for these data is proposed.

Dream recall and repression-sensitization.

Several studies employing Welshs Repression scale (Domhoff & Gerson, 1967; Singer & Schonbar, 1961; Tart, 1962) noted sensitizers recalled more dreams than repressors. The present study attempted to extend these findings both by employing a larger sample and by using another measure of repression-sensitization, Byrnes (1961) Repression-Sensitization (RS) scale. Ss were 244 males and 253 females between the ages of 17 and 20 yr., newIy enrolled students. Immediately after completing the RS, as part of a class requirement, Ss estimated their dream recall on a 1 to 7 scale, 7 representing maximum recall. The product-moment correlation between RS ( M = 51.50, SD = 19.79) and dream recall ( M = 4.13, S D = 1.15) was very small ( r = -.012). Separate correlations by sex were also nonsignificant for males ( T = ,011) and females ( r = -.019). Data for males and females were: Mns of 52.24, 50.79 and SDs of 19.46, 20.04; dream recall Ms of 4.00, 4.25 and SDs of 1.12, 1.16. A 2 X 2 extreme-group analysis involving top and bottom 25% of RS scores and sex, with dream recall scores as the dependent variable, yielded nonsignificant factors (sex, F = 2.55, df = 1/232); RS and the RS x sex interaction ( F s < 1.00). The general lack of significance is difficult to explain; Singer & Schonbar (1961) and Tarts (1962) significant results may be due to small Ns and possible biased sampling. Support for this statement is suggested by Domhoff and Gerson (1967) as borderline significance (one-tailed t test) when a larger sample (N = 190) was employed. Assuming acceptable valjdiry of the RS scale (Byrne, 1961), dream recall does not seem related to repression-sensitization as measured.

Dream Recall and Cognitive Style

Schonbar (1965) and Witkin, Dyk, Faterson, Goodenough, and Karp (1962) found dream recall related to field independence. However, these findings may lack generality due to limited Ns. Field independence and dream recall were investigated by employing an alternative measure of the former, the Identical Figures Task (IFT) which consists of a series of 24 geometrical figures, each containing two identical designs which S must locate. The time limit is 4 min. Preliminary data (Pieszko, 1965) indicates reasonable reliability (stability coefficient .79) and validity in terms of IFT-Gottschaldt correlations. Split-half reliability (Bone, unpublished data) is also high ( N = 76, r = .87). The IFT was administered to 65 males and 63 females, and Ss estimated their current dream recall on a 7-point scale. The correlation for the total group was nonsignificant ( r = .05), as were those for males ( s = .09) and females (r = .02). IFT means (5.62, 4.95, 5.29) and standard deviations (3.07, 2.59, 2.85) and dream recall means (3.91, 4.22, 4.06) and standard deviations (1.17, 1.01, 1.10) were obtained for male, female and combined data respectively. A 2 X 2 analysis of variance involving sex and extreme IFT scores (N = 40), wich dream recall scores as the dependent variable, yielded nonsignificant main effects and interaction (Fs < 1.00). These results are not consiscent wich previous findings but should not be taken as conclusive until replicated using other measures of field independence (viz., Rod-and-frame Test).

FIRSTBORNS, ONLY CHILDREN, SEX AND FIELD INDEPENDENCE

Recent critiques of research on birth order (Sampson, 1965; Warren, 1966) have questioned the practice of including only children in the category of oldest or firstborn. Also, many investigators combine scores for males and females despite a growing concern with sex differences. The purpose of this study was to determine the comparability of only and firscborns, of both sexes, with respect to the traic of field independence which has previously been linked to effects of birth order (Stewart, 1967). A 2 by 2 factorial design involving sex and birth order (only versus oldest) was utilized with rod-and-frame scores as the dependent variable. Oldest Ss were selected from volunteers from two-sibling families to control for family size. Each S was tested individually and a field independence score was derived by compucing the mean deviation from the upright over 8 trials. An analysis of the data showed both effects of birth order and sex nonsignificant (Fs < 1.00), however, the birth X sex interaction was significant (F = 4.60, df = 1/36, p < .05 ), with mean rod-and-frame scores of 9.67 and 8.32 for males and 7.44 and 10.85 for females for only and firstborn classifications respectively. These results support previous criticisms (e.g., Sampson, 1965) of combining data for only and firscborns at least for field independence. These data also lend support to the finding of differential birth order-field independence effects (Stewart, 1967).

A human analogue of discrimination contrast

In the present study, one group of human Ss received differential reward conditioning for accurately tracing numbered-nonnumbered star patterns, whereas a second and third control group received either just small or large reward in both discriminanda, respectively. The present study replicated the findings of previous animal contrast studies in that a significant negative S—contrast effect, as well as a trend towards a negative S+ contrast effect, was shown.

A human analogue of discrimination contrast: S− contrast as a function of magnitude of S+

In the present study, two groups of human Ss received differential reward conditioning for accurately tracing red or blue star patterns. A control group received only small reward in both discriminanda. The present study replicated the findings of previous animal and human studies in that a significant negative S− contrast effect was found. In addition, it was shown that this effect increased as a function of the magnitude of reward in the S+ discriminandum. The latter finding has been obtained previously with infrahuman Ss but evidently not with human Ss; hence, the generality of negative discriminative contrast effects have been extended significantly.

Noncontingent Nonreinforcement of a Response as the Aversive Event in a Cer Paradigm

The present experiment investigated the effects of noncontingent nonreinforcement as the aversive event in a CER paradigm. The results showed a significant response-facilitation effect during early training, but none during later training with a high rate-producing, high-density reinforcement schedule. The present results imply that a low rate-producing, high-density reinforcement schedule is not a necessary condition for response facilitation.