13.07.2015 Views

Contents

Contents

Contents

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

CHAPTER 8: Complex Designs 253Keep in mind that each group formed by the combination of variables representsa random group of participants. The design looks like this:Interrogator ExpectationSuspect Status Guilty InnocentActual guilt 1 2Actual innocence 3 4Kassin et al. (2003) measured several dependent variables so that they coulddetermine if there was converging evidence in support of the behavioral confirmationtheory. For example, they measured dependent variables for the interrogatorsand suspects, and for new, additional participants who listened to thetape-recorded interrogations (much like potential jurors might hear). We willfocus on three dependent variables from their experiment to illustrate main effectsand interactions. Let’s see what they found.Main Effects and Interaction Effects• The overall effect of each independent variable in a complex design iscalled a main effect and represents the differences among the averageperformance for each level of an independent variable collapsed across thelevels of the other independent variable.• An interaction effect between independent variables occurs when the effectof one independent variable differs depending on the levels of the secondindependent variable.Key ConceptIn any complex factorial design it is possible to test predictions regarding theoverall effect of each independent variable in the experiment while ignoring theeffect of the other independent variable(s). The overall effect of an independentvariable in a complex design is called a main effect. We will examine two maineffects Kassin and his colleagues observed in their experiment for two differentdependent variables.Prior to their interrogation of the suspect, student interrogators were giveninformation about interrogation techniques, including a list of possible questionsthey could ask about the theft. Twelve questions were written as pairs(but presented randomly in the list). One question of the pair was written insuch a way that the suspect’s guilt was presumed (e.g., “How did you find thekey that was hidden behind the VCR?”) and the second question in the pairwas written so as not to presume guilt (e.g., “Do you know anything about thekey that was hidden behind the VCR?”). Student interrogators were asked toselect six questions they might later want to ask. Thus, students could selectfrom 0 to 6 questions that presumed guilt. Based on the behavioral confirmationtheory, Kassin et al. predicted that interrogators in the guilty-expectation conditionwould select more guilt-presumptive questions than would interrogators

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!