Revised PSAT Debuts in October

University Testing

With great fanfare, the Educational Testing Service (ETS) and College Board are introducing a "new" PSAT including a Writing Skills section that will be administered for the first time to high school juniors on October 14 and 18. A review of the new exam's specifications, however, shows that it is neither "new" nor a genuine measure of "writing." But score data from similar exams indicate that even the minor change in format is likely to result in millions more in tuition aid going to females each year.


The revised test was rushed to market as part of the settlement of FairTest's civil rights complaint charging the testmakers with illegally assisting gender discrimination through use of PSAT scores as the sole criterion to select National Merit Scholarship semi-finalists (see Examiner, Fall 1996 and Winter 1994-94). The only significant change is the addition of one 30-minute "Writing Skills" section containing 39 multiple-choice items. This October's PSAT will still include two Math and two Verbal sections derived from the SAT. A National Merit selection score will be calculated by adding scores from the Math, Verbal and Writing Skills sections, all of which are will be reported on a scale of 20 to 80.


Possibly because of the fast development time, two of the three "Writing Skills" sections ("identifying sentence errors" and "improving sentences") were taken unchanged from the Test of Standard Written English (TSWE), an exam that was eliminated when the SAT was revised in 1994. The third "Writing Skills" section, "improving paragraphs" is nothing more than an expanded version of "improving sentences." A test- taker must fill in bubbles in response to every question, making the section a potential measure of copy editing but not actual writing.


Though the PSAT changes are trivial, they appear to be crafted to "substantially address" FairTest's gender discrimination complaint, as required by the National Merit settlement. The last time the TSWE was administered, for example, females averaged more than a point and a half higher than males. Thus, adding a section with nearly identical items to the PSAT will substantially offset the advantage boys have on Math due to the test's gender biases (see Examiner, Winter 1996-97). Moreover, including the "Writing Skills" score in the National Merit selection formula, instead of doubling Verbal where females also score lower than males, further levels the playing field.


Closing the PSAT gender gap will result in more young women becoming National Merit Semifinalists and ultimately winners of the prestigious and valuable scholarships. The precise amount of additional money going to these academically talented females, who already average better grades than males in both high school and college, cannot be determined until test scores are reported, but the sum is certain to be substantial. The National Merit Scholarship Corporation awards more than $25 million each year.


The ease with which the gender gap can be narrowed by trivial changes in test format leads to two obvious questions. Why haven't similar changes been made in similar exams such as the SAT and GRE? And, more fundamentally, why are instruments on which results can be so quickly "adjusted" ever used to make high-stakes decisions?


  Average PSAT Scores
 Class of 1998
   Male Female Gender Gap 
 Verbal  48.6  47.9  0.7
 Math  50.9  47.8  3.1