ERIC Number: ED424337
Record Type: Non-Journal
Publication Date: 1997-May
Reference Count: N/A
The ETS Gender Study: How Females and Males Perform in Educational Settings.
Cole, Nancy S.
The Educational Testing Service (ETS) Gender Study is the result of 4 years of work by several researchers using data from more than 400 tests and other measures from more than 1,500 data sets involving millions of students. The study focuses on nationally representative samples that cut across grades (ages), academic subjects, and years in order to control factors that may have introduced confusion and contradictory results into previous studies of gender differences in educational settings. Results indicate that gender differences are not quite as expected. For nationally representative samples of 12th graders, the gender differences are quite small for most subjects, small to medium for a few subjects, and quite symmetrical for females and males. There is not a dominant picture of one gender excelling academically, and in fact, the average performance difference across all subjects is essentially zero. The familiar mathematics and science advantage for males was found to be quite small, significantly smaller than 30 years ago. However, a language advantage for females has remained largely unchanged over that time period. Also, gender differences for component skills of academic disciplines were often different than for the discipline as a whole. Gender differences were shown to change as students grew older and moved to higher grades. Patterns of gender differences in performance are similar to patterns of differences in interests and out-of-school activities, suggesting that a broad constellation of events relates to observed differences. Results show larger gender differences for self-selected groups taking high-stakes tests than for nationally representative samples, reflecting primarily the wider spread of male scores. Results indicate that neither guessing, speededness, nor the multiple-choice format per se accounts for the gender differences. However, results on presently used open-ended questions sometimes reflected no gender effect and sometimes reflected effects in which females' performances exceeded those of males and vice versa. Implications of these findings are discussed. A list of 67 resources for further reading is included. (Contains 8 figures and 19 endnotes.) (SLD)
Publication Type: Reports - Research
Education Level: N/A
Authoring Institution: Educational Testing Service, Princeton, NJ.