George M. Alliger

Construct validation of two instruments designed to measure job involvement and work centrality

The authors clarified the relationship between 2 concepts that have been confused, and often used interchangeably, in the literature. Job involvement was defined as the degree to which one is cognitively preoccupied with, engaged in, and concerned with ones present job. Work centrality was defined as the degree of importance that work, in general, plays in ones life. Questionnaire measures consistent with these definitions were constructed and tested. Data were collected from 313 human services employees. Confirmatory factor analysis demonstrated that, consistent with the authors hypothesis, job involvement and work centrality appear to be 2 distinct constructs. There was also evidence to confirm the hypothesis that there are 2 distinct components of job involvement: job involvement-role and job involvement-setting

Type II Error Problems in the Use of Moderated Multiple Regression for the Detection of Moderating Effects of Dichotomous Variables

Monte Carlo simulation procedures were used to assess the power of moderated multiple regression (MMR) to detect the effects of a dichotomous moderator variable under conditions of: (1) between- group differences in within-group relationships between two variables (i.e., |ρ XY (1) -ρ XY (2) |= .20, .40, .60); (2) different combined sample sizes for the two groups (N 1 + N 2 = NT = 30, 60, 90, 180, 300); and (3) differing proportions of cases (P-i ) in the two groups (i.e., P 1 = .10, .30, .50). Results showed that, consistent with our a priori predictions, the power of MMR increased as: (1) total sample size (NT) increased; (2) the difference between within-group correlation coefficients increased; and ( 3) the difference between the proportion of cases in each group decreased. Moreover, the simulation showed that these three variables had interactive effects on power. The major implication of our findings is that in cases where tests of moderating effects are conducted with MMR and the proportion of cases in each group differs greatly, inferences of no moderating effect may be erroneous: Such inferences may be the result of low statistical power rather than the absence of a moderating effect.

Toward an ADA-appropriate job analysis

Abstract This article describes how current job analysis methods can be used to incorporate the mandates of the Americans with Disabilities Act (ADA) of (1990). First, terms in the ADA such as “qualified individuals with disabilities,” “essential functions,” and “reasonable accommodation” are discussed in reference to their implications for job analysis. Second, survey results gathered from individuals with disabilities are presented to suggest some ways job analysis can be used to identify essential functions and reasonable accommodations. Third, some general suggestions are offered for performing an ADA-appropriate job analysis. Finally, several current job analysis techniques chosen to be representative of the field are evaluated on three criteria: (a) whether the technique has an outcome versus a process focus, (b) whether the technique identifies essential functions, and (c) whether the technique identifies potential reasonable accommodations for a particular job. The job analysis techniques evaluated include: (a) Functional Job Analysis (FJA) (Fine 1988), (b) task analysis (e.g., CODAP, standard task analyses), (c) Threshold Traits Analysis (TTA) (Lopez et al. 1981), (d) the Position Analysis Questionnaire (PAQ) (McCormick, Jeanneret, & Mecham 1972), and (e) the Critical Incident Technique (CIT) (Flanagan 1954).

Reactions to Overt Integrity Test Items

A total of 102 participants rated overt integrity test items on three dimensions (invasion of privacy, ease of faking, job relatedness). Integrity test items were grouped into seven categories. Participants identified items that required direct admissions of counterproductive behaviors as being the most invasive but also the easiest to fake. Job relatedness and invasion of privacy were both found to be correlated with ease of faking perceptions. Implications of these findings and suggestions for future research are discussed.

Do Zero Correlations Really Exist among Measures of Different Intellectual Abilities

Whether measures of different intellectual abilities are positively intercorrelated has been a topic of interest and debate since the turn of the century. The data from an article by Guilford pivotal to this debate are reexamined. It is argued that, contrary to the original claim of the article, the data set of over 7,000 correlations does not provide support for the existence of zero correlations among tests of intellectual abilities.

The Small Sample Performance of Four Tests of the Difference Between Pairs of Meta-Analytically Derived Effect Sizes

Four tests for the difference between pooled estimators of effect size from separate meta-analyses are discussed, and small sample performance compared via Monte Carlo simulation. In terms of Type I error and power, a z-test based on the difference between pooled estimators appears most accurate, while a confidence interval approach modified to permit testing between two mean effect sizes and Hedges and Olkin’s (1985) chi-square test with one degree of freedom appear roughly equivalent, and each of these is superior to a z-test based on an r-transformation. The modified confidence interval approach has particular promise for graphical examination of presumed moderators in meta-analyses.

Confounding among Measures of Leniency and Halo.

Because the variability of rating distributions decreases with leniency, common indices of the psychometric performance of rating data may be confounded. Lenient rating distributions, for example, should show lower average scale intercorrelations (a measure of halo) than do less lenient rating distributions. Although psychometric theory would predict such confounding, it does not permit the estimation of the magnitude of this phenomenon. Consequently, the interrelationships among halo and leniency measures were examined using simulated rating data. Specifically, effects of leniency on 3 levels of halo were examined. As expected, results showed that as leniency increased, halo decreased when measured by dimension intercorrelations but increased when measured by standard deviations across dimensions. The magnitude of the convergent validities calculated for halo and leniency measures also decreased as leniency increased. In addition, although halo and leniency were orthogonal when no leniency was present, substantial confounding of these constructs occurred as leniency increased. Implications of these results for the use and integration of the various measures are discussed.

A Proposed Method to Combine Ranking and Graphic Rating in Performance Appraisal: The Quantitative Ranking Scale

A method is proposed to combine a relative and an absolute approach to performance appraisal; specifically, graphic rating and ranking are combined. In two studies which examined this method, called the Quantitative Ranking Scale (QRS), less leniency generally was found in QRS than in graphic rating. The psychometric performance of the QRS was best when the distinction between the QRS and pure ranking was least. Nonetheless, the system would appear to hold promise as an alternative to graphic rating that could easily be adapted to other rating formats, such as behavioral anchoring or computerized rating.

Generating Correlated Bivariate Random Normal Standard Scores in Quickbasic

A Quick BASIC program for generating random bivariate normal standard scores with given correlation is described. The program uses the Box-Muller (1958) generation method. Compiled run-time and source code versions of the program are available.

A Program for Testing Categorized Data for Normality

Most current statistics packages do not appropriately compute tests for normality on categorized data. This program for personal computers calculates two test statistics and associated significance levels for categorized data: the Kolmogorov-Smimov D, and an omnibus test based on skew and kurtosis. This program would be useful in examining, for example, sample test score or rating distributions.