PSYCH 312 1st Edition Lecture 11 Outline of Last Lecture I. Confound vs. extraneous question treeOutline of Current Lecture I. Control achieved througha. Participant assignment b. Experimental designc. Logic of experimentationCurrent Lecture -ControloDesign details used to ensure that results are due to IV and not confounds-Control strategies-Participant assignment-Experimental design details-Logic of the experimental design-Control: participant assignmentoAccomplished by random assignment to conditions-Population-Sample (random selection)-Random assignment (designed to make groups similar at the outset of the study--increases internal validity -Control-TreatmentoDoes not remove differences among participants-Differences wont differentially affect conditionsoSeeks to equate them across conditionsoIf we know the relevant dimensions that may influence your results (outside IV), then tryto balance factors across group-Mental rotaion task-Gender effect-Time estimation study-Time participant arrives to study might be factor--try to balance across levels of IV-Could control using matching-Control: through experimental designThese 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.oType of design can impose specific control strategies -Good design should-Provide evaluation of effects of IV on DV-Help rule out confoundsoPretest-posttest control-group design-Randomly assign participants to groups -Give pretest to each group before treatment -Compare for similarity -Give posttest to each group after treatment-Compare to evaluate treatment effect-Pretest-posttest designoHA: new reading program will increase reading comprehension in 2nd grade children-Randomly assign 52 children to control/experimental groups-Experimental (26 children)-reading pretest-TREATMENT(IV)-reading posttest-Control (26 children)-reading pretest-NO TREATMENT (no IV)-reading posttestoBenefits of this design-Allows to determine if 2 groups equal on DV before treatment -Possible that pretest may influence outcome of experiment-If this is concern, use Solomon four group design-Solomon four-group designoArrangement checks for possibility that pretest influences results of experiment-Group 1 Pretest Treatment (IV) PosttestGroup 2 Pretest No treatment (no IV) PosttestGroup 3- Treatment PosttestGroup 4- No treatment Posttest-To determine whether IV affected the DV (treatment effect)-Compare groups 1 & 2 posttest score-Treatment effect=different-No treatment effect=similar-To determine whether pretest confounded any effect of IV-Compare grp 1 & 3 posttest score-No effect=similar-Effect=different-To determine whether pretest alone affected DV-Compare grp 2 & 4 posttest score-Effect=different-No effect=similar-Control: through the logic of experimental designoFollowing pattern in solomon 4 grp design provides greatest confidence in concluding that IV had effect on DV:-Grp 1 & 2 different posttest -Grp 1 & 3 similar posttest-Grp 2 & 4 similar