c.Longitudinal designs compare the same group of subjects over time. It is “long-term” within subjects researchi.Benefits: it controls for individual differences, which increases power; also controls for generational effects, which can be counfoundsii.Drawbacks: very long timed.Cross-sectional designs compare different subjects at different stages, and each person is studied only at one point in time. This is a between subjects designi.Benefits: less timeii.Drawbacks: need more subjects, less powerful than longitudinale.Sequential designsare combined longitudinal and cross-sectional where each subject is followed for the same amount of time, but it’s shorter than longitudinal. For example, one cohort is followed from ages 1-2, another from ages 3-44.What are the threats to internal validity when using these designs (pretest sensitization, history, maturation, measurement, regression to the mean)?a.Selection effects and attrition differences between your subjects in a between subjects designb.History – any event that occurs between the pre and posttest (same time of manipulation)c.Maturation – time changes in subjects. Could be age or fatigue. Not applicable in longitudinal studies because you’re looking at maturation changesd.Measurement – the instrument (test, measure, judge) changes over timee.Regression to the mean – extreme scores come closer to the mean in later testsf.Testing sensitization – can happen in any study with a pretest or repeated measures5.What is a single case study and why do we use them? Examples?a.A single case design examines the effect of treatment on a single subject. These studies have a small number of subjects. The treatment and baseline conditions are compared, and the studies are best if there is a stable baselineb.We often use these studies in applied medical treatment, therapy and counseling, behavior modification, and education6.What is a reversal design? a.In a reversal design(also known as an ABA design), there is some sort of baseline, then treatment is introduced, then performance is measured, then the baseline is re-introducedi.Improved by adding more rounds of baseline/treatment; ABABAB > ABABii.Usually is the most ethical to end on a treatment7.What are multiple baselines (Ss, behaviors)?a.In studies with multiple baselines, you use multiple subjects (small N) and observe thechange from treatment to baseline with no reversalb.Allows the experimenter to examine multiple behaviorsc.These tests are good if going back to the baseline is unethical or impossibleCorrelation1.Know the measures of central tendency and variability.
a.A measure of central tendencyis one number that represents what the whole sample looks likei.Mean– average of the scoresii.Median– the middle score when all the scores are arranged from largest to smallest; When you have skewed data, sometimes the median is a more accurate representation of the sample than the mean isiii.Mode– most often occurring score. Use with nominal scalesiv.