PNB 3RM3 Lecture Notes - Lecture 23: Statistical Inference, Dependent And Independent Variables
Document Summary
Single subject research continued: multiple-baseline design (ab design) Now that there is no reversal, you need to provide the treatment at different times instead. Have the introduction of treatment differ across participants. By showing that there is a difference in their behaviour every single time you introduce it, you can make good causal conclusions. By measuring your baseline for different lengths of time, and therefore introducing treatment at different times, you"re making sure there isn"t some extraneous factor that coincides with the switch. Different baselines can be measured in different participants: measuring different dependent variables. Focus is in visual inspection (cf. inferential statistics). Changes in level: does the measure of the overall dependent variable go up or down following treatment: trend. Was there a reversal, slowing down, or increase in the pattern of behaviour following treatment (ideally we want to see a reversal: latency.