Classroom Test Analysis

This program analyzes data consisting of one or more test scores. It can weight items differentially (e.g., 2 points for Item #1, 3 points for Item #2, etc.) and accepts positive, negative and decimal fraction scores.

The TOTAL output consists of three parts: A summary of statistics, a test frequency distribution, and a listing of students' scores. Any or all three parts of the output may be printed. The first part consists of a summary of the following test statistics: The number of items on the test, the mean, median, standard deviation, Kuder-Richardson reliability 21 (KR-21), standard error of measurement, the possible and obtained low and high scores, and the number of scores. Only those scores that fall within the range specified by the user are included in the analysis, so that the user has the option of disregarding certain scores. Blank and invalid scores (those falling outside the specified range) are counted but are omitted from the analysis. (See the ITEM ANALYSIS program description for definitions of the statistical terms mentioned above.)

The second part of the TOTAL output displays a test frequency distribution. The raw scores are ordered from high to low, with the corresponding standard score, percentile rank, percentage of people in the total group tested who received the given score, the frequency, and the cumulative frequency. The accompanying histogram portrays the frequency of the scores at each score value. (See the ITEM ANALYSIS program for a description of the statistical terms mentioned above.)

The first two parts of the TOTAL TEST analysis can be processed overall, i.e., for the entire class or group of data. It can also be processed by groups, such as class sections. The summary of test statistics and the frequency distribution can be processed for one to ten test scores per TOTAL program. Raw and/or standard scores can be summed, optionally applying weights to each score; a new summary of test statistics and a frequency distribution will be produced for the summed scores. Since different examinations usually have unequal standard deviations and different numbers of items on each examination, raw scores cannot be compared directly across examinations. When raw scores are summed to determine final score distributions and grades, the examinations with the larger standard deviations will always carry the most weight in determining the final score distribution. Since standard scores are comparable across examinations, it is strongly recommended to sum standard scores instead of raw scores or percent scores.

The third part of the TOTAL output is a listing of the students, sorted alphabetically or by student number.The raw score, standard score, and percentile rank for each test, and optionally for the summed raw and/or standard scores, is presented for each student. If grades are based on an examination score distribution, the summed raw score distribution, or the summed standard score distribution, or determined by assigning a certain proportion of the class different letter grades, then the assigned grades can print out in the listing of individual scores. However, this method of assigning grades is not recommended. Refer to Ebel & Frisbie (1986) or consult the staff of Measurement and Evaluation if you are interested in learning about valid grading practices.

References

• Ebel, R.L., & Frisbee, D.A. (1986). Essentials of educational measurement (4th ed.). Englewood Cliffs, NJ: Prentice-Hall, Inc.
• Gronlund, N.E., & Linn, R.L. (1990). Measurement and evaluation in teaching (6th ed.). NY: MacMillan.