Twitter | Search | |
C. Kirabo Jackson
I often get this question: "High school students have many teachers. As such, how can one identify individual teacher impacts on outcomes like absences?" I created this GIF to try to explain it. Once you think about it in this way, it's pretty simple :)
Reply Retweet Like More
Dapel Nov 15
Replying to @KiraboJackson
Hi , not sure you received my e-mail
Reply Retweet Like
Ewout ter Haar Nov 15
Replying to @KiraboJackson
In this example, the fact that she has a factor 4 bigger impact in team B is evidence that contextual effects maybe much larger than individual impact?
Reply Retweet Like
Sarah Cohodes Nov 15
Replying to @KiraboJackson
This is awesome!
Reply Retweet Like
Steven Glazerman Nov 15
Replying to @KiraboJackson
Love this. I sometimes talk about "the Scottie Pippen problem" when describing estimating the impacts of teachers who have great peers. Really dating myself but it worked well in Chicago in the 90s when I started doing this.
Reply Retweet Like
C. Kirabo Jackson Nov 15
Replying to @EduGlaze
Yes, that does date you :)
Reply Retweet Like
Markus Stoor Nov 15
How do we know in the real world? (With lots of data we might use this for research on teacher characteristics but the data will be lacking for individual evaluation in most real world examples. Or?)
Reply Retweet Like
C. Kirabo Jackson Nov 15
It can certainly be used to get an estimate for individual teachers in the real world. Whether one would want to use if for evaluation is a different question....but this is true for regular teacher value-added (or any proposed measure of teacher quality).
Reply Retweet Like
Maria Prada Nov 15
Replying to @KiraboJackson
This is great! Thanks
Reply Retweet Like
Joel Chan, owning it while I'm honing it Nov 15
Replying to @KiraboJackson
Nice! If you're talking to someone who is into sports, you might say it's the same as 'wins above replacement'
Reply Retweet Like
C. Kirabo Jackson Nov 15
Replying to @JoelChan86
Exactly!
Reply Retweet Like
John Faig Nov 15
Replying to @KiraboJackson
I use an analogy that compares a school to a restaurant. It is a team effort and difficult to attribute precise contributions to the overall experience.
Reply Retweet Like
Martin Nov 15
Isnโ€™t this quite simplified though? Iโ€™m guessing such a neat pattern will rarely appear, and more importantly: wouldnโ€™t the particularities of each students situation, the broader social environment and coincidences create a lot of statistical noise?
Reply Retweet Like
C. Kirabo Jackson Nov 15
Replying to @tjaulow @salonium
Yes, this is simplified precisely to make it easy to understand. That was my rationale behind making the gif in the first place. The question of whether there is noise is separate from whether one can obtain an estimate with some real usable signal.
Reply Retweet Like
Gabs, Preschool Teacher ๐Ÿงธ Nov 15
Replying to @KiraboJackson
๐Ÿ‘๐Ÿ‘๐Ÿ‘ Love it. Great job
Reply Retweet Like
Martin Nov 15
Ah, I see. So itโ€™s quite theoretical. A bit like microeconomics (imagine country A producing apples and country B producing oranges...)?
Reply Retweet Like
Robert Nathenson Nov 15
Replying to @KiraboJackson
This is great!
Reply Retweet Like
Barbara Biasi Nov 15
Replying to @KiraboJackson
I wish I could use this as an exhibit in my old JMP! (I went for boring math instead)
Reply Retweet Like
Andrew Barber ๐ŸŒน๐ŸŒ Nov 15
Replying to @KiraboJackson
For which states can you link students and teachers? I hear itโ€™s quite hard to do and unions fight it, but I have some research that could really benefit from a value-added analysis.
Reply Retweet Like
Scott Ogawa Nov 15
Replying to @KiraboJackson
Conclusion: The great teacher of all time is Robert Horry.
Reply Retweet Like