What is the survey response scale?
The survey questions are scaled from 1-5, with 5 as the best score. Prior to Michaelmas Term 2018 the scale was the opposite whereby 1 was the best score.
How can I compare my results from previous years to the new scale?
This table (xlsx) shows the translation of scores from the old scale to the new scale.
What does ’% Agree’ mean?
The ‘% Agree’ is the percentage of respondents who either answered ‘Definitely agree’ or ‘Mostly agree’ and excluding those who answered ‘Not applicable’ from the calculation. For example, X% of students on Y course agree with the statement ‘I have received helpful comments on my work’. This measure is flagged as the ‘cumulative frequency (cf)’ in the PDF reports received by teachers.
What does standard deviation mean?
The standard deviation is a measure of the spread of the responses away from the mean (average) score. Consider the following two sets of responses: in case one, six respondents return the scores 5,5,5,3,3,3; in case two, six respondents return the scores 5,4,4,4,4,3 . In both cases the mean score is 4.0, but case two has smaller deviations away from the mean, so has a lower standard deviation.
What are weighted averages?
A weighted average (of scores in three courses for example) is an overall average across the three courses, taking account of the numbers of responses to each unit. Suppose, for example, that course C1 has mean score 4.4 with 170 responses, course C2 has mean score 4.0 with 20 responses and course C3 has mean score 3.6 with 10 responses, the unweighted average of the three scores 4.4, 4.0 and 3.6 (taking no account of the numbers of responses) is 4.0. But the weighted mean is (4.4x170 + 4.0x20 +3.6x10)/(170+20+10) = 4.47. This is closer to the score 4.4 for course C1, because many more students contributed to the score of that course.
What are typical response rates for the survey?
Typical rates are in the region of 40% for the lecture survey and 60% for the class/seminar survey.