Playing the game with mindset data
It’s been a while since I wrote about the work I did to measure growth mindset in my classes, so I wanted to write an update on some of my findings.
First, Sam Shah recently gave our a similar survey that I borrowed from Bowman in his math classes, and posted his results in summary form on his blog.
Thanks to some wonderful friends at my school who have given this survey to their classes (and their children’s classes), I managed to get a few more insights and a better idea of what a baseline on this instrument should be.
Here are the self reported baseline Growth Mindset Index scores for all of the data I have from my school. Remember that the Growth Mindset Index is a statistical measure I created to measure how closely a respondent’s answers to these questions align with someone who would strongly agree with all growth mindset statements and strongly disagree with all fixed mindset statements.
|Class||N||GMI pre-class score (%)||GMI mid-class score (%)|
|Burk 9th grade physics||35||64||75|
|11th grade biology||49||64||64|
I still find the visual comparison of this data to be one of the most striking, so I’m putting all the data I’ve collected into the image below. Green boxes are aligned with growth mindset, red boxes are aligned with fixed mindset. Pre-class scores are on the left, post class scores on the right (click to enlarge).
First a few important qualifications about the data:
- All of the pre-class measurements were made by asking students to remember what they felt going into this class. For some of the biology classes, these questions were asked very late, so it is likely students may have had some difficulty remembering this information.
- The mid-class measurement was made on at the same time as the pre-class measurement—students were asked to record how they felt previously and now. It’s possible looking at their responses for how they felt previously affect their responses to how they feel now.
- My data was taken 1-2 days after a full class discussion about Carol Dweck’s work on mindset and reading a summary of her findings.
Some additional questions and musings:
- Based on this very limited set of data, GMI scores seem to be very stable across time. I’m wondering if it would be possible to see any differences between young elementary aged populations, and older high school aged populations.
- Since none of the biology classes did any work to discuss growth mindset, this data seems to support the idea that GMI scores are stable as students move through a course. This also seems reasonable, since most students’ self-image of how they learn is unlikely to change in a course.
- This data also supports the idea that instruction and discussion of mindset does make dramatic differences in students self-reported mindsets. The only set of data to report a shift in mindset was my class, where we had previously discussed this topic in class.
- Are students just playing the game? It may very well be that after students read an article about mindset, and participate in a discussion about the topic, that they realize the topic is very important to me, and as a result, they alter their responses to the survey to reflect this, either consciously or subconsciously. My question is—does this matter? There’s a good bit of evidence out there that simply having students shift their language and pretend to adopt a new attitude can have a dramatic affect on attitude. Lots of anecdotal evidence tells me that students aren’t playing a game when it comes to these ideas, but I’d like to explore this question further.