SPHR 2102 Lecture Notes - Lecture 8: Interrupted Time Series, Measuring Instrument, Internal Validity
Document Summary
Effect of experimental manipulation on single research participant. Measured from baseline period to treatment periods. Confidence of treatment effect is direct function of stability of baseline measure. Same issue here constant variability or stability followed by variability suggest lack of experimental control. Ideally, highly stable responding (at different levels) between. Greater overlaps in response measures across conditions decreases conditions. Overlap between scores of adjacent conditions confidence of experimental effect. Number of data points in each condition. Confidence re: stability is function of number. Delay in trend changes decreases confidence in treatment effect. The use of praise as a treatment to measure the improvement of a child"s school performance. Figure 11. 1 - hypothetical data from aba reversal design. Introduce manipulation at different points of time. Determine if manipulation is the cause of change. Procedure used with single subject can be replicated with others. Traditional single-case research presents results from each subject individually. Figure 11. 2 - hypothetical data from multiple baseline.