4 ms·

The ultimate purpose is to improve education until no body ever again spews this bullshit again. The sample size here was 22,000. Is that enough? (the answer,

The ultimate purpose is to improve education until no body ever again spews this bullshit again. The sample size here was 22,000. Is that enough? (the answer, by the way, is "It depends on the effect size you want to measure". I can do a study with n=2 that is pretty convincing. Question: "are these pills actually cyanide?". Study design: "I give one pill each to two healthy undergrads, after they tried to appear smart by saying 'correlation does not prove causation'". Observation: "The ethics board should definitely not have approved this!" Discussion: "Background death rate from non-violent causes for 22 year old males in the US is 0.01096/100000/day. Pill is poisonous with p<0.000000000000012")

ourmandave 10y ago
You're going to have a hell of a time getting participates to fill out follow up surveys for a long term analysis.
- baldeagle 10y ago
  Get them to 'donate their body to science' and you can keep observing them forever, or until the cryo machines stop working.
wjnc 10y ago
Well the sample size should probably be mentioned as 22k / # students per class (~700 classes?), since the students results are purported to be correlated. The statistics is meagre as well. Regressing percentages, counts and dollars all non-normalised is bad form at the least. That said, I like the study. True random design takes guts and effort and gives data to be useful for years. We're still discussing RAND... On an important subject as well. Lot more to be found [1]. Thing that wonders me how state vs national differences are corrected for. On the one hand they shouldn't matter, since it is a random design. But the differences with national averages were huge. And far larger than the measured impact of the intervention. [1] https://www.povertyactionlab.org/evaluation/effects-kindergarten-classroom-earnings-united-states https://www.povertyactionlab.org/evaluation/effects-kinderga...

ourmandave 10y ago

You're going to have a hell of a time getting participates to fill out follow up surveys for a long term analysis.

baldeagle 10y ago
Get them to 'donate their body to science' and you can keep observing them forever, or until the cryo machines stop working.

wjnc 10y ago

Well the sample size should probably be mentioned as 22k / # students per class (~700 classes?), since the students results are purported to be correlated. The statistics is meagre as well. Regressing percentages, counts and dollars all non-normalised is bad form at the least. That said, I like the study. True random design takes guts and effort and gives data to be useful for years. We're still discussing RAND... On an important subject as well. Lot more to be found [1]. Thing that wonders me how state vs national differences are corrected for. On the one hand they shouldn't matter, since it is a random design. But the differences with national averages were huge. And far larger than the measured impact of the intervention. [1] https://www.povertyactionlab.org/evaluation/effects-kindergarten-classroom-earnings-united-states https://www.povertyactionlab.org/evaluation/effects-kinderga...