Identifying mediators in variable chains as part of a causal mediation analysis can shed light on issues of causation, assessment, and intervention. However, coefficients and effect sizes in a causal mediation analysis are nearly always small. This can lead those less familiar with the approach to reject the results of causal mediation analysis.…

Educational researchers commonly use the rule of thumb of "design effect smaller than 2" as the justification of not accounting for the multilevel or clustered structure in their data. The rule, however, has not yet been systematically studied in previous research. In the present study, we generated data from three different models…

Although the use of simulation to teach the sampling distribution of the mean is meant to provide students with sound conceptual understanding, it may lead them astray. We discuss a misunderstanding that can be introduced or reinforced when students who intuitively understand that "bigger samples are better" conduct a simulation to…

Developed for users of the Longitudinal Surveys of Australian Youth (LSAY), this user guide consolidates information about the LSAY 2009 cohort into one document. The guide aims to address all aspects of the LSAY data including: how to access the data; data restrictions; variable naming conventions; the structure of the data; documentation;…

The Longitudinal Surveys of Australian Youth (LSAY) is a research program that tracks young people as they move from school into further study, work and other destinations. This "User guide" has been developed for users of the LSAY data. The guide endeavours to consolidate existing technical documentation and other relevant information…

The Longitudinal Surveys of Australian Youth (LSAY) is a research program that tracks young people as they move from school into further study, work and other destinations. It uses large, nationally representative samples of young people to collect information about education and training, work, and social development. It includes surveys…

When a simple random sample of size n is employed to establish a classification rule for prediction of a polytomous variable by an independent variable, the best achievable rate of misclassification is higher than the corresponding best achievable rate if the conditional probability distribution is known for the predicted variable given the…

Researchers persist in using stepwise regression in spite of problems with this approach. As noted by B. Thompson (1995), three problems accompany the use of stepwise applications. The first is that computer packages may use incorrect degrees of freedom in their computations, resulting in a greater likelihood of obtaining a spurious statistical…

The serious problems associated with the use of stepwise methods are well documented. Various authors have leveled scathing criticisms against the use of stepwise techniques, yet it is not uncommon to find these methods continually employed in educational and psychological research. The three main problems with stepwise techniques are: (1)…

This report is about the problem of making transition or enrollment rate gains comparable. It is shown that measures based on the proportions themselves, i.e. the difference between proportions, the proportion ratio and the residual gain ratio do not make the gains comparable. Instead a non-linear transformation has to be done. Two such…

The accuracy and variability of 4 cross-validation procedures and 18 formulas were compared concerning their ability to estimate the population multiple correlation and the validity of the sample regression equation in the population. The investigation included two types of regression, multiple and stepwise; three sample sizes, N = 30, 60, 120;…

Examined methodological issues related to the detection and evaluation of behavioral toxicity in infants and children, focusing on the selection of appropriate variables and strategies to control for confounding, sampling strategies and the problem of "overcontrol" for confounding; and the evaluation of dose-response relations and…

The effects of the violation of the assumption of normality coupled with the condition of multicollinearity upon the outcome of testing the hypothesis Beta equals zero in the two-predictor regression equation is investigated. A monte carlo approach was utilized in which three differenct distributions were sampled for two sample sizes over…

Often quantitative data in the social sciences have only ordinal justification. Problems of interpretation can arise when least squares multiple regression (LSMR) is used with ordinal data. Two ordinal alternatives are discussed, dominance-based ordinal multiple regression (DOMR) and proportional odds multiple regression. The Q[superscript 2]…

This report is the methodology report for the National Longitudinal Study of the High School Class of 1972 follow-up in 1986. The fifth follow-up survey of the National Longitudinal Study of the High School Class of 1972 (NLS-72) took place during spring and summer of 1986. A mail questionnaire was sent to a subsample of 14,489 members of the…