Masculinity,Femininity, and Multiple Regression: Comment on Zeldow,Daugherty, and Clark's "Masculinity,Femininity, and Psychosocial Adjustment in Medical Students: A 2-Year Follow-Up"

Inappropriate applications and interpretations of multiple regression analyses by Zeldow and his associates (Zeldow, Clark, & Daugherty, 1985; Zeldow, Daugherty, & Clark, 1987) are discussed as are several general interpretive and statistical issue. To perform and interpret multiple regressors, several characteristics of this statistical approach must be taken into account. Of critical importance is the order in which variables are entered, the choice of regression approaches, the direction of an effect upon entering the model, and the manner in which the interpretation of interaction effects varies as a function of the presence or absence of main effects. Because Zeldow et al. (1985, 1987) did not consider these factors, their regression analyses were not run as stated and are difficult to interpret. Five hypothetical outcomes in androgyny research and an example involving actual data are presented.     

Masculinity,Femininity, and Multiple Regression: A Reply to Holmbeck

We respond to Holmbeck's (this issue) criticism of our work. Holmbeck is correct to observe that we mislabel the hierarchical regression analyses that we actually performed in Zeldow et al. (1985, 1987) as stepwise regressions. We also reported some correlation coefficients in Zeldow et al. (1985) which were incorrect. However, further speculation by Holmbeck that alternate analyses would "dramatically" alter our results and that we are guilty of various interpretive errors are shown to be wrong. Most of Holmbeck's conclusions about the potential misuses of multiple regression analyses in androgyny research are supported. However, the substantive conclusions of our earlier articles remain unchanged.     

Gender roles in relation to assertiveness and eysenckian personality dimensions: Replication with a spanish population sample

The objective of the present study was to evaluate the cross-national replicability of the usual pattern of associations observed in Anglo-Saxon samples between masculinity and femininity on the one hand and difficulty and distress in assertiveness and the major Eysenckian dimensions of personality on the other hand. Participants were 925 SpanishSs (54% females; and 95.5% “European or other white”). Both the masculinity theory of psychological well-being and the notion thathigh femininity would not be implicated in self-assessed psychologicaldistress ordysfunction were supported by the data. Higher-order analysis showed that masculinity loaded highly on Positive Affect, whereas femininity had its primary loading on Constraint. Tentative support was found for the femininity-humility hypothesis advanced by P. B. Zeldow, S. R. Clark, and D. C. Daugherty in 1985.     

Masculinity, femininity, type A behavior, and psychosocial adjustment in medical students

The freshman class of a midwestern medical school completed measures of masculinity and femininity, Type A behavior, and a variety of dependent variables concerning psychological well-being, adjustment, and interpersonal satisfaction. Appropriate statistical treatment of the data revealed strong and consistent masculinity effects on neuroticism, depression, self-esteem, confidence, hedonic capacity, locus of control, and relationship satisfaction. Femininity main effects varied in number as a function of the statistical method employed and involved a more diverse group of variables than is typically reported. Additive androgyny formulations of mental health were supported; balance androgyny formulations were not. No evidence for a Type A X Masculinity effect on adjustment was found. Discussion focuses on the correct interpretation of masculinity and femininity scales, comparability of analysis of variance (ANOVA) and multiple regression statistical analyses, and the viability of the concept of androgyny.     

The Velten Mood Induction Procedure and cognitive manipulation: Our response to Clark (1985)

Clark (1985) makes several thoughtful points about studies using the Veiten Mood Induction Procedure (VMIP; Veiten, 1968). We still disagree with his position, and interpret the evidence as supporting that the VMIP is a cognitive manipulation. Furthermore, our cognitive priming position may offer a more parsimonious and coherent explanation of the VMIP findings on mood induction, memory and other mood-linked behaviors.     

The Detection and Interpretation of Interaction Effects Between Continuous Variables in Multiple Regression

issues in the detection and interpretation of interaction effects between quantitative variables in multiple regression analysis are discussed. Recent articles by Cronbach (1987) and Dunlap and Kemery (1987) suggested the use of two transformations to reduce "problems" of multicollinearity. These transformations are discussed in the context of the conditional nature of multiple regression with product terms. It is argued that although additive transformations do not affect the overall test of statistical interaction, they do affect the interpretational value of regression coefficients. Factors other than multicollinearity that may account for failures to observe interaction effects are noted.     

Personality and Family-Environment Predictors of Self-Reported Eating Attitudes and Behaviors

We assessed the contributions of personality and family environment to variations in self-reported eating attitudes and behaviors. Female college undergraduates (N = 137) completed the NEO Personality Inventory (NEO-PI; Costa & McCrae, 1985). Family Environment Scale (Moos & Moos, 1986), Eating Disorder Inventory (EDI; Garner & Olmsted, 1984), and revised Eating Attitudes Test (EAT-26; Garner, Olmsted, Bohr & Garfinkel, 1982). Correlation and multiple regression analyses showed that among the personality variables, NEO-PI Neuroticism and Extraversion made the largest unique contributions to the EDI subscales and EAT-26. The family-environment measures made significant contributions to those EDI subscales that are reflective of broader emotional and interpersonal problems, rather than eating disorders per se. Suppression effects were found for NEO-PI Extraversion and Neuroticism, underscoring the need for researchers to assess comprehensive sets of etiologic factors—and associations among them—to properly interpret complex predictor/criterion relationships.     

Masculinity, Femininity, and Marital Satisfaction: An Examination of Theoretical Models

ABSTRACT This study sought to investigate the relationship between masculinity, femininity, and marital satisfaction. A number of polynomial multiple regression analyses were performed in an effort to determine the validity of six theoretical models linking sex roles to marital satisfaction. These are the femininity model, masculinity model, sex-typed model, additive androgynous model, interactive androgynous model, and curvilinear model. The sample was composed of 117 couples who completed the Bem Sex-Role Inventory (Bem, 1974) and the Dyadic Adjustment Scale (Spanier, 1976). For men, the results showed that marital satisfaction was related to (a) their self-described levels of femininity and masculinity, (b) the level of self-described femininity of their wives, and (c) the presence of feminine qualities as well as a limited optimal level of masculine qualities which they perceived in their wives. For women, marital satisfaction was associated with (a) the number of self-described feminine qualities and (b) the level of masculinity, as well as an optimal level of femininity, which they perceived in their husbands. Further-more, small actual-ideal discrepancies in levels of masculinity and femininity ascribed to partners constituted reliable predictors of marital satisfaction for both men and women.     

Alcoholism,Aging, and Cognition: A Review of Evidence for Shared or Independent Impairments

Little is known about the combined effects of aging and alcoholism on cognitive function, in spite of the potential importance of this knowledge. Studies reported to date have focused on whether aging and alcoholism have shared mechanisms (i.e., the premature aging hypothesis) or independent mechanisms. Most have concluded that the effects of aging and alcoholism on cognition operate through independent mechanisms. However, there are reasons to doubt this conclusion. Many of the studies were not optimally designed to discover shared mechanisms because they used a limited range of ages, included several different groups in their analyses, or had low statistical power. Furthermore, interpretation of the patterns of main effects and interactions between aging and alcoholism is not straightforward. Assumptions must be made about the nature of the impairments in order to interpret the data. This review uses several theories of cognitive aging as an organizing heuristic for interpreting the extant data on aging, cognition, and alcoholism. Some potential shared mechanisms for cognitive deficits in aging and alcoholism are identified, such as information processing speed and proactive interference in working memory. In contrast, some potential independent mechanisms are also identified, such as decay from working and long-term memory. These mechanisms should be used as a base to guide necessary further research.     

Masculinity, femininity, and psychosocial adjustment in medical students: a 2-year follow-up

As part of a longitudinal study of medical students, relations between masculinity and femininity and psychosocial well-being were investigated over a 21-month interval. Of a class of medical students, 82% (N = 99) completed measures of masculinity and femininity during orientation and, 21 months later, completed a broad array of measures of psychological well-being, interpersonal functioning, humanistic attitudes toward patient care, and alcohol consumption. Hierarchical multiple regression analyses with interaction terms revealed main effects of masculinity on self-esteem, extraversion, and confidence, and main effects of femininity on hedonic capacity, interpersonal satisfaction, sharing of personal problems, and alcohol consumption. Little evidence for additive or balance androgyny formulations was found. Findings for masculinity were consistent with earlier findings from this study and others, but they were weaker in magnitude. Findings for femininity were surprisingly robust and enrich the construct validity of this measure. Measures of impaired mood and of alcohol and drug impairment were also studied in relation to masculinity and femininity. Both variables contribute to the prediction of depressed mood; femininity also contributes to the prediction of drug involvement. Neither scale has sufficient sensitivity or specificity to be used by itself as a test of impairment.     

Masculinity and Femininity: A Bipolar Construct and Independent Constructs

ABSTRACT The present investigation is a reanalysis of data from Antill and Cunningham (1979, 1980, Marsh, Antill, & Cunningham, 1987) consisting of responses to five masculinity-femininity (MF) instruments, two self-esteem instruments, and two social desirability scales Correlations between M and F for the five instruments varied from 23 to approximately - 10, support for distinguishable (nonbipolar) M and F factors was found for four of the instruments Applying confirmatory factor analysis (CFA) and hierarchical CFA (HCFA), the present study examined the dimensionality of MF and the influence of method/ halo effects m response to specific instruments The best fitting model identified three higher order factors, in support of traditional personality theories one factor was a bipolar MF construct, but m support of androgyny theory the other two factors were distinguishable M and F factors The factor structures were reasonably invanant for men and women, and methodological implications of this important finding were examined In subsequent analyses, the higher order MF factors were related to self-esteem, social desirability, and gender in order to further test interpretations of the MF factors.     

Cognition and belief in paranormal phenomena: gestalt/feature-intensive processing theory and tendencies toward ADHD, depression, and dissociation

Belief in paranormal phenomena and cryptids--unknown animals such as Bigfoot--may predispose individuals to interpret real-world objects and events in the same way that eyewitness identification can be biased by unrelated information (P. James and N. Thorpe, 1999). Psychological tendencies toward attention deficit hyperactivity disorder (ADHD), dissociation, and depression, even at subclinical levels, may be associated systematically with particular paranormal or cryptozoological beliefs. The authors evaluated these psychological tendencies using the Conners Adult ADHD Rating Scales (C. K. Conners, D. Erhardt, and E. Sparrow, 1999), the Dissociative Experiences Scale (L. Coleman & J. Clark, 1999), and the Beck Depression Inventory-II (A. T. Beck, 1996). They performed regression analyses against beliefs in ghosts, unidentified flying objects (UFOs), extrasensory perception (ESP), astrology, and cryptids. ADHD, dissociation, and depression were associated with enhanced tendencies toward paranormal and cryptozoological beliefs, although participants who believed in each of the phenomena differed from one another in predictable and psychologically distinguishable ways. Cognitively biasing influences of preexisting psychological tendencies may predispose individuals to specific perceptual and cognitive errors during confrontation of real-world phenomena.     

Feelings of control: Contingency determines experience of action

The experience of causation is a pervasive product of the human mind. Moreover, the experience of causing an event alters subjective time: actions are perceived as temporally shifted towards their effects [Haggard, P., Clark, S., & Kalogeras, J. (2002). Voluntary action and conscious awareness. Nature Neuroscience, 5(4), 382-385]. This temporal shift depends partly on advance prediction of the effects of action, and partly on inferential “postdictive” explanations of sensory effects of action. We investigated whether a single factor of statistical contingency could explain both these aspects of causal experience. We studied the time at which people perceived a simple manual action to occur, when statistical contingency indicated a causal relation between action and effect, and when no such relation was indicated. Both predictive and inferential “postdictive” shifts in the time of action depended on strong contingency between action and effect. The experience of agency involves a process of causal learning based on statistical contingency.     

Right-wing authoritarianism,fundamentalism and prejudice revisited: Removing suppression and statistical artefact

This paper challenges a finding reported by several researchers, that fundamentalism could be associated with a reduction in racial prejudice after controlling for authoritarianism (RWA). We argue that the presence of fundamentalism in the construct definition of the conventionalism cluster of RWA leads to higher associations between fundamentalism and conventionalism than with other aspects of RWA. This creates a statistical artefact that distorts the results of multiple regression analyses that include both fundamentalism and RWA as independent variables. To test this hypothesis, 299 participants completed measures of prejudice as well as fundamentalism and the three RWA clusters (conventionalism, authoritarian aggression, and submission). In regression analyses using fundamentalism and the combined RWA scale we replicate previous findings that when RWA is controlled, higher fundamentalism leads to lower prejudice. After removing the overlapping method variance in the scales, this pattern is eliminated and the commonly observed positive relationship between fundamentalism and prejudice is found. We describe the statistical artefact, its antecedents, and its theoretical implications, and outline how investigations in this important area should proceed.     

Asymmetries in cue competition in forward and backward blocking designs: Further evidence for causal model theory

A hallmark feature of elemental associative learning theories is that multiple cues compete for associative strength when presented with an outcome. Cue competition effects have been observed in humans, both in forward and in backward blocking procedures (e.g., Shanks, 1985) and are often interpreted as evidence for an associative account of human causal learning (e.g., Shanks & Dickinson, 1987). Waldmann and Holyoak (1992), however, demonstrated that cue competition only occurs in predictive, and not diagnostic, learning paradigms. While unexplainable from an associative perspective, this asymmetry readily follows from structural considerations of causal model theory. In this paper, we show that causal models determine the extent of cue competition not only in forward but also in backward blocking designs. Implications for associative and inferential accounts of causal learning are discussed.     

The power of statistical tests for moderators in meta-analysis

Calculation of the statistical power of statistical tests is important in planning and interpreting the results of research studies, including meta-analyses. It is particularly important in moderator analyses in meta-analysis, which are often used as sensitivity analyses to rule out moderator effects but also may have low statistical power. This article describes how to compute statistical power of both fixed- and mixed-effects moderator tests in meta-analysis that are analogous to the analysis of variance and multiple regression analysis for effect sizes. It also shows how to compute power of tests for goodness of fit associated with these models. Examples from a published meta-analysis demonstrate that power of moderator tests and goodness-of-fit tests is not always high.     

Beliefs and inferences: A test of a rational-emotive hypothesis 1. Performing in an academic seminar

Subjects were asked to imagine that they were going to present an academic seminar. They were further asked to imagine (a) that they adhered to a rational belief or an irrational belief; (b) that they had made or had not made an effort in preparing for the seminar and (c) that their performance counted or did not count towards their final examination grade. Whilst in role, subjects were asked to make inferences about various aspects of their performance and the responses of others. While the results supported the hypothesis that imagining that one is holding an irrational belief leads to more negative inferences than holding a rational belief, it was also found that not making an effort in preparing for the siminar led subjects to make more negative inferences than making an effort. In addition, there were several two-way and three-way significant interactions between the independent variables. The results supported Ellis's (1985) recent formulation concerning the complex relationship between events and inferences (A), beliefs (B) and emotional and behavioral consequences of beliefs (C).Windy Dryden PhD is Senior Lecturer in the Department of Psychology, Goldsmiths' College, University of London, England. Julia Ferguson is Research Assistant in the Department of Psychology, Goldsmiths' College, University of London, England. Tony Clark was an Undergraduate Student in the Department of Psychology, Goldsmiths' College, University of London, England. He graduated in 1987.     

Job analysis for managerial selection: A guidelines-based approach

This research note describes a job analysis study which was conducted for the purpose of deriving selection standards for entry-level management jobs in a public utility. It is aimed toward practitioners with the responsibility of conducting legally-defensible job analyses. Each step of the study was guided by the legal and professional guidelines on testing and selection (viz.,Uniform Guidelines, 1978;Principles, 1986;Standards, 1985). Procedures employed to identify job content, job groups, the importance of various work behaviors, and a set of important knowledge, skills, abilities, and personal characteristics (KSAPs) are described.The authors would like to thank the following individuals for their help in the data collection phase of the study described here: Kimberly Clark, Lynn Clark, Gary Hayes, Candice Poole, Susan Rizhaupt, Phillip Randall, Sheri Elmore, and Ann Money.     

Behavioral and Attitudinal Correlates of Masculinity: A National Survey of Male Counselors

In this study the authors investigated the relationship between the Male Role Norm Scales (MRNS; Thompson & Pleck, 1987) and gender-related attitudes and behaviors in a sample of 367 male counselors drawn from American Counseling Association (ACA), formerly the American Association for Counseling and Development (AACD), members. Results indicated that this group of men did not, in general, endorse traditional male roles as measured by the MRNS. When a degree of affiliation with these roles was found, however, it was predictive of violence proneness, intimacy style, and degree of perceived similarity to father. Implications for training and counseling are discussed.