In educational effectiveness research, multilevel models are increasingly used because these models take the multilevel structure of the data into account. In this paper, the effect of ignoring one or more levels of variation in hierarchical linear regression is explored, using a model with four hierarchical levels (individual pupil, class group,…

As an extension of B. Thompson's 1998 invited address to the American Educational Research Association, this paper cites two additional common faux pas in research methodology and explores some research issues for the future. These two errors in methodology are the use of univariate analyses in the presence of multiple outcome variables (with the…

This paper acknowledges that research on school dropouts presents problems in the definitions and computation of dropout indices. It considers a variety of methods for computing dropout rates, using dropout rates which are real and which were available at the end of the 1986-1987 school year. Fall enrollment, average daily membership, and total…

This presentation is an informal discussion of a questionnaire context effect which led to an erroneous conclusion in analysis of data from the Monitoring the Future project; an annual survey of high school seniors that has occurred since 1975, and that focuses heavily on their drug use and attitudes about drugs. In 1980, the researchers reported…

The Bean and Metzner Nontraditional Undergraduate Student Attrition Model appeared to be a good model to address community college student attrition because it takes into account the environmental variables that colleges must consider to enhance persistence. A study was conducted to determine how well the Bean and Metzner Model fit with community…

This 4-year longitudinal study of logical reasoning found complex interrelationships among different cognitive processes of children ages 6 to 15. Piaget's stage theory is discussed in the introduction, with a focus on the concrete operational stage in middle childhood. In the study, a representative array of logical concept tasks and short-term…

The advantages that computer adaptive testing offers over linear tests have been well documented. The Computer Adaptive Test (CAT) design is more efficient than the Linear test design as fewer items are needed to estimate an examinee's proficiency to a desired level of precision. In the ideal situation, a CAT will result in examinees answering…