Health Sciences 3801A/B Lecture Notes - Lecture 2: Confidence Interval, Sample Size Determination, Repeated Measures Design
Document Summary
Topics: two designs (repeat measures vs. between groups, calculating a t ratio, calculating confidence interval, effect size and cohens d, sample size. There is always one uncertainty about the size of he mean difference. Standard error: is a measure of how much the mean differences is likely to vary its calculation depends on the design. T ratio: a t ratio is calculated by diving the mean difference by its standard error, md represents the effects of your condition, se represents the effects of chance, fluctuations from where we have no real explanation. Indicates the number of standard deviations by which groups differ (cid:862) (cid:373)all(cid:863) effect (cid:894)d= 0. 2(cid:895: the effect is not perceptible to the eye of an observer. medium effect (d=0. 5: the effect is perceptible to the eye of a sensitive observe. strong (d= 0. 8 or higher: the effect is perceptible to the eye of a casual observer, should strive for this on the clinical level.