PROMPTING TECHNIQUES TO INCREASE THE RETURN RATE OF MAILED QUESTIONNAIRES1

To increase the return rate of questionnaires mailed to clergy and physicians concerning their mental-health practices, different prompts were used after the questionnaire was received during four mail-outs to four randomly drawn samples of clergy and physicians. For each mail-out, the sample was divided into experimental (received prompt) and comparison (no prompt) groups, and one type of prompt or combination was used. Non-returnees of the questionnaire in the experimental group received either: (a) a single telephone call, (b) a memo, (c) a package (personal letter and new questionnaire) or package plus a telephone call, or (d) a double call. Comparison physicians and clergy were mailed only the original questionnaire. Relative to their respective comparison group's return rate, which averaged 22% across the four mail-outs (range 18% to 24%), the single call and package alone about doubled the overall return rate, the package and call increased the return rate about two-and-a-half fold, and the double call almost tripled the return rate. The memo was ineffective. A cost-effectiveness analysis indicated that the double-call procedure was less expensive than the single call, and much less expensive than the package alone or package with a call in securing returns. An analysis of the pattern of returns showed clearly that when prompts were not delivered (comparison groups), very few returns were received after about seven days from the initial mail-out. Most returns from prompts (experimental groups) were received by several days after the prompt. The results were seen as salient to the problem of reducing selection or volunteer bias in questionnaire studies and subsequent research demonstrating the effectiveness of telephone calls made about a week after distribution of surveys in securing high return rates was discussed.     

A cross-cultural examination of psychological differentiation in males and females1

A review of cross-cultural studies of sex differences in field dependence-independence reveals that ecological, social-conformity, and possibly biological pressures account for such differences in performance as are found between the sexes. The ecological model (pertaining particularly to non-westernized, subsistence-economy groups) is seen as the most parsimonious of the three models put forth, as it includes the other two models within it by implication. Mechanisms by which each of these factors may operate and suggestions for further research in each of these areas are advanced.     

Problem-solving performance of young boys and girls as a function of task appropriateness and sex identity

The problem-solving performance of young boys and girls was examined in three conditions in which good performance was indicated as being appropriate for boys, for girls, or for both boys and girls. The relations between task scores and five sex-role measures were also examined. As predicted, the boys performed significantly more poorly when the task was indicated as one that girls did better. The girls' performance was not affected by the conditions of the study. The girls' problem-solving scores did not differ significantly from the boys' scores. There was some relationship between the task scores and the sex-role measures for boys, but not for girls. Associations among the sex-role measures were not high.This research was supported by a DGES-FCAC research grant from the government of Quebec, Canada.     

Sex roles and the psychiatric referral process

Three studies were undertaken to explore the relationship between sex-role standards and the psychiatric referral process. Based on observations from previous literature it was hypothesized that children who exhibited behavioral characteristics inappropriate to their sex would be more likely to be referred to psychiatric facilities than would children who exhibited behavioral characteristics appropriate to their sex. The first study examined the records of an outpatient child-guidance clinic. In accord with the hypothesis, it was found that more boys than girls were referred for being emotional or passive and more girls than boys were referred for being defiant and verbally aggressive. In the two subsequent studies, samples of parents and graduate students in clinical and school psychology read hypothetical case studies in which identical behavior problems were attributed either to a boy (Bob) or to a girl (Barbara). The data from the two samples indicated that the child exhibiting the behavior inappropriate to his/her sex was seen as more severely disturbed, as more in need of treatment, and as having a less successful future than the child exhibiting sex-role appropriate behaviors.The first phase of the research was done in collaboration with Merrily Kaplan, whose ideas contributed much to the thrust of the rest of the research. The second phase is based on the first author's undergraduate honors thesis which was supported by NIMH Undergraduate Research Fellowship MH 08027. The authors gratefully acknowledge the assistance of Dorothy Bianco and William Cutler in distributing questionnaires, the cooperation of the New Britain Child Guidance in permitting us to examine their records, and the helpful comments of Benjamin Braginsky, Lorelei Brush, Abigail Stewart, and David Winter, who read an earlier draft of this paper.     

Motivational Factors Affecting Fluctuations of Female Sexual Activity at Menstruation

The hypothesis was tested that pre-menstrual and post-menstrua I increases in sexual activity are due to anticipation of deprivation and rebound following deprivation (Ford & Beach, 1951). Thirty-five married women completed daily questionnaires about their sexual activity for at least two menstrual cycles. Evidence supported the hypothesis that the post-menstrual peak in intercourse is a rebound phenomenon. There was less evidence to support the hypothesis that a pre-menstrual rise represents anticipation of deprivation. In addition, there was a tendency to increase autosexual activity during periods of sexual abstention, which may represent a compensation phenomenon.     

Daydreaming: a measurable concept

Research has been reported which supports the psychometric properties of the Imaginal Processes Inventory. The purpose was to confirm and extend this research as well as investigate the interrelationships between daydreaming and depression, locus of control, and visual imagery. This inventory (7 scales), Beck Depression Inventory, Rotter Locus of Control Scale (I-E), and Gordon's Test of Visual Imagery Control were administered to 100 female undergraduates. A sample of 39 subjects were retested on the scales an average of 6.8 wk. after the first administration. Correlations with Imaginal Processes Inventory and test-retest data were consistent with Giambra's (1977) findings for males and support the reliability and generalizability of the scores. Other results include significant intercorrelations between the various scales of this inventory and the others. Implications for an understanding of the process of depression are discussed.     

Invariance tests for bisection and fractionation scaling

If we are to accept the power function as the psychophysical law, then estimates of the exponent of a particular form of the power law should be independent of changes in basic independent variables. In the present study, various power law hypotheses were tested for the bisection and fractionation scaling of brightness. The results indicated that for bisection, estimates of exponents were dependent on the particular interval bisected, suggesting rejection of the simple power law for bisection. For fractionation, two additional forms of the power law were tested, each form involving a threshold parameter. One was the Φ-law (involving a translation on the intensity axis), and the other was the ψ-law (involving a translation on the psychological axis). The Φ-law provided a poor fit to the data, whereas the ψ-law appeared to fit well for at least one S. Analysis for individuals showed that for all five Ss, the variance due to standards was appreciably larger for the Φ-law.