regression to the mean

[bgaynes]

We have done a survy evaluating quality of vision following cataract surgery. We used an instrument previously validated by the NIH on normal people without cataracts. The scores range from 0 (unhappy) to 100 (very happy) in categories such as vision comfort, clarity etc. Interestingly, the means for each category from our study are very close to the means from the large population survey done by the NIH. Could we be seeing a "regression to the mean" effect with our sample?

[Dason]

I'm not exactly sure "regression to the mean" applies in this situation. It just sounds like what you're describing is that the group of post cataract surgery people tend to on average have similar vision scores compared to normal people.

For more on what "regression to the mean" actually describes you could read the wikipedia article.

[bgaynes]

I'm not exactly sure "regression to the mean" applies in this situation. It just sounds like what you're describing is that the group of post cataract surgery people tend to on average have similar vision scores compared to normal people.

For more on what "regression to the mean" actually describes you could read the wikipedia article.

OK, I will take a look.

[Link]

Just for kicks...would you mind defining your definition of "regression to the mean" (or looking it up) for us?

[interestednew]

We have done a survy evaluating quality of vision following cataract surgery. We used an instrument previously validated by the NIH on normal people without cataracts. The scores range from 0 (unhappy) to 100 (very happy) in categories such as vision comfort, clarity etc. Interestingly, the means for each category from our study are very close to the means from the large population survey done by the NIH. Could we be seeing a "regression to the mean" effect with our sample?

I dont know anything about your sample so making any kind of inference would be impossible without that. However might this simply suggest that on average people after cataract surgery have pretty normal vision? I would also look carefully at the instrument. Ideally you would want the subjects in your sample to use the instrument before and after the surgery.