PSY 250 Lecture : PSY250ch7continued.doc
Document Summary
Interval around point estimates to indicate actual population value different from point estimate: sampling error is always present, confidence interval, range of numbers inferred from sample with a probability of capturing the population parameter, normally distributed data, approximately 95% of individual measurements lie within two standard deviations of the mean, benefit of confidence intervals is that know the probability of true population parameter, 95% confidence interval, 95% confident that it will include the population parameter. Doesn"t always work out, but the more people you have, the better variables the chances are: matching, measure the extraneous variable and then assign people to groups based on hat measurement, not possible to do it for every possible extraneous variable, take more participants and more time, repeated testing on same people, testing the same person in all of your conditions super matching, repeated testing causes own problems.