May 2, 2010

Mostly-male book images may reduce girls’ science scores

Part of the rea­son boys tend to out­score girls in sci­ence clas­ses may be that most text­books show pre­dom­i­nantly male sci­en­tists’ im­ages, a small ex­plor­a­to­ry study has found.

The stu­dy, on 81 young high-school stu­dents, saw the “gen­der gap” ap­par­ently re­versed when youths were tested based on a text con­tain­ing only female sci­ent­ist im­ages, in­ves­ti­ga­tors said. The gap re­turned in its usu­al form when ma­le-only im­ages were used—and van­ished when the pho­tos showed equal num­bers of men and wom­en sci­en­tists, re­search­ers said.

Part of the rea­son boys tend to out­score girls in sci­ence clas­ses may be that most text­books show pre­dom­i­nantly male sci­en­tists’ im­ages, a small ex­plor­a­to­ry study has found. (Image courtesy Vir­gi­nia Dept. of Ed.)

The in­ves­ti­ga­tors cau­tioned, based on the small sam­ple size and oth­er fac­tors, that it’s un­real­is­tic to ex­pect it would be so easy to erase the gen­der gap in real life.

None­the­less, the find­ings hint that “pro­vid­ing stu­dents with di­verse role mod­els with­in text­book im­ages” may be an im­por­tant step, the re­search­ers wrote in re­port­ing their re­sults. The stu­dy, by Jes­si­ca J. Good of Rut­gers Uni­vers­ity in New Jer­sey and col­leagues, is pub­lished in the March-Ap­ril is­sue of the Jour­nal of So­cial Psy­chol­o­gy.

Oth­er re­search­ers have pro­posed that so­ci­e­ty can wipe out the pe­r­for­mance gap—which has al­ready shrunk­en in re­cent years—by mak­ing stronger ef­forts to give both sexes si­m­i­lar re­sources and op­por­tun­i­ties. A 2004 re­port by the U.S. Cen­ter for Educa­t­ion Sta­tis­tics not­ed that the pre­vi­ous year, sci­ence scores for eighth-grade boys ex­ceeded those for eighth-grade girls in 28 out of 34 coun­tries sur­veyed.

In the study on text­book im­ages, ninth- and tenth-grade stu­dents, 29 male and 52 fema­le, were asked to read a three-page chem­is­try text with one pho­to per page. Stu­dents were ran­domly as­signed one of three ver­sions of the read­ing: one whose pic­tures showed all male sci­en­tists, anoth­er with only female sci­en­tists and one with equal num­bers of sci­en­tists of both sexes. The text it­self was the same in all cases.

The stu­dents, who had no pri­or for­mal chem­is­try train­ing, were next di­rect­ed to take a short test on the read­ing.

Girls did sig­nif­i­cantly bet­ter when us­ing the text with wom­en-only im­ages, the in­ves­ti­ga­tors re­ported. Boys did bet­ter with the men-only im­ages, though the dif­fer­ence here did­n’t reach a sta­tis­ti­cally sig­nif­i­cant lev­el. Over­all, av­er­age scores were high­er for girls than boys among all stu­dents who got the wom­en-only ver­sion.

The com­mon pre­dom­i­nance of ma­le-sci­ent­ist im­ages in text­books is a case of what some read­ers would pe­rceive as “stereo­type threat,” a phe­nom­e­non first de­scribed by re­search­ers at Stan­ford Uni­vers­ity in Cal­i­for­nia in the mid-1990s, ac­cord­ing to Good and col­leagues.

Ster­e­o­type threat oc­curs when a test-taker is pre­sented with, or freshly re­minded of, a ster­e­o­type that re­flects neg­a­tively on his or her abil­i­ties in the sub­ject mat­ter at hand. Stud­ies have found that ster­e­o­type threats push down the test-taker’s score, in the same di­rec­tion the ster­e­o­type would pre­dict.

Thus a pre­dom­i­nance of ma­le-sci­ent­ist im­ages in the ma­jor­ity of sci­ence text­books may re­in­force pop­u­lar no­tions that girls are worse at sci­ence, and then lead to re­sults in line with those ideas, said Good and col­leagues.

Ster­e­o­type threats have been found to af­fect mi­nor­i­ties as well as fema­les. And the new find­ings sug­gest ster­e­o­type threat might work both ways—hurt­ing not only those dis­fa­vored by a com­mon ster­e­o­type, but those fa­vored as well. In par­tic­u­lar, al­though the pop­u­lar ster­e­o­type is that boys are the top pe­rformers in sci­ence, Good’s re­sults hinted that boys’ scores, too, might suf­fer if they saw pic­tures that cut against the flat­ter­ing ster­e­o­type.

A sim­ple so­lu­tion that pre­s­ents it­self, though it re­quires more re­search, would be “mixed-gen­der text­book im­ages,” the re­search­ers wrote. These “may rep­re­sent a sim­ple and cost-ef­fec­tive way to rem­e­dy the neg­a­tive ef­fects of stereo­typic text­book im­ages.”

They cau­tioned that not­with­stand­ing the lat­est re­sults, oth­er stud­ies have found that re­mov­ing ster­e­o­type threats does­n’t com­pletely elim­i­nate pe­r­for­mance gaps among dif­fer­ent groups, though it helps.

How ex­actly ster­e­o­type-threat ef­fects work is un­known, Good and col­leagues said, al­though there is ev­i­dence that they ope­rate largely sub­con­scious­ly. Pos­si­ble rea­sons may in­clude anx­i­e­ty or in­tru­sive thoughts caused by the ster­e­o­type threat, they wrote. Anoth­er ex­plana­t­ion may be that there is a sub­con­scious ten­den­cy to con­form to so­ci­e­tal ex­pecta­t­ions.

“Re­search should in­ves­t­i­gate the in­flu­ence of di­verse role mod­els pre­sented in text­books as a way of im­prov­ing pe­r­for­mance of mul­ti­ple ster­e­o­typed groups, not just wom­en,” the in­ves­ti­ga­tors con­clud­ed. “Although elim­i­nat­ing gen­der bi­as in text­books will most likely not erad­i­cate the gen­der gap in sci­ence in­ter­est and achieve­ment, it will beg­in to chip away at an ev­er crum­bling founda­t­ion.”



Post a Comment