Academic journal article Educational Research Quarterly

The Effects of Gender Bias and Errors in Essay Grading

Academic journal article Educational Research Quarterly

The Effects of Gender Bias and Errors in Essay Grading

Article excerpt

The present study investigated the effects of gender bias and number of errors as unintentional determinants on essay grades. Twenty-two undergraduate education majors were asked to grade four essays written by 6th grade students. Essays were similar in content, but written in either stereotypically male or stereotypically female handwriting. One of each type contained ten errors (six spelling; four grammatical) while the others only contained two (one spelling; one grammatical). Subjects were instructed to grade the essays on content alone. Numerical grades awarded to the essays were recorded A significant main effect for 'perceived sex of the student" was found Essays believed to be written by boys were graded higher than those believed to be written by girls. No other significant differences were found

Research suggests that there are numerous variables, aside from actual content, that may influence the grades that teachers assign to student essays. Among these are handwriting quality (Chase, 1986; Eames & Loewenthal, 1990; Marshall & Powers,1969; Marshall, 1972), the numbers of errors appearing in the essay (Marshall & Powers, 1969), the examiner's expertise on the subject of the essay (Eames & Loewenthal, 1990), the expectations or past performance of the essay writer (Chase,1986; Guttmann & Boudo, 1988), and the writer's family background-(Guttmann & Boudo, 1988), race (Chase, 1986), and sex (Chase, 1986; Goldberg, 1968; Guttmann & Boudo, 1988).'

Two of these variables (the perceived sex of the author of the essay and the number of errors appearing in the essay) were investigated in this study. Because explicitly telling subjects the sex of the ,students before they graded the essays might have hinted at the experimenter's expectations, the sex was suggested by the handwriting. The rationale for this approach is drawn from a study done by Lester, McLaughlin, Cohen, & Dunn (1977). In their first experiment, judges first rated 56 samples of handwriting in terms of masculinity and femininity; in a second experiment, judges rated 20 writing samples as being either male or female; and in a third experiment, 34 handwriting samples were judged in the same manner as the second study. In all three studies, inter judge agreements were high. In fact, in the second study, the inter judge agreement was 70%. From this information, one can conclude that people have similar preconceptions of what male and female handwriting should look like.

Loewenthal (1980), has suggested that although some men might possess stereotypically "feminine" handwriting and some females might possess stereotypically "masculine" handwriting, for the most part, sex can be guessed from the handwriting alone. Loewenthal also goes on to reveal that the basic feature of handwriting that people base these guesses on is circularity: the more circular the writing, the more feminine the writing seems. Thus, the present study may be one of the first studies to investigate sex bias without explicitly informing subjects of the sex of the writer. Moreover, Archer & McCarthy (1988) noted that it has not yet been determined whether sex perceived through handwriting can elicit sex bias.

It has, however, been determined that there is sex bias in the evaluation of written work (Chase, 1986; Goldberg, 1968; Guttmann, 1988). Goldberg (1968) conducted a study that investigated women's prejudice against other women. Six articles were taken from professional literature (each article from a different field) and modified to be the same length. Two sets of the articles were compiled - each set containing the same six articles. In one set, an article bore a female author's name, while in the other set, the same article bore a male author's name. Forty randomly selected college females were asked to read each article in their "booklet" and then answer nine questions that called for critical evaluations of the articles. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.