Sample size too large ?
October 17, 2006 at 5:36 am #44922
Under what conditions should sample sizes be considered too large ?
Other than the practical considerations like time, cost, etc, at what point do too many samples result in an overly narrow estimate of the confidence interval – eg a dataset of 200,00-300,000 samples showed a correlation between an automated measurement vs phase of the moon, as the ANOVA had narrowed the CIs down to a fraction of a mm.
I have heard anecdotaly that this may occur with sample sizes greater than 300-500, but havent found any reference to this.
The forum ‘General’ is closed to new topics and replies.