Analysis of the Revised Student Descriptive Questionnaire: Phase Ii: Predictive Validity of Academic Self‐report

Abstract

An initial study phase dealing with the suitability of the revised (1985) Student Descriptive Questionnaire (SDQ) assessed the accuracy of student self‐report for that instrument and found it to be of sufficient accuracy to fulfill its intended purposes. This second phase of study examines the adequacy of the revised SDQ in terms of the predictive validity of its student academic self‐report information against a criterion of first‐year college achievement (GPA). Findings indicate that the validities are consonant with earlier study results using the original version of the SDQ, as well as with other similar self‐report instruments used with college applicants. This was evident in overall magnitudes of the validities obtained, the precision of prediction in multiple regression equations, and in differences between validities for school‐reported and student‐reported high school academic achievement (as GPA and class rank). Multiple regression correlations and their associated regression weights were used to contrast validities for school‐ and student‐report prediction models. These results indicated slight but consistent superiority for the school‐reported information. In multiple prediction based on either school‐ or student‐reported information, apparent differences were found: (1) in the relative contributions of the predictor variables (SAT‐V, SAT‐M, and high school academic achievement); (2) in the relative importance of the predictors for subgroups based on race and, to a lesser extent, on sex; and (3) in differing roles for high school GPA and class rank as academic predictor variables. Other predictive validities were explored for SDQ information involving students' academic self‐concept and student‐reported grades in separate high school subject‐matter areas. Differential validities based on groups defined by high school academic curriculum (i.e., math/science vs. non‐math/science) in conjunction with intended choice of college academic major were also examined.

PREDICTIVE VALIDITY OF SAT® I: REASONING TEST FOR TEST‐TAKERS WITH LEARNING DISABILITIES AND EXTENDED TIME ACCOMMODATIONS

“…A similar relationship was found between student-and college-reported HSGPA in Sample 1 (r = .76, p < .001). Other research examined the relationship between HSGPA and FGPA, and showed that correlations between school-reported HSGPA and FGPA were .04 points larger than the correlation between student-reported HSGPA and FGPA (Donlon, 1984;Freeberg, Rock, and Pollack, 1989). Although these relationships are not identical, they are close and for the purpose of this study self-reported HSGPA will be used as a proxy for actual HSGPA in Sample 2.…”

Comparing the Quality of Students' Experiences During Cooperative Learning and Large-Group Instruction

“…Although GPA is not an ideal indicator of achievement, it is the indicator that is used most widely because "it is a relatively well-defined criterion, is widely understood, and is easily and quickly obtainable" (Young, 1993, p. 151). Although self-reported HSPGA is not as accurate as school-reported GPA, Freeberg, Rock, and Pollack (1989) demonstrated that its predictive validity is nearly as predictive of first-year college achievement, therefore serving as a valid indicator of high school achievement when school records are not available. In addition, Peterson (1993) showed that students' prior achievement, measured as self-reported HSPGA, affects students' motivation for small-group learning.…”

PREDICTIONS OF FRESHMAN GRADE‐POINT AVERAGE FROM THE REVISED AND RECENTERED SAT® I: REASONING TEST

“…FGPA was similarly coded from 0 to 4.3, though 4.0 was the top score for many colleges that did not use A+ grades. Previous research suggests that using the self-reported HSGPA from the SDQ results in multiple correlations (combining SAT scores and HSGPA to predict FGPA) that are about .03 to .04 points smaller than multiple correlations that use the actual school-reported HSGPA (Freeberg, Rock, & Pollack, 1989).…”