It is a common practice in item response theory (IRT) to treat estimates of item parameters, say "B" circumflex, as if they were the known, true quantities, "B." However, ignoring the uncertainty associated with item parameters can lead to biases and over-confidence in subsequent inferences such as ability estimation,…

Describes and compares eight methods of estimating the parameters of the Zipf distribution. (CLB)

Explores posterior analysis in estimating factor score in a confirmatory factor analysis model with polytomous, censored or truncated data, and studies the accuracy of Bayesian estimates through simulation. Results support these Bayesian estimates for statistical inference. (SLD)

Many evaluation students are still being taught the use of tests of statistical significance without being warned about their limitations. This paper discusses other estimates of treatment effects necessary to interpret between-group differences correctly. Sources to improve evaluation practice are also suggested. (SLD)

Methods by which structural models correct for the effects of attenuation due to measurement error are reviewed, and implications of such disattenuation for interpreting the results of structural equation models are considered. Recommendations are made for improving the practice of disattenuation, and caution is urged in drawing inferences based…

This paper investigates methods that avoid using multiple groups to represent the missing data patterns in covariance structure modeling, attempting instead to do a single-group analysis where the only action the analyst has to take is to indicate that data is missing. A new covariance structure approach developed by B. Muthen and G. Arminger is…

Computation of a one-way analysis of variance (ANOVA) "F" ratio from descriptive statistics in the absence of raw data is corrected from two sources. Means associated with inferential statistical hypotheses are identified as estimable population parameters. (Author)

In many practical applications of item response theory, the parameters of overlapping subsets of test items are estimated from different samples of examinees. A linking procedure is then employed to place the resulting item parameter estimates onto a common scale. It is standard practice to ignore the uncertainty associated with the linking step…

In the process of developing a conditionally-dependent item response theory (IRT) model, the problem arose of modeling an underlying multivariate normal (MVN) response process with general correlation among the items. Without the assumption of conditional independence, for which the underlying MVN cdf takes on comparatively simple forms and can be…

The statistical analysis of processing tree models is advanced by showing how the parameters of estimation and hypothesis testing, based on the likelihood functions, can be accomplished by adapting the expectation-maximization (EM) algorithm. The adaptation makes it easy to program a personal computer to accomplish the stages of statistical…

Bayesian methods for estimating dose response curves in quantal bioassay are studied. A linearized multi-stage model is assumed for the shape of the curves. A Gibbs sampling approach with data augmentation is employed to compute the Bayes estimates. In addition, estimation of the "relative additional risk" and the "risk specific…

Results from two Monte Carlo studies in item response theory (comparisons of computer item analysis programs and Bayes estimation procedures) are analyzed with inferential methods to illustrate the procedures' strengths. It is recommended that researchers in item response theory use both descriptive and inferential methods to analyze Monte Carlo…

Studied the impact of changes in item difficulty on National Assessment of Educational Progress (NAEP) estimates over time through a Monte Carlo study that simulated the responses of 1990 NAEP mathematics respondents to 1990 and 1996 NAEP mathematics items. Results support the idea that these changes have not affected the NAEP trend line.…

The quality of approximations to first- and second-order moments based on latent ability estimates is discussed. The ability estimates are based on the Rasch or the two-parameter logistic model, and true score theory is used to account for the fact that the basic quantities are estimates. (SLD)

Presents a problem that describes a real-world situation from World War II that tries to estimate the number of German tanks, half-tracks, and aircraft, and a simulation of that problem in which the selection of a suitable estimate is less apparent. (ASK)