Efficacy of modified {crepid} SDys on the basis of archival organizational data.

This study examined the effect of estimating {crepid} (Cascio & Ramos, 1986) SDy using the original procedure or using modified versions that replaced data obtained specifically for {crepid} SDy estimation with the following organizational archival data: performance evaluations, job analysis ratings, or both evaluations and ratings. Those four SDy estimates and the estimate obtained using Hunter and Schmidt's (1982) 40% rule converged at a level nearly one-fifth of that for the Procedure-B (Burke & Frederick, 1984) SDy. Moreover, judges preferred supplying Procedure-B percentile estimates, but perceived that their {crepid} ratings were more accurate. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Direct estimates of SDy and the implications for utility analysis.

Utility analysis suggests that human resources policies can have an economically significant impact on business organizations. Confidence in such conclusions, however, requires an accurate estimate of SDy. This article provides a validity check on prevailing subjective methods of SDy estimation by directly estimating SDy from unique field data. Using both simulated and field data, the range of potential bias associated with predictor unreliability in regression analyis is illustrated, and how to calculate corrected values is shown. The methodological problems of directly estimating SDy with organizational data is discussed, and a range of estimates for SDy is provided. Direct estimation of SDy yielded values ranging from 74% to 100% of mean salary, which are considerably greater than conventional subjective judgments. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Is cost accounting the answer? Comparison of two behaviorally based methods for estimating the standard deviation of job performance in dollars with a cost-accounting-based approach.

Accurate estimation of the standard deviation of job performance in dollars (SDy) can improve the precision of utility estimates of expected payoffs from personnel programs. The purpose of this study was to compare directly the estimates of SDy obtained using a cost-accounting-based estimate of SDy, the Global Estimation Model, and the CREPID procedure. The study was conducted in a large, soft-drink bottling company. Each method for estimating SDy was applied to the job classification, route salesman, producing three independent estimates of SDy. These estimates were tested for significant differences. Results indicated that the Global Estimation Model estimate and the cost-accounting-based estimate were not significantly different, whereas the estimate produced by the CREPID procedure was significantly smaller. Limitations of the cost-accounting-based estimate are identified and results are discussed in terms of their implications for the theory and practice of utility analysis in organizations. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Individual differences in output variability as a function of job complexity.

The hypothesis was tested that the standard deviation of employee output as a percentage of mean output (SDp) increases as a function of the complexity level of the job. The data examined were adjusted for the inflationary effects of measurement error and the deflationary effects of range restriction on observed SDp figures, refinements absent from previous studies. Results indicate that SDp increases as the information-processing demands (complexity) of the job increase; the observed progression was approximately 19%, 32%, and 48%, from low to medium to high complexity nonsales jobs, respectively. SDp values for sales jobs are considerably larger. These findings have important implications for the output increases that can be produced through improved selection. They may also contribute to the development of a theory of work performance. In addition, there may be implications in labor economics. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Has the problem of judgment in utility analysis been solved?

N. S. Raju et al (1990) presented a new approach to utility analysis that they claimed has the advantage of circumventing the use of expert judges to estimate the standard deviation of job performance in dollars (?Y, or SDy). The present authors demonstrate that the Raju et al approach simply shifts the judgment problem from that of estimating the standard deviation of the criterion (?Y) to that of estimating the coefficient of variation of the criterion (?Y/μY, or SDy/?). The authors also critique 3 arguments advanced by Raju et al: (1) that the most appropriate estimate of the average value of employee job performance is average compensation, (2) that the CREPID method and the 40% and 70% rules represent special cases of the Raju et al approach, and (3) that the practice of correcting the validity coefficient for criterion unreliability in studies using the F. L. Schmidt et al (1979) method results in overestimation of utility. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Refinements in validity generalization methods: Implications for the situational specificity hypothesis.

Using a large database, this study examined 3 refinements of validity generalization procedures: (1) a more accurate procedure for correcting the residual standard deviation (SD) for range restriction to estimate SDp, (2) use of r? instead of study-observed rs in the formula for sampling error variance, and (3) removal of non-Pearson rs. The 1st procedure does not affect the amount of variance accounted for by artifacts. The addition of the 2nd and 3rd procedures increased the mean percentage of validity variance accounted for by artifacts from 70 to 82%, a 17% increase. The cumulative addition of all 3 procedures decreased the mean SDp estimate from .150 to .106, a 29% decrease. Six additional variance-producing artifacts were identified that could not be corrected for. In light of these it was concluded that the obtained estimates of mean SDp and mean validity variance accounted for were consistent with the hypothesis that the true mean SDp value is close to zero. These findings provide further evidence against the situational specificity hypothesis. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A comparison of economic utility estimates for alternative SDy estimation procedures.

Compared the per-selectee and total utility estimates for an assessment center that was used to select midlevel sales managers when 2 consensus-seeking procedures, the F. L. Schmidt et al (see record 1981-02231-001) procedure and 40% and 70% of mean salary, were used for estimating the standard deviation of job performance in dollars (SDy). Using the Schmidt et al procedure as the base, the overall utility estimates varied from 21 to 125% of the Schmidt et al values. For most cases, the different SDy estimation procedures produced somewhat similar utility estimates. The resulting estimated dollar gains from the use of the assessment center to select sales managers were substantial in all cases. The dollar gains from the current selection strategy were, at a maximum, only 63% of estimated top-down selection utility gains. Findings illustrate the value of economic utility analysis for evaluating human resource programs regardless of the type of SDy estimation procedure used. (21 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Two modified procedures for estimating standard deviations in utility analyses.

F. L. Schimdt et al (see record 1981-02231-001) developed a procedure for obtaining rational estimates of the standard deviation of job performance in dollars (SDy). A problem in using their procedure is that it can yield large variation in percentile point estimates across judges. Two modified procedures using consensual feedback were tested with 26 high-level managers to determine if the new procedures could reduce this variability. The normality assumption of the utility estimates yielded by the Schmidt et al and modified procedures was also investigated, and SDy estimates from the Schmidt et al and 2 modified procedures were compared to each other and to a performance and salary distribution. Results indicate that one of the modified procedures offers an improved practical means of obtaining rational SDy estimates in utility analyses. (19 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Selection utility in the occupation of U.S. park ranger for three modes of test use.

Presents an analysis of the selection-utility implications of employing a valid test as opposed to a structured employment interview in the selection of park rangers in the US Park Service. Values of the standard deviation of job performance in dollars (SD–By–b) were estimated using the method presented by the 1st author and colleagues (see record 1981-02231-001). The impact on employee output of 3 modes of selection test use was examined: (a) top–down selection, (b) minimum required test scores equal to the mean, and (c) minimum score at 1 standard deviation (SD) below the mean. Top–down selection produced an increase in average productivity of about 13%. This 13% increase translated into large savings in dollar terms. Where the minimum required score was set at the mean, dollar value of output gains was only 45% as large as dollar value for top-down selection. The percentage increase in output was 5.9%. When the cutoff score was 1 SD below the mean, dollar value of output gains was only 16% of the top–down figure, and the percentage increase in output was only 2.1%. It is concluded that employers using the minimum test score cutoff method of test use are doing so at substantial economic cost. (24 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

The effect of overdominance on characterizing deleterious mutations in large natural populations

Alternatives to the mutation-accumulation approach have been developed to characterize deleterious genomic mutations. However, they all depend on the assumption that the standing genetic variation in natural populations is solely due to mutation-selection (M-S) balance and therefore that overdominance does not contribute to heterosis. Despite tremendous efforts, the extent to which this assumption is valid is unknown. With different degrees of violation of the M-S balance assumption in large equilibrium populations, we investigated the statistical properties and the robustness of these alternative methods in the presence of overdominance. We found that for dominant mutations, estimates for U (genomic mutation rate) will be biased upward and those for h (mean dominance coefficient) and s (mean selection coefficient), biased downward when additional overdominant mutations are present. However, the degree of bias is generally moderate and depends largely on the magnitude of the contribution of overdominant mutations to heterosis or genetic variation. This renders the estimates of U and s not always biased under variable mutation effects that, when working alone, cause U and s to be underestimated. The contributions to heterosis and genetic variation from overdominant mutations are monotonic but not linearly proportional to each other. Our results not only provide a basis for the correct inference of deleterious mutation parameters from natural populations, but also alleviate the biggest concern in applying the new approaches, thus paving the way for reliably estimating properties of deleterious mutations.     

The role of prescribed and nonprescribed behaviors in estimating the dollar value of performance.

Methods for assessing the standard deviation of dollar-valued work performance (SDy) do not address the issue of whether or not performance of nonprescribed behaviors should be included in the estimation of SDy. We used a policy-capturing approach to see whether supervisors considered nonprescribed behaviors when making dollar judgments about work performance. Seventeen supervisors attached dollar values to 50 profiles of hypothetical computer programmers. Each profile consisted of performance ratings on 10 dimensions of prescribed behavior and 3 dimensions of nonprescribed behavior. Results indicated that most of the supervisors do take nonprescribed behaviors into account when making dollar judgments of work performance. Estimates of SDy derived from the policy-capturing method correlated .45 (p?SDy made by subjects, using the global estimation method (Schmidt, Hunter, McKenzie, & Muldrow, 1979). Implications are discussed for both SDy estimation and treatment of nonprescribed work behavior. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Longitudinal assessment of growth in children born to mothers with human immunodeficiency virus infection

OBJECTIVES: To describe and to evaluate the longitudinal growth of children born to mothers with human immunodeficiency virus (HIV) infection. DESIGN: Measurements of weight, length (measured in infants in a recumbent position) and height (measured in older children in an upright position), and head circumference were documented and evaluated longitudinally using generalized estimating equations in a group of children born to HIV-infected mothers. Children infected with HIV were compared with uninfected children and with National Center for Health Statistics standards. SETTING: Primary care clinic in an urban hospital devoted to the medical care of children born to HIV-infected mothers. PATIENTS: One hundred nine children born to HIV-infected mothers, 59 HIV-infected and 50 uninfected, between birth and 70 months of age. RESULTS: The mean birth weights of both groups were below the 50th percentile. While the mean weight-for-age curve of uninfected children attained the 50th percentile by age 24 months, the mean birth weight-for-age curve of HIV-infected children remained below the 50th percentile. Weight gain became significantly different between the two groups by age 36 months. The mean birth length-for-age curves of HIV-infected and uninfected children was also below the 50th percentile. The mean height-for-age curve of uninfected children attained the 50th percentile by age 40 months, while that of HIV-infected children remained well below the 50th percentile. Linear growth between HIV-infected and uninfected children diverged earlier than weight, becoming significantly different by age 15 months. CONCLUSIONS: Although children born to HIV-infected mothers are born with weight and length below the 50th percentile, uninfected children catch up, while HIV-infected children remain below the 50th percentile and experience an earlier and more pronounced decrease in linear growth (height-for-age) than in weight-for-age.     

Social desirability bias in dietary self-report may compromise the validity of dietary intake measures

BACKGROUND: Self-report of dietary intake could be biased by social desirability or social approval thus affecting risk estimates in epidemiological studies. These constructs produce response set biases, which are evident when testing in domains characterized by easily recognizable correct or desirable responses. Given the social and psychological value ascribed to diet, assessment methodologies used most commonly in epidemiological studies are particularly vulnerable to these biases. METHODS: Social desirability and social approval biases were tested by comparing nutrient scores derived from multiple 24-hour diet recalls (24HR) on seven randomly assigned days with those from two 7-day diet recalls (7DDR) (similar in some respects to commonly used food frequency questionnaires), one administered at the beginning of the test period (pre) and one at the end (post). Statistical analysis included correlation and multiple linear regression. RESULTS: Cross-sectionally, no relationships between social approval score and the nutritional variables existed. Social desirability score was negatively correlated with most nutritional variables. In linear regression analysis, social desirability score produced a large downward bias in nutrient estimation in the 7DDR relative to the 24HR. For total energy, this bias equalled about 50 kcal/point on the social desirability scale or about 450 kcal over its interquartile range. The bias was approximately twice as large for women as for men and only about half as large in the post measures. Individuals having the highest 24HR-derived fat and total energy intake scores had the largest downward bias due to social desirability. CONCLUSIONS: We observed a large downward bias in reporting food intake related to social desirability score. These results are consistent with the theoretical constructs on which the hypothesis is based. The effect of social desirability bias is discussed in terms of its influence on epidemiological estimates of effect. Suggestions are made for future work aimed at improving dietary assessment methodologies and adjusting risk estimates for this bias.     

Ridge regression: Application to nonexperimental data.

The value of regression analysis as a statistical tool may be significantly diminished when the set of independent or predictor variables are approximately collinear. The presence of approximate linear relationships in a set of predictors is not unusual for data that arise in social science research. This situation, often referred to as multicollinearity, may be difficult to detect and is even more difficult to remedy. A method called ridge regression has recently been developed and may be useful for solving some of the problems associated with multicollinearity. The ridge estimators have smaller variances (in comparison with those of ordinary regression), but they are biased. When the reduction in variance exceeds the square of the bias, ridge estimates are preferred. In this article, the ridge method is described. Its application is demonstrated by analyzing data from an employee satisfaction survey, (30 departmental units, each with 35 employees and a supervisor), and the ridge solution is compared with the regular regression solution. (27 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Estimation of effects of quantitative trait loci in large complex pedigrees

A method was derived to estimate effects of quantitative trait loci (QTL) using incomplete genotype information in large outbreeding populations with complex pedigrees. The method accounts for background genes by estimating polygenic effects. The basic equations used are very similar to the usual linear mixed model equations for polygenic models, and segregation analysis was used to estimate the probabilities of the QTL genotypes for each animal. Method R was used to estimate the polygenic heritability simultaneously with the QTL effects. Also, initial allele frequencies were estimated. The method was tested in a simulated data set of 10,000 animals evenly distributed over 10 generations, where 0, 400 or 10,000 animals were genotyped for a candidate gene. In the absence of selection, the bias of the QTL estimates was < 2%. Selection biased the estimate of the Aa genotype slightly, when zero animals were genotyped. Estimates of the polygenic heritability were 0.251 and 0.257, in absence and presence of selection, respectively, while the simulated value was 0.25. Although not tested in this study, marker information could be accommodated by adjusting the transmission probabilities of the genotypes from parent to offspring according to the marker information. This renders a QTL mapping study in large multi-generation pedigrees possible.     

Bootstrap Standard Error and Confidence Intervals for the Correlation Corrected for Range Restriction: A Simulation Study.

The standard Pearson correlation coefficient is a biased estimator of the true population correlation, ρ, when the predictor and the criterion are range restricted. To correct the bias, the correlation corrected for range restriction, rc, has been recommended, and a standard formula based on asymptotic results for estimating its standard error is also available. In the present study, the bootstrap standard-error estimate is proposed as an alternative. Monte Carlo simulation studies involving both normal and nonnormal data were conducted to examine the empirical performance of the proposed procedure under different levels of ρ, selection ratio, sample size, and truncation types. Results indicated that, with normal data, the bootstrap standard-error estimate is more accurate than the traditional estimate, particularly with small sample size. With nonnormal data, performance of both estimates depends critically on the distribution type. Furthermore, the bootstrap bias-corrected and accelerated interval consistently provided the most accurate coverage probability for ρ. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Biased estimators, biased experts, and the ivory tower: A reply to Grove (2001).

This reply responds to W. M. Grove's (2001) critique of H. O. F. Veiel and R. F. Koopman's (2001) article on bias in widely used methods of estimating premorbid IQ. In this reply, the authors show that Grove is misrepresenting part of Veiel and Koopman's arguments, extending them to show that the proposed adjustment to regression estimates of IQ not only is unbiased but also is the maximum-likelihood estimate of the true IQ, and argue that Grove's notion of the acceptability of biased methods in judicial proceedings reflects a fundamental misapprehension of their nature and purpose. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A note on the economic utility of generalized validity coefficients in personnel selection.

This study compared economic utility estimates that were based on noninteractive, interactive, independent multiplicative, and Taylor Series Approximation (TSA) 1 and 2 validity generalization results for clerical selection procedures at a large international manufacturing company. On the basis of estimates of the mean true validity and lower bound 90% credibility value, magnitude and percentage differences in resulting utility estimates across validity generalization procedures were relatively small for almost all comparisons. Regardless of the specific validity generalization parameter estimate used in estimating a utility value, the change in economic utility, going from the organization's current selection procedure (i.e., a verbal ability test) to an alternative procedure, was sizable in most cases. These results clearly demonstrate the practical similarity in utility terms of alternative validity generalization procedure results as well as the sizable economic value of minimum-level generalized validity coefficients. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Statistical properties of d ' and β estimates of signal detection theory.

This study provides practical guidelines for estimating and performing inferential tests using sensitivity, d', and response bias, β, estimates of signal detection theory obtained in single-participant designs. In a Monte Carlo study, statistical bias and sampling variability of these estimates were examined across a wide range of distributional overlap. Two factors were manipulated: N (number of presentations of each stimulus) and C (decision criterion along the perceptual axis). Both d' and β estimates were relatively unbiased with small sampling variance for N values of at least 100; for smaller N values, both were positively biased. The d' estimates remained unaffected by C, and β estimates were unaffected by the means of the signal?+?noise distribution (effectively d') throughout their estimable ranges. Limiting cases and recommendations for use are discussed. (PsycINFO Database Record (c) 2010 APA, all rights reserved)