A four-factor structure for parametric tests on independent groups.

A prior article by P. A. Games (1978) presented numerous statistical hypothesis-testing procedures for independent groups. These resulted from crossing 3 factors: technique used, method of control of Type 1 errors, and parameter used. Tests on means, correlations, proportions, and variances are represented by using θ for the parameter, and θ? for the statistic. The prior article was limited to cases in which the variance of the statistic V k(θ) was a constant and hence was limited to equal ns. The present article contains the extention to unequal V k(θ?) cases and hence permits unequal ns. More general large-sample forms of the t and F statistics, namely, the t′ and the df uF′ statistics, are recommended for the unequal V k(θ?) cases. The B. L. Welch solution for general contrasts on means is presented as a minor adaptation of t′. The Welch solution for an omnibus test on means is a minor adaptation of df uF′. Examples are given that illustrate the application of t′ and df uF′. (27 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

The effects of dieting on eating behavior: A three-factor model.

Reviews and critiques restraint theory and develops a 3-factor model of dieting behavior. The factors (frequency of dieting and overeating, current dieting, and weight suppression) are embedded within a 3-dimensional grid that also considers mechanisms mediating the effects of dieting and the influence of weight status. The eating behavior exhibited by restrained eaters stems from their frequent dieting and overeating in the past rather than from their current state of dietary or cognitive restraint. Evidence is reviewed indicating that current dieting and weight suppression have different effects on eating than does restraint. The 3-factor model is used to reinterpret findings consistent with restraint theory and to explain findings inconsistent with restraint theory. Finally, clinical and research implications of the 3-factor model are discussed. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Sources of stress in policewomen: A three-factor model.

It has been argued that models of stress need to be occupation specific. Policewomen are an occupational subgroup who experience stress differently from male officers. Sixteen sources of felt stress were rated by 206 policewomen. Exploratory factor analysis suggested a three-factor model of operational, interpersonal, and management/organizational stress. Confirmatory factor analysis with a second sample of 213 policewomen confirmed the three-factor structure. Interpersonal stressors accounted for most variance in ratings of felt stress. This suggests that the interpersonal climate experienced by female officers contributes significantly to rated stress levels, and consideration of this finding should be given in interventions to reduce work stress for female officers. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A note on measurement scales and statistical tests.

The author defends the use of parametric tests (Boneau, 1960), and has been challenged on more than one occasion to justify the use of the t test in many typical psychological situations where there are measurement considerations. Intelligence is often given as an instance, the point being that intelligence is actually measured by an ordinal scale, that equal differences between scores represent different magnitudes at different places on the underlying continuum. This is seen as somehow invalidating the use of the t test with such scores. Burke (1953) has presented an argument which should have ended further discussion, but, in view of the present concern, a restatement of the argument and the addition of a few comments would seem indicated. The present concern seems to have been stimulated by the publication by psychologists of two texts in the field of statistics (Senders, 1958; Siegel, 1956) both of which are organized around Stevens' (1951) system of classifying measurement scales. Siegel and Senders belabor the point that parametric statistics, specifically the t and F tests should be avoided when the measurement scales are no stronger than ordinal, a state of affairs purportedly typical in psychology. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Single-sample tests for many correlations.

When more than one correlation coefficient is tested for significance in a study, the probability of making at least one Type I error rises rapidly as the number of tests increases, and the probability of making a Type I error after a Type I error on a previous test is usually greater than the nominal significance level used in each test. To avoid excessive Type I errors with multiple tests of correlations, it is noted that researchers should use procedures that answer research questions with a single statistical test and/or should use special multiple-test procedures. A review of simultaneous-test and multiple-test procedures for correlations (e.g. Bartlett and Rajalakshman's test, multistage Bonferroni procedure, union-intersection tests, and the rank adjusted method) is presented, and several new procedures are described. (40 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Randomized single-subject experiments and statistical tests.

The importance of intensive study of the individual subject in counseling research and the importance of the development of appropriate designs and analytical procedures for single-subject research has been recognized. Single-subject experiments are not always possible or ethical, but when they can be performed, treatment times should be randomly assigned to treatments, as this randomization is essential to provide the basis for a statistical test. The assumption of random sampling is not met in single-subject experiments. However, the data from a randomized single-subject experiment can be analyzed by use of a randomization test because such tests do not require random sampling. Randomized single-subject experiments can be helpful in determining what is best for a particular client, and systematic replications of such experiments can help in generalizing to other subjects. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

The power of statistical tests in meta-analysis.

Calculations of the power of statistical tests are important in planning research studies (including meta-analyses) and in interpreting situations in which a result has not proven to be statistically significant. The authors describe procedures to compute statistical power of fixed- and random-effects tests of the mean effect size, tests for heterogeneity (or variation) of effect size parameters across studies, and tests for contrasts among effect sizes of different studies. Examples are given using 2 published meta-analyses. The examples illustrate that statistical power is not always high in meta-analysis. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Boyhood gender identity development: A statistical contrast of two family groups.

Investigates the genesis of boyhood cross-gender behaviors in an analysis that compares cross-sectional data on families of boys with extensive cross-gender behavior with data on families of boys with more masculine behaviors. A measure of extent of boyhood femininity is developed by discriminating these two groups according to measures of each boy's sex-typed behaviors. Two methods are used in evaluating the explanatory value of variables hypothesized as affecting boyhood gender identity: (a) Between-group differences on these variables clarify potential sources of atypical gender identity in males; (b) Within-group correlations between these variables and extent of boyhood femininity yield information on similarities and differences in the origins of boyhood cross-gender behaviors for each group. No definitive support is found for either modeling or social learning theories of gender identity development. However, striking differences in within-group correlations between the two family groups suggest a qualitative difference in the processes that produce gender identity among boys with typical versus atypical amounts of cross-gender behavior. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

The statistical analysis of data from small groups.

The authors elaborate the complications and the opportunities inherent in the statistical analysis of small-group data. They begin by discussing nonindependence of group members' scores and then consider standard methods for the analysis of small-group data and determine that these methods do not take into account this nonindependence. A new method is proposed that uses multilevel modeling and allows for negative nonindependence and mutual influence. Finally, the complications of interactions, different group sizes, and differential effects are considered. The authors strongly urge that the analysis model of data from small-group studies should mirror the psychological processes that generate those data. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

The converse inequality argument against tests of statistical significance.

Critics have put forth several arguments against the use of tests of statistical significance (TOSSes). Among these, the converse inequality argument stands out but remains sketchy, as does criticism of it. The argument states that we want P(HΔ) (where H and D represent hypothesis and data, respectively), we get P(DΗ), and the 2 do not equal one another. Each of the terms in 'P(DΗ)?≠?P(HΔ)' requires clarification. Furthermore, the argument as a whole allows for multiple interpretations. If the argument questions the logic of TOSSes, then defenses of TOSSes fall into 2 distinct types. Clarification and analysis of the argument suggest more moderate conclusions than previously offered by friends and critics of TOSSes. Furthermore, the general method of clarification through formalization may offer a way out of the current impasse. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Misuse of statistical tests in three decades of psychotherapy research.

Reviews the misuse of statistical tests in psychotherapy research studies published in the Journal of Consulting and Clinical Psychology in the years 1967–1968, 1977–1978, and 1987–1988. It focuses on 3 major problems in statistical practice: inappropriate uses of null hypothesis tests and p values, neglect of effect size, and inflation of Type 1 error rate. The impressive frequency of these problems is documented, and changes in statistical practices over the past 3 decades are interpreted in light of trends in psychotherapy research. The article concludes with practical suggestions for rational application of statistical tests. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

An empirical comparison of statistical tests for assessing the proportional hazards assumption of Cox's model

In the analysis of survival data using the Cox proportional hazard (PH) model, it is important to verify that the explanatory variables analysed satisfy the proportional hazard assumption of the model. This paper presents results of a simulation study that compares five test statistics to check the proportional hazard assumption of Cox's model. The test statistics were evaluated under proportional hazards and the following types of departures from the proportional hazard assumption: increasing relative hazards; decreasing relative hazards; crossing hazards; diverging hazards, and non-monotonic hazards. The test statistics compared include those based on partitioning of failure time and those that do not require partitioning of failure time. The simulation results demonstrate that the time-dependent covariate test, the weighted residuals score test and the linear correlation test have equally good power for detection of non-proportionality in the varieties of non-proportional hazards studied. Using illustrative data from the literature, these test statistics performed similarly.     

Mitochondrial sulfhydryl groups. A possible endogenous probe of conformational changes in the mitochondrial membrane

The protein-bound sulfhydryl (SH) groups of the mitochondrial membrane were determined with Ellman's reagent in energized and non-energized configurational states of mitochondria and submitochondrial particles. When beef heart mitochondria were energized by respiration, there was a decrease in titratable protein-bound SH groups which varied according to substrate: NADH-linked substrates induced a decrease of about 10 nmol per mg of protein,succinate about 7, and ascorbate-tetramethyl-p-phenylene-diamine about 3. Similar changes occurred in phosphorylating submitochondrial particles. A decrease in SH titer was also observed in non-energized conditions, induced by hypotonic treatment and by some reagents inhibiting electron transport and oxidative phosphorylation and inducing orthodox configuration. These changes in protein-bound SH groups might be useful in analyzing the conformational states of mitochondrial membranes.     

Using significance tests to evaluate equivalence between two experimental groups.

Equivalency testing, a statistical method often used in biostatistics to determine the equivalence of 2 experimental drugs, is introduced to social scientists. Examples of equivalency testing are offered, and the usefulness of the method to the social scientist is discussed. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Understanding roles: A psychodynamic model for role differentiation in groups.

The basic roles that emerge in groups are described in terms of a typology based on 3 biological imperatives: evaluation (i.e., good or bad), gender, and hierarchy. The mental matrix model based on these imperatives identifies 14 cardinal and emotional deep roles. The author argues that deep roles are born out of a biogenetically explicated mythic (narrative) structure of mind that is modified by early child–family interactions. In interaction with the environment, these primordial role fantasies become features of group reality. Their basis in the evolution of the human species and their imprint in folklore as well as in personality structure are illustrated. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A more flexible regression-to-the-mean model with possible stratification

We consider a regression-to-the-mean model that includes both additive and multiplicative treatment effects. We allow either or both of these treatment effects to be stratified by ranges of the first measurement. We focus on the situation where there is a very large sample on the first measurement and a relatively small subsample for the second measurement is selected, which often occurs in screening trials. We propose some asymptotically efficient estimators for the parameters of the model that are very simple to compute. We begin with a discussion of the full model, and more on tests and estimation for reduced models follows. An example from a large screening trial is discussed.     

A competency-based model for teaching skills in the administration of intelligence tests.

The Criteria for Competent WISC-R Administration (CCWA) is introduced as a cost-effective instrument for assessing competence in administering the WISC-R. A training model demonstrated an effective way to teach such skills by utilizing the CCWA as a measuring and a teaching device. After 3 administrations of the WISC-R, this model brought 8 graduate students to criterion level for competent administration. (12 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A possible integrative model for the systematic and measurable improvement of education.

Offers a possible rationale for cooperation and interdependence between professional educational practitioners who are working to improve the products and processes of education. A possible generic model for educational management is presented, identifying 6 steps for problem solving. Tools currently being used for the quantifiable improvement of education are briefly discussed and an attempt is made to relate these tools to the suggested process model of education. A possible taxonomy of educational planning is also offered. (29 ref.) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A model of emergence and dysfunctional effects of emotional conflict in groups.

The current study tested a model of group effectiveness in which emotional conflict, negative affective tone in groups, and group mean agreeableness were proposed as key antecedent variables. Data collected from 84 project groups supported the proposition that group-level agreeableness was significantly related with emotional conflict, and that emotional conflict indirectly affects group outcomes through group-level negative affective tone and contextual-discretionary behaviors, as opposed to task-related behaviors. Findings are discussed in terms of how group members' personality foster emotional conflict in groups and the implications of these findings for group effectiveness research as well as the management of project groups. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Make today count: A collaborative model for professionals and self-help groups.

Presents a narrative report of a field study in which members of a self-help research team provided services of significant, positive impact to one chapter of Make Today Count, a self-help group composed of cancer patients, their spouses, and health care professionals. A collaborative model is formulated. Important elements of this model include adequate knowledge of self-help processes, enhancement of rapport through acceptance of the group, and adoption of a consultive approach. Nonproductive modes on interaction observed between other professionals and chapter members are discussed. The present results are contrasted with negative results reported previously by professionals who attempted to collaborate with a similar self-help population, but who applied a different model. (16 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)