Effects of nonnormal data on parameter estimates and fit indices for a model with latent and manifest variables: An empirical study

Research in covariance structure analysis suggests that nonnormal data will invalidate chi‐square tests and produce erroneous standard errors. However, much remains unknown about the extent to and the conditions under which highly skewed and kurtotic data can affect the parameter estimates, standard errors, and fit indices. Using actual kurtotic and skewed data and varying sample sizes and estimation methods, we found that (a) normal theory maximum likelihood (ML) and generalized least squares estimators were fairly consistent and almost identical, (b) standard errors tended to underestimate the true variation of the estimators, but the problem was not very serious for large samples (n = 1,000) and conservative (99%) confidence intervals, and (c) the adjusted chi‐square tests seemed to yield acceptable results with appropriate sample sizes.     

Evaluating Intercept-Slope Interactions in Latent Growth Modeling

The effects of misspecifying intercept-covariate interactions in a 4 time-point latent growth model were the focus of this investigation. The investigation was motivated by school growth studies in which students' entry-level skills may affect their rate of growth. We studied the latent interaction of intercept and a covariate in predicting growth with respect to 3 factors: sample size (100, 200, and 500), 4 levels of magnitude of interaction effect, and 3 correlation values between intercept and covariate (.3, .5, and .7). Correctly specified models were examined to determine power and Type I error rates, and misspecified models were examined to evaluate the effects on power, parameter estimation, bias, and fit indexes. Results showed that, under correctly specified models, power increased as expected with increasing sample size, larger magnitude of interaction, and larger intercept-covariate correlation. Under misspecification, omitting a non-null interaction results in significant change in the estimation of the direct effects of both covariate and intercept in both magnitude and direction, with results dependent on sign of parameter values for main effects and interaction. Including a spurious interaction does not affect estimation of direct effects of intercept and covariate but does increase standard errors. The primary problem in ignoring a non-null interaction lies in misinterpretation of the model, as interactions yield important insights into the nature of the processes being studied.     

A non-linear model of information sharing practices in academic communities

A new model of information sharing practices in academic communities is based on Latour's circulatory system of scientific facts, and some elements of Foster's non-linear model of information-seeking behavior. The main proposition of this model is that information-sharing practices and context simultaneously shape each other. The proposed model supports Foster's conceptualization of information practices as non-linear processes, but its emphasis on the interdependence between context and information practices provides a more effective means to capture complex negotiations involved in information-sharing practices. The proposition is that the major reason for nonlinearity in information practices is a continuous shifting of actors' interests, pressures, and concerns. Capturing these dynamic relations becomes possible through this model. The model also offers a way to generate a number of research questions and hypotheses, and as such it could be a useful tool for empirical studies on information sharing in academic communities.     

Relationships of hip abductor strength,neuromuscular control,and hip width to femoral length ratio with peak hip adduction angle in healthy female runners

ABSTRACT

A large peak hip adduction angle during running is a risk factor for several overuse injuries in women. The purpose of this study was to determine if female runners with a large peak hip adduction angle have differences in eccentric hip abductor muscle strength, hip neuromuscular control, and/or hip width to femoral length ratio (HW:FL) compared to those with a small angle. Hip adduction during running, hip strength, hip control, and HW:FL were measured in sixty healthy female runners (1.66 ± 0.06 m; 63.2 ± 8.3 kg; 27 ± 6 years). Data from twenty runners with the largest and twenty with the smallest peak hip adduction angles were analysed. Between-group differences in hip strength, control, and HW:FL were determined using independent t-tests (p < 0.05). Variables that were significantly different between groups were entered into a regression model. Runners in both groups had similar hip strength (p = 0.90) and control (p = 0.65). HW:FL was greater in the large peak angle group (p = 0.04), but only explained a small amount of peak hip adduction angle variance for all sixty runners (R² = 0.05). Alarge peak hip adduction angle in some healthy female runners may simply be instinctive as there were no deficiencies in the strength or neuromuscular control constructs assessed.     

Adding results to a meta-analysis: Theory and example

Meta-analysis has been used as a research method to describe bodies of research data. It promotes hypothesis formation and the development of science education laws. A function overlooked, however, is the role it plays in updating research. Methods to integrate new research with meta-analysis results need explication. A procedure is presented using Bayesian analysis. Research in science education attitude correlation with achievement has been published after a recent meta-analysis of the topic. The results show how new findings complement the previous meta-analysis and extend its conclusions. Additional methodological questions adddressed are how studies are to be weighted, which variables are to be examined, and how often meta-analysis are to be updated.     

九月的一天

As on every weekday Tom Tuckett was awakened by his annoying, buzzing alarm clock at 5:30. He got out of bed and proceeded to his bathroom for his daily cold shower to wake him up. He turned the knob all the way to the right to make sure the water would be as cold as it could be. He had started this ritual when he moved to New York to take a teaching job at PS 181 in Manhattan.     

More on time series designs: A reanalysis of mayer and Kozlow's data

A recent article by Mayer and Kozlow introduces a time series experiment to science education researchers. This article examines in greater detail design considerations and reanalyzes their data using time series analysis. The first applications of time series experimental methodology to science education research were reported by Mayer and Lewis (1979) and by Mayer and Kozlow (1980). This article discusses Mayer and Kozlow's work and presents a reanalysis of their data which is intended to be heuristic, not critical. (All data were supplied by Dr. Mayer and permission was granted to use the data.) Both design and analysis will be examined. It is important to distinguish time series experimental design from time series analysis. The former, as presented by Campbell and Stanley (1966) and detailed by Glass, Willson, and Gottman (1975), is concerned with structuring treatment conditions and observations over time to make causal inferences. Time series analysis is a set of statistical techniques from which an appropriate technique can be selected to analyze collected data in the framework of a time series design. Just as data from in a randomized experiment might be analyzed using a one-way analysis of variance or a Kruskal- Wallis H-test, data from a time series design may be subjected to several alternative analysis techniques.     

A meta-analysis of the relationship between science achievement and science attitude: Kindergarten through college

A meta-analysis of research results for correlation between science achievement and attitude was conducted. Forty-three studies were utilized and 15 variables were coded for each correlation found in each study, yielding 280 coefficients. Overall relationship is moderate (0.16), with differences between elementary, junior high, senior high, and college subjects. Causal ordering results support achievement causing attitude in grades three to eight and mixed results thereafter. Sex differences exist through high school. Other effects are mixed across grade level.     

Measurement Invariance Across Groups in Latent Growth Modeling

This Monte Carlo study investigated the impacts of measurement noninvariance across groups on major parameter estimates in latent growth modeling when researchers test group differences in initial status and latent growth. The average initial status and latent growth and the group effects on initial status and latent growth were investigated in terms of Type I error and bias. The location and magnitude of noninvariance across groups was related to the location and magnitude of bias and Type I error in the parameter estimates. That is, noninvariance in factor loadings and intercepts was associated with the Type I error inflation and bias in the parameter estimates of the slope factor (or latent growth) and the intercept factor (or initial status), respectively. As noninvariance became large, the degree of Type I error and bias also increased. On the other hand, a correctly specified second-order latent growth model yielded unbiased parameter estimates and correct statistical inferences. Other findings and implications on future studies were discussed.