13.07.2015 Views

Contents

Contents

Contents

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

CHAPTER 8: Complex Designs 257STRETCHING EXERCISE IIn this exercise you are asked to examine Tables 8.1,8.2, and 8.3 to answer the following questions.1 (a) In Table 8.1, what are the means for the maineffect of the suspect-status independentvariable?(b) How does the main effect of the suspect-statusvariable compare to the main effect of theinterrogator-expectation variable forthese data?(c) Is an interaction effect likely present in thesedata?2 (a) In Table 8.2, what are the means for themain effect of the interrogator-expectationindependent variable?(b) How does the main effect of the interrogatorexpectationvariable compare to the main effectof the suspect-status variable for these data?(c) Is an interaction effect likely present in thesedata?3 (a) In Table 8.3, what are the means for themain effect of the interrogator-expectationindependent variable?(b) What are the means for the main effect of thesuspect-status independent variable?(c) Kassin et al. (2003) observed these main effectsto be statistically significant. Using the meansyou computed, describe the main effects of theinterrogator-expectation and suspect-statusvariables in Table 8.3.We are now in a position to describe the conclusions that Kassin et al.(2003) made based on their data analyses of all their data. Using behavioralconfirmation theory, they hypothesized that interrogators’ expectations ofguilt would cause them to conduct an interrogation that would confirm theirbeliefs. Their results supported this hypothesis; overall, interrogators who suspectedguilt conducted more aggressive interrogations. In turn, suspects in theguilty-expectation condition became more defensive and were perceived asguilty by the neutral observers. That the interrogators in the guilty- expectationcondition were even more aggressive when trying to obtain a confession forsuspects who were actually innocent demonstrates the power of their expectationsof guilt and the power of the behavioral confirmation process. In thecriminal justice context, police interrogations that are based on a preexistingbias of the suspect’s guilt can trigger a biased chain of events that may lead totragic conclusions, including false confessions by innocent people.Describing Interaction Effects• Evidence for interaction effects can be identified using descriptivestatistics presented in graphs (e.g., nonparallel lines) or tables (subtractionmethod).• The presence of an interaction effect is confirmed using inferential statistics.How you choose to describe the results of an interaction effect depends onwhich aspect of the interaction effect you want to emphasize. For example,Kassin et al. (2003) emphasized the effect of the interrogation-expectation variableon innocent and guilty suspects to test their predictions based on behavioralconfirmation theory. That is, the manipulation of interrogators’ expectations ofa suspect’s guilt or innocence allowed them to test their predictions that interrogatorswould seek to confirm their expectations. By adding the second independentvariable, Kassin et al. accomplished two things. First, the study more

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

Saved successfully!

Ooh no, something went wrong!