Maintains that Franklin Wilson's 1985 finding that white flight resulting from school desegregation is short-term and unaffected by the nature of the plan is incorrect. Presents new data showing that the effects are long-term and depend mainly upon whether the plan is voluntarily implemented or enforced by the government. (GEA)

Based on data from 411 Filipino secondary school students, the hypothesis was tested that press for achievement from parents, peers, and teachers mediates the relationship between school grades from one year to the next. No support was found for this contention after linear and curvilinear regression models were examined. (Author/LMO)

Critiques the use of regression procedures in which researchers use a number of independent variables to explain reading comprehension. Discusses the problem of simultaneous equation bias and provides examples of how this problem may give researchers a distorted picture of the comprehension process. Offers an approach for dealing with the problem.…

Reports the results of a study which examined the relationship of both student and course characteristics to student rankings of instructors. The 1984 data (n=1,333) showed that 67% of the variation in students' evaluations was explained by instructor-specific variables. Presents detailed regression analysis tables and includes copies of the…

Examines determinants of tuition at American private colleges using 1982-83 data. Hedonic regression analysis involving over 20 quality measures shows that private college tuition responds in economically sensible ways to changes in public tuition and in quantity and quality of the faculties, facilities, and student bodies at private colleges.…

Reports the results of a study of entrance characteristics and degree performance for 341 students studying for the Bachelor of Arts degree at Kingston Polytechnic between 1977 and 1985. Results showed that age, sex, "A" level qualifications, and mathematical background were significant factors in explaining the variation in performance…

The applicability of a new multiple-group factor analysis of dichotomous variables is shown and contrasted with the item response theory approach to item bias analysis. Situations are considered where the same set of test items has been administered to more than one group of examinees. (Author/BS).

Estimated age-adjusted Wechsler Adult Intelligence Scale-Revised intelligence quotient (IQ) scores using linear regression techniques and continuous norming of two samples of psychiatric inpatients (total N=150) in Veteran's Administration hospital. Results support use of estimation procedure in clinical and research settings desiring brief but…

A study of the effectiveness of ridge regression over ordinary least squares regression as applied to both cognitive and noncognitive admissions data is reported. Separate race equations and a general equation were used. The analysis used did not improve on existing regression analyses. (MSE)

Two sophisticated--autoregressive integrated moving average (ARIMA), straight-line regression--and two naive--simple average, monthly average--forecasting techniques were used to forecast monthly circulation totals of 34 public libraries. Comparisons of forecasts and actual totals revealed that ARIMA and monthly average methods had smallest mean…

A statistical model is proposed that describes the determination of an educational outcome variable as a nonlinear function of explanatory variables defined at different levels of a survey data hierarchy, such as students and classes. The theoretical and practical derivation of the model is discussed, and an example is given. (Author/BW)

A model was constructed in which it was assumed that cost and enrollment at public two-year colleges are mutally influential. The model was then estimated by both ordinary least squares regression and two-state least squares regression. The results were found to differ only slightly. (Author/MLW)

Findings of empirical examination of author productivity data to determine conformity to Lotka's law indicate that: most data didn't fit inverse square function; two constants in Lotka's formulation (slope n, constant c) must be derived from observed distribution; inverse square law was special case of inverse exponential relationship. (37…

This study was intended to serve as an example of cross-validating results from student persistence prediction models that employed commonly available pre-college student characteristics. The study investigated whether the accuracy of predicting student persistence would vary because of the use of present-year vs. previous-year parameters on…

Blending means smoothing together subword segments to try to identify a spoken word. Research suggests that beginning readers need to blend to combine the phonemes they "sound out" into a recognizable approximation of a known word. Popular wisdom presumes blending is easiest when the segments of words are whispered and when syllables are…