Psych 325 1st edition Lecture 23 Current LectureThursday, April 30, 2015Part II: Factorial designs-Factorial DesignsoStudying multiple IV's at onceoDescribed as "N" X "N"-# of N's = # of IV's-N= # of levels of IV-No limit to "N's"-All combinations of the IV's are studiedoBetter than running multiple studies-More control over external influences-Doesn't require as many people-Types of Effects in Factorial DesignsoMain effect-Effect of one IV on the DV-Effect of the other IV's held constant-Simple, straightforward effectoTwo-way interaction effect-Effect of two IV's on the DV-Effect of one IV depends on the level of the other IVoMain effects tell you nothing about interaction effectsoA significant interaction means main effects aren't whole story-Our Favorite Design: 2x2 Factorialo2 IV's, each with two levelsoMain advantages-Simple and easy to interpret-Efficient, doesn't require too many people-Lots of info, especially on generality-Easiest way to test for interaction effects-Why We Love Factorials: Can Test For Interaction Effects!oHelps you better understand how the variables work-May focus on effect of one IV-Other IV reflects different situations or other variables that affect how it worksoLet's you know the extent that an effect will generalize to other situationsoHelps you further develop questions-Why did interaction occur?-Can test with additional studies-High Order Factorial DesignsThese notes represent a detailed interpretation of the professor’s lecture. GradeBuddy is best used as a supplement to your own notes, not as a substitute.oYou can have as many IV's as you wantoWhat if you have 3 IV's?-3 main effects-3 two-way interactions-1 three-way interactionCombined influence of 3 IV's on DVTwo-way interaction varies depending on levels of a third IV-How Many IV's Are Enough?oFactorials are efficient, but if you have too many variables, you'll run into problems-Requires a large number of participants-Participants may need to do too many things-Certain combinations of IV's may be weirdoWhat should you do?-Don't test an IV just because you can-Start small and establish the phenomenon-Only look for meaningful interactions-Study should only be as complex as necessary-How Many Levels Should Your IV Have?oTwo is good-Can conclude that IV has effect-Establishing phenomenonoWhat if two doesn't work?-May be a curvilinear relationship-Higher or lower may workoAll or nothing vs. high and low?-Depends on the question and variable-High and low are easier to
View Full Document