High school seniors' performance on a test of selected consumer mathematics problems
The problem of this study was to identify selected consumer mathematics concepts, to develop an instrument for measurement of those concepts, and to determine the level of high school seniors' ability to solve problems using those concepts.
An outline of 40 items was constructed and two problems were written for each outline. Pilot Test I was constructed by using the 80 problems in four 20-problem tests. Pilot Test I, using open-ended questions, was administered to obtain distracters for multiple-choice items used in Pilot Test II. Pilot Test II was conducted to test the plausibility of the distracters. Pilot Test III was given to refine the items and determine the best items for inclusion in the final test.
In the construction of the final test only the best 35 of the original 80 items were used. The test was administered to 502 students in five high schools in Knoxville City and Knox County schools; however, only 442 of the students' answer sheets could be used in the analysis.
The final test was handscored and all answers were transferred to IBM cards. These IBM cards were used in the analysis of the study which was accomplished through the use of an SPSS-6 and SAS program.
The reliability coefficient of the test was .86, which met the criterion that it should not be below .80. The criterion for discrimination was that at least 50 percent of the items have indexes above .40; and 100 percent were to have indexes above .20. Seventy-seven percent of the items were above .40, with the remaining items between .20 to .40. The criterion for difficulty indexes was that at least 50 percent of the items were to fall in the .40 to .70 range, with no items below .20 or above .80. Seventy-four percent of the items fell within the .40 to .70 range, with 89 percent in the .20 to .80 range. Only four items failed to meet the criterion.
Findings of the study revealed:
1. A significant difference was found between mean scores for male and female students in the entire sample of 442 students. Males scored significantly higher on the test than the females.
2. A significant difference was found between mean scores for college-bound and non-college-bound students in the entire sample. College-bound students scored significantly higher on the test than non-college-bound students.
3. High school seniors in this study could not solve consumer mathematics problems with a high degree of competence. Only 44.3 percent of the students correctly answered more than 65 percent of the test items.
Thesis76.L583.pdf
3.96 MB
Adobe PDF
2928fc84d2c32f323693dfc16da55c14