ATTENTION FACULTY:
If you collect student feedback on your course, please think hard about what you ask and how you ask it.
Otherwise, you’re wasting everyone’s time… or worse.
Some suggestions:
(a🧵)
#MedEd
If you collect student feedback on your course, please think hard about what you ask and how you ask it.
Otherwise, you’re wasting everyone’s time… or worse.
Some suggestions:
(a🧵)
#MedEd
ONLY SURVEY STUDENTS WHO PARTICIPATED.
These days, many students don’t go to class or use institutional resources. Like, at all.
Requiring these students to complete your course evaluation will only generate meaningless feedback for you - and waste their time. Let them opt out.
These days, many students don’t go to class or use institutional resources. Like, at all.
Requiring these students to complete your course evaluation will only generate meaningless feedback for you - and waste their time. Let them opt out.
BE SURE STUDENTS CAN ANSWER THE QUESTIONS YOU’RE ASKING.
Students are well-positioned to comment upon, say, the instructor’s enthusiasm.
But they don’t know whether the concepts that were taught were the ones that *should* be taught.
You need to evaluate that another way.
Students are well-positioned to comment upon, say, the instructor’s enthusiasm.
But they don’t know whether the concepts that were taught were the ones that *should* be taught.
You need to evaluate that another way.
MAKE SURE A SURVEY IS THE BEST WAY TO COLLECT DATA.
Surveys are great for evaluating learner sentiment. Just don’t measure quantities with surveys that could be measured directly.
(e.g., instead of asking whether students accessed resources, just count the page clicks)
Surveys are great for evaluating learner sentiment. Just don’t measure quantities with surveys that could be measured directly.
(e.g., instead of asking whether students accessed resources, just count the page clicks)
KNOW WHAT YOU’LL DO WITH THE RESPONSES.
Think of the survey item as a diagnostic test. You should have a distinct idea about what you’ll do if the feedback says one thing or another.
And if you aren’t gonna change anything, regardless of the sentiment expressed - don’t ask.
Think of the survey item as a diagnostic test. You should have a distinct idea about what you’ll do if the feedback says one thing or another.
And if you aren’t gonna change anything, regardless of the sentiment expressed - don’t ask.
*** - I can’t stress this last point enough. If you ask students, course after course, what they think should be done to improve a course, but in reality, you have no intention of doing what they’re gonna suggest - it breeds mistrust and discontent to keep asking them.
MAKE SURE THE SURVEY ITEMS ARE CLEAN.
If you’re using loaded words, confusing stems, or double-barreled items, you won’t accurately measure your learners’ sentiment.
If you’re using loaded words, confusing stems, or double-barreled items, you won’t accurately measure your learners’ sentiment.
RESIST THE TEMPTATION TO “OBJECTIVELY” COMPARE COURSES.
Unless you have control over all the confounding variables, it’s likely not meaningful to compare Course A’s Likert score of 3.72 to Course B’s Likert of 3.51.
(Heck, those figures may not even be statistically different).
Unless you have control over all the confounding variables, it’s likely not meaningful to compare Course A’s Likert score of 3.72 to Course B’s Likert of 3.51.
(Heck, those figures may not even be statistically different).
Same thing goes for longitudinal comparisons of the same course.
Maybe this year’s class gave you a higher 5-point Likert score than last year’s… but there are probably many differences between those classes beyond the instruction they received in your course.
Maybe this year’s class gave you a higher 5-point Likert score than last year’s… but there are probably many differences between those classes beyond the instruction they received in your course.
ANCHOR LIKERT SCALES WITH REALITY.
A 5-point scale is much more meaningful everyone’s on the same page about what the numbers mean. Anchor the numbers with a description of what a rating at that number *should* look like in real terms.
A 5-point scale is much more meaningful everyone’s on the same page about what the numbers mean. Anchor the numbers with a description of what a rating at that number *should* look like in real terms.
MAKE THE SURVEY AS SHORT AS POSSIBLE.
Then make it shorter.
Ask yourself how much time YOU would want to spend on a survey with no reward.
(Plus, you’d probably rather have meaningful responses to 5 questions than rushed responses to 25.)
Then make it shorter.
Ask yourself how much time YOU would want to spend on a survey with no reward.
(Plus, you’d probably rather have meaningful responses to 5 questions than rushed responses to 25.)
• • •
Missing some Tweet in this thread? You can try to
force a refresh
