Whether "more is ever too much" for the number of indicators (p) per factor (p/f) in confirmatory factor analysis (CFA) was studied by varying sample size (N) from 50 to 1,000 and p/f from 2 to 12 items per factor in 30,000 Monte Carlo simulations. For all sample sizes, solution behavior steadily improved (more proper solutions and more…

A sequence of simulations was carried out to aid in the diagnosis and interpretation of equating differences found between random and matched (nonrandom) samples for four commonly used equating procedures: (1) Tucker linear observed-score equating; (2) Levine equally reliable linear observed-score equating; (3) equipercentile curvilinear…

Experimental studies of response rates to mail surveys were reviewed and differences in response by population type were described. Cases were selected for review if they were experimental studies that manipulated a response enhancement factor. Results suggest significant differences in typical response rates for different populations. Higher…

The sampling distribution is a common source of misuse and misunderstanding in the study of statistics. The sampling distribution, underlying distribution, and the Central Limit Theorem are all interconnected in defining and explaining the proper use of the sampling distribution of various statistics. The sampling distribution of a statistic is…

Conventional wisdom suggests the omnibus F-test needs to be significant before conducting post-hoc pairwise multiple comparisons. However, there is little empirical evidence supporting this practice. Protected tests are conducted only after a significant omnibus F-test while unprotected tests are conducted without regard to the significance of the…

Type I error rates were estimated for three tests that compare means by using data from two independent samples: the independent samples t test, Welch's approximate degrees of freedom test, and James's second order test. Type I error rates were estimated for skewed distributions, equal and unequal variances, equal and unequal sample sizes, and a…

As significance testing comes under increasing criticism, some researchers are turning to other indices to evaluate their findings. Included among the alternative options are the interpretation of effect size estimates and the evaluation of sample specificity (invariance testing). Using a hypothetical data set of 64 cases and two predictor…

A specific application of a general paradigm described by R. D. Cook (1986) and R. McCulloch (1985) in assessing local influence is given. Snow geese flock size is estimated as "X" by an observer and "Y" by a photograph. "Y" is believed to be the true flock size. The problem is to obtain true flock size "Z"…

The results of a study of find alternative techniques for testing distributional normality are presented. A group of statistical techniques--some established and some new--were compared using empirical techniques. One new technique which appears to have higher power than the Lilliefors test was subjected to a better definition. Distributions under…

Teacher performance data was used in this investigation to determine whether the stability of variance components and generalizability coefficients could be obtained by increasing sample sizes. A section of the Teacher Performance Assessment Instruments (TPAI) was utilized to assess 2 competencies for each of 750 teachers Each competency, defined…

The distributional characteristics of the Kaiser-Rice measure of sampling adequacy (MSA) were investigated with sample correlation matrices from multivariate normal populations where the level of correlation (LC) was systematically varied. Two additional variables were manipulated--sample size (SS) and number of variables (NV). Ten matrices were…

Four procedures are discussed for obtaining a confidence interval when answer-until-correct scoring is used in multiple choice tests. Simulated data show that the choice of procedure depends upon sample size. (GDC)

A theory-driven approach to validity is proposed. The central argument is that a model or theory should be formulated in a program evaluation and the modeling process should include the identification of potential threats to validity in research. (Author/LMO)

Describes a sampling strategy which involves linked pairs of persons used to obtain bereaved respondents for a study examining loss and coping responses following a recent natural disaster. The sampling procedure appeared not to produce an obvious bias and was very beneficial in meeting the research objectives. (Author/NRB)

This note shows that, contrary to what has been claimed, Bentler and Bonnett's nonnormed fit index is dependent on sample size. Specifically for a constant value of a fitting function, the nonnormed index is inversely related to sample size. A simple alternative fit measure is proposed that removes this dependency. (Author/LMO)