Assessing problem-solving skills on selected questions from the Scholastic Aptitude Test

The types of strategies that students use while solving selected questions from previously administered Scholastic Aptitude Tests (SAT) were examined. The primary aim of this study was to identify various measures of test-taking skills as developed from observations during verbalized problem solving. The secondary purpose was to determine if verbalization had a detrimental impact on students' test performance.

Sixteen high school juniors participated in this study. Students were randomly selected on the basis of sex, race and classified on ability level as measured by scores obtained on the Preliminary Scholastic Aptitude Test (PSAT). Three (3) sets of verbal and quantitative questions were administered--timed and verbalized; untimed and verbalized; and untimed and solved silently. Subjects were individually tested and videotaped.

On the general measure of Good Test-taking, upper level students and females employed more effective test taking behaviors and outperformed their counterparts on the actual test questions. Upper level Blacks outscored all other groups on both the skills measure and the test questions.

As expected, upper level students outperformed lower level students on the Verbal Skills measure and on the verbal test questions. Upper level black students attained the highest average score on Verbal Skills as well as on the verbal problem sets.

On the Math Skills measure, upper level students outscored the lower level students. Although females scored significantly higher, they scored the same as the males on the set of math questions. A closer examination revealed that although the upper level females outscored the males by four to one on the math skills measure, they scored the same as their male counterparts on the math questions.

Verbalization during problem solving for verbal questions was not significant. However, thinking out loud while solving math problems did have a significant impact. Lower level students, especially lower level females, benefitted from verbalizing during problem solving.

The results provide insight to those behaviors that students display while engaging in problem solving and provides a basis on which to further examine other student populations. Although the findings are encouraging, additional research is recommended before generalizations can be made across populations.