

ODA Home > Institutional Research > FCQ > Statistics for the Section Itself Statistics for the Section Itself We report the number of sheets requested (which usually corresponds to the number of students eligible to rate the course) and the number returned.The RETURN RATE is number of sheets returned  number of sheets requested.For the Boulder campus, the average section return rate (for sections with tenured and tenuretrack instructors) is 77%. The return rate falls below 67% for fewer than 25% of the sections. Be wary if the return rate is under 70%, especially if there's reason to think that students who did not return forms are different from those who did. For each question, we report the number of returned forms with no answer (labelled BL for "blank" on the reports) or with NA (not applicable) answers. This number is usually very small except on questions 6 and 7, on student and issue diversity. Be wary if the question you're looking at was blank or NA on more than 10% of returned forms. This is very rare except on the diversity items. AVERAGE (AVG): The section average is the arithmetic mean of all ratings on one question. Ratings that are blank or NA (not applicable) are not counted in calculating the average. In calculating the average the ratings count A=4, B=3, C=2, D=1, F=0. The average is also shown as a letter grade. The formula for converting numbers to letter grades is A+ = 3.88  4.00 C+ = 2.25  2.50 A = 3.63  3.88 C = 1.75  2.25 A = 3.50  3.63 C = 1.50  1.75 B+ = 3.25  3.50 D+ = 1.25  1.50 B = 2.75  3.25 D = 0.75  1.25 B = 2.50  2.75 D = 0.50  0.75 F = 0.00  0.50.The average can be misleading when there are many responses in the AB range and many in the DF range, but few in the middle (see STANDARD DEVIATION). Because the FCQ scale is restricted to 04, the average is usually very close to the median, another frequentlyused measure of the "typical" response. We do not report the median. STANDARD DEVIATION (STD) = an index of agreement or disagreement among student raters. If all raters agreed exactly (e.g., 100% answered "B" on a particular question), the STD would be zero. The STD is expressed in terms of points on the FCQ 04 rating scale. The higher the STD, the more the raters disagreed with one another. Section STD's for the FCQ generally range between 0.55 and 1.00. As a rule of thumb, check the distribution of responses for the question very carefully if the STD is over 1.10. The raters may be split in their opinions, with many answering A or B, many others answering D or F, and few in between. In such a case the average rating does not reflect a group consensus; in essence there is no "average" or "typical" response. STANDARD ERROR OF THE MEAN (SEM) is often called the "margin of error." It measures how accurately the average reflects the "true average" for the section. There is a 68% chance that the "true average" for the section is between (your average  SEM) and (your average + SEM). This is commonly known as a "confidence interval." The SEM is a function of the section STD for a question and the number of AF responses to that question. The higher the STD and the lower the number of responses, the higher the SEM will be. It's calculated as STD/(square root of N). The section SEM usually ranges between 0.10 and 0.30. Examples:
N of responses A B C D F #AF Avg STD SEM Confidence interval 3 5 10 5 2 25 2.08 1.09 0.22 1.86  2.30 9 4 0 4 8 25 2.08 1.74 0.35 1.73  2.43 30 50 100 50 20 250 2.08 1.09 0.07 2.01  2.15 
Last revision 05/19/16 ODA Home  Institutional Research  Reporting & Analytics  Contact  Legal & Trademarks  Privacy 15 UCB, University of Colorado Boulder, Boulder, CO 803090015, (303)4928631 © Regents of the University of Colorado 