CRCJ 3001 Chapter Notes - Chapter 9: Confounding, Measuring Instrument, Internal Validity
Document Summary
Order effects as a confounding variable: can distort results of a research study by diminishing or exaggerating a real effect, ex. order effects creates what resembles treatment effect but actually is just order effect. 9. 2 dealing w time-related threats & order effects. Switch to a b-s design: b-s design (w separate group for each treatment) best strategy whenever a researcher has reason to expect substantial order effects, completely eliminates any threat of confounding from order effects. Limitations of counterbalancing: counterbalancing & variance, distorting treatment means. Advantages of w-s designs: requires relatively few participants compared to b-s designs, ex. Individual diffs consistent across treatments measured & removed from rest of variance in data = reduce negative effects of large variance. 9. 4 applications & statistical analysis of w-s designs. Two-treatment designs: advantages, design easy to conduct & results easy to understand, researcher can maximize diff btwn treatments by selecting 2 clearly diff treatment conditions.