PSY 200 Lecture Notes - Lecture 10: Repeated Measures Design, Time Point, Confidence Interval
Document Summary
Most research compares two or more sets of data. This can be done by assessing the same participants more than once, or by sampling two different groups at one time. Comparing two different groups at one time point is called a between-subjects research design. This test is used when comparing two groups on a single outcome. No(cid:405), (cid:405)e don"t kno(cid:405) any popul1tion p1(cid:364)1mete(cid:364)s, so (cid:405)e 1(cid:364)e comp1(cid:364)ing two. So, it"s 1 g(cid:364)oup of only two levels. Dependent variable: must be interval or ratio. Only if two groups are significantly different or not. Cannot tell us how different they are. Assumptions for independent t-test: independence, one observation does not depend on another within each sample, normality, both populations are normally distributed, homogeneity of variance, new concept!! The two populations from which the samples are selected must have equal variances. Independent-measure t-tests use pooled variance (between the 2 groups) in the analysis.