GGW’s “Did It Work?”

This exciting feature allows you, in the comments section, to guess what happened in a randomized trial.  One per day all week.  Get it while it lasts.

If you want to participate, write if you think the intervention “worked.”  That is, measurably raised performance in some way, or not.  Other thoughts welcome.  Wonk bonus: include effect size in SDs!

I’ll wander back a day later and write the result.  No Googling!

Did It Work: A recent randomized experiment had 8 year olds get an hour of reading tutorial a week.  Tutors were volunteers from “business.”

(But not from Bain Capital.  Nor Solyndra).

-Guestblogger Mike Goldstein

10 Replies to “GGW’s “Did It Work?””

  1. I really like the idea of teachers having control over their own school, even if that means two teachers and twenty students in a large, private home. When this happens, it will signify full professional status for teachers.

    Whether or not this happens depends on the economy. If it continues as it is, districts will have no problem hiring teachers and paying and treating them poorly. However, if the economy improves and the predicted teacher shortage materializes, I believe teachers will be in a position to insist on full professional status. I hope I live to see this.

  2. For a playground legend with more hops than a 49 year-old HOFer, this is an easy slam dunk: effect size =0.01 & not statistically significant. You’re fattening us up though.

  3. Since you didn’t talk about any protocol for the tutoring, I have multiple predictions, based on three different potential designs.

    1) If there was no guidance to the tutors, ES = 0.

    2) If the guidance was very general (e.g., “read with the child and ask questions”), ES = 0.03-0.04, not statistically significant.

    3) If the tutors were asked to follow a specific protocol with prior research (e.g., where the previous research had paid staff such as graduate research assistants), the ES will be a little lower than in the more rigorous research.

  4. My guess: if the volunteers were given no training, no differential effect. If they received substantial training along the lines of #3 in Sherman’s post, I predict a substantial ES.

  5. (But not from Bain Capital. Nor Solyndra)

    Not funny and insulting.

    As for the “intervention,” not working for me. I choose the red pill.

  6. How many weeks did they receive tutoring?
    If just one week, no significant effect.

  7. Per questions – I didn’t see detailed info on the training, and the total dosage seems to be 12.5 hours.

    This one comes from Ireland. Probably something for everyone in result.

    1. Change in decoding, fluency, and reading rate ranged from +.14 to +.22

    2. Change in comprehension and accuracy not significant, though.

    Of interest –

    3. Scholars who wrote paper lead with: “This follow-up trial has found clear evidence that the refined Time to Read program is effective.”

    http://www.eric.ed.gov/PDFS/ED519114.pdf

    4. Meanwhile, New America Foundation blog on same study has this headline: “New Study Shows Volunteer Reading Tutors Get Limited Results”

    Eye of beholder? I lean more towards 3 than 4.

Leave a Reply

Your email address will not be published.