Educational Testing Service (ETS), working with 21 U.S. law schools, evaluated the predictive validity of the GRE® General Test using a sample of 1,587 current and graduated law students. Results indicated that the GRE is a strong, generalizably valid predictor of first-year law school grades and that it provides useful information even when…

In this research report, we present an empirical argument for the use of a contributory scoring approach for the 2-essay writing assessment of the analytical writing section of the "GRE"® test in which human and machine scores are combined for score creation at the task and section levels. The approach was designed to replace a currently…

Scores on essay-based assessments that are part of standardized admissions tests are typically given relatively little weight in admissions decisions compared to the weight given to scores from multiple-choice assessments. Evidence is presented to suggest that more weight should be given to these assessments. The reliability of the writing scores…

We analyzed a pool of items from an admissions test for differential item functioning (DIF) for groups based on age, socioeconomic status, citizenship, or English language status using Mantel-Haenszel and item response theory. DIF items were systematically examined to identify its possible sources by item type, content, and wording. DIF was…

Data from 787 international undergraduate students at an urban university in the United States were used to demonstrate the importance of separating a sample into meaningful subgroups in order to demonstrate the ability of an English language assessment to predict the first-year grade point average (GPA). For example, when all students were pooled…

Research on examinees' response changes on multiple-choice tests over the past 80 years has yielded some consistent findings, including that most examinees make score gains by changing answers. This study expands the research on response changes by focusing on a high-stakes admissions test--the Verbal Reasoning and Quantitative Reasoning measures…

Essay scores generated by machine and by human raters are generally comparable; that is, they can produce scores with similar means and standard deviations, and machine scores generally correlate as highly with human scores as scores from one human correlate with scores from another human. Although human and machine essay scores are highly related…

This study examined the relationship between scores on the TOEFL Internet-Based Test (TOEFL iBT[R]) and academic performance in higher education, defined here in terms of grade point average (GPA). The academic records for 2594 undergraduate and graduate students were collected from 10 universities in the United States. The data consisted of…

Descriptions of validity results based solely on correlation coefficients or percent of the variance accounted for are not merely difficult to interpret, they are likely to be misinterpreted. Predictors that apparently account for a small percent of the variance may actually be highly important from a practical perspective. This study combined two…

Automated scoring models for the "e-rater"® scoring engine were built and evaluated for the "GRE"® argument and issue-writing tasks. Prompt-specific, generic, and generic with prompt-specific intercept scoring models were built and evaluation statistics such as weighted kappas, Pearson correlations, standardized difference in…

In order to estimate the likely effects on item difficulty when a calculator becomes available on the quantitative section of the Graduate Record Examinations® (GRE®-Q), 168 items (in six 28-item forms) were administered either with or without access to an on-screen four-function calculator. The forms were administered as a special research…

Two methods of showing the ability of high school grades (high school grade point averages) and SAT scores to predict cumulative grades in different types of college courses were evaluated in a sample of 26 colleges. Each college contributed data from three cohorts of entering freshmen, and each cohort was followed for at least four years.…

The impact of allowing more time for each question on the SAT I: Reasoning Test scores was estimated by embedding sections with a reduced number of questions into the standard 30-minute equating section of two national test administrations. Thus, for example, questions were deleted from a verbal section that contained 35 questions to produce forms…

Descriptions of validity results for the GRE® General Test based solely on correlation coefficients or percentage of the variance accounted for are not merely difficult to interpret, they are likely to be misinterpreted. Predictors that apparently account for a small percentage of the variance may actually be highly important from a practical…

Using data from a sample of 10 colleges at which most students had taken both the SAT I: Reasoning Test and SAT II: Subject Tests researchers simulated the effects of making selection decisions using SAT II scores in place of SAT I scores. Students in each college were treated as forming the applicant pool for a more select college, and the top…