PNB 3RM3 Lecture Notes - Lecture 33: Null Result
Document Summary
Confounds can both produce results and obscure them. Use block randomization if you think randomization would produce error. Three design errors can produce null results by creating low between- groups variability from: Levels of independent levels were not different enough to create conditions different enough to produce effect. Use of a measure that produces a floor/ceiling effect. People are either puddled up at the top or bottom of our measure with no variability between groups. You need to use a measure that can detect differences between groups . Make sure independent variable will produce effects that will result in variability between groups. Make sure effect is due to manipulation and not something else. Make them different from each other and make more margins. Can obscure any between group differences you would have found between measures. Because too much overlap on the dv between-groups.