PSYCH 312 1st Edition Lecture 12Outline of Last Lecture I. Control achieved througha. Participant assignmentb. Experimental designc. Logic of experimentationOutline of Current Lecture I. 3 classes of experimental designsa. Between subjectb. Within subjectc. MixedII. Types of between subject designsa. Completely randomizedb. Multilevel completely randomizedc. Factorial Current Lecture - Outlineo3 general classes of experimental designs-Between-subject-Within-subject-MixedoTypes of between-subject designs-Completely randomized-Multilevel completely randomized -Factorial -General classes of designsoBetween-subjects-Different participants are assigned to different conditions-Different individuals receive different levels of the IV (including the "no IV" control groupoWithin-subjects-The same participants serve in all conditions-Same individuals receive all levels of the IV (including the "no IV" control group)-Participants serve as their own controlsoMixedThese 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.-Examines at least 2 IV's at same time-At least 1 IV is run between-subjects (i.e. differs across groups)-At least 1 IV run within-subjects (all participants experience all levels of this IV)-Types of between-subject designsoCompletely randomizedoMulti-level completely randomizedoFactorial-Completely randomizedoExamines 1 IV with 2 levels (IV; no IV)-Participants randomly assigned to either control or treatment group-1/2 assigned to control, 1/2 to treatment-Different participants in each group (level of IV)-Multi-leveled completely randomizedoExamines 1 IV with 3 or more levels-Random assignment of participants to different groups (1 group for each level of IV)-E.g., if 3 levels: 1/3 participants in each conditionoBetween-subject ANOVA (F ratio)-Determines whether mean score of at least one condition differs from another-H0: no difference across all levels of IV-HA: difference between 2 or more levels of IV-If no significant difference is found for IV-F ratio is non-significant-Fail to reject the null (H0)-If significant difference is found for IV-F ratio is significant, still don’t know which groups differ-Must run independent t-tests comparing means of each group to find which ones differ from each other-Factorial designsoExamines the effects of 2 or more IV's (factors) on the same DV in the same study oEach IV must have at least 2 levels (IV, no IV) but could have moreoNOTE:-In factorial designs, factors (IV) can be run:-Between-subjects-Within-subjects-Mixed (between & within)-Between-subject factorial designsoExample: does effect of music on problem solving performance depend on age?-Factor 1: age-Young, old-Factor 2: music-Rock, classical, no music-Called a "2 x 3 factorial design"-2 x 3 between subject factorial design Rock Classical No musicYoung 20 20 20Old 20 20 20-Between subject factorial designsoAnalysis: Two-way between subject ANOVA-Test for 3 effects-Main effect for factor 1 (age) (f ratio)-Does age alone affect performance?-If significant-no t-test needed (only 2 levels)-Main effect for factor 2 (music) (f ratio)-Does music alone affect performance?-If significant-Follow with independent t-test to find which levels differ (e.g., Tukey HSD)-Music x age interaction (f ratio)-Is any effect of music on performance dependent upon age-If significant-Follow with independent t-test to find which of 6 conditions differ from each
View Full Document