Psychology 242 Introduction to Research Psychology 242 Dr

Psychology 242 Introduction to Research Psychology 242, Dr. Mc. Kirnan 1 Complex Research Designs

Psychology 242 Introduction to Research Complex Research Designs: Multiple Independent Variables. ü Click “slide show” to start this presentation as a show. ü Remember: focus & think about each point; do not just passively click. © Dr. David J. Mc. Kirnan, 2014 The University of Illinois Chicago Mc. Kirnan. UIC@gmail. com Do not use or reproduce without permission 2

Psychology 242 Introduction to Research Complex Research Designs: Multiple Independent Variables. 3 4/14/09 Topics: q Testing hypotheses about > 1 independent variable q Factorial Designs: Main effects, Additive Effects, Interactions Examples of complex experiments q The interaction of drug use & attitudes on sex risk among gay men q The interaction of “nature” and “nurture”: Genetics & stress and depression q 3 Independent variables: alcohol and behavioral disinhibition 1121/16

Psychology 242 Introduction to Research Multiple independent variables Testing hypotheses about > 1 q independent variable q Factorial Designs: Main effects, Additive Effects, Interactions Examples of complex experiments q The interaction of drug use & attitudes on sex risk among gay men q The interaction of “nature” and “nurture”: Genetics & stress and depression q 3 Independent variables: alcohol and behavioral disinhibition Psychology 242, Dr. Mc. Kirnan Multiple independent variables 4

Psychology 242 Introduction to Research The complexity of life 5 Our world is far more complex than we can express in an experiment: ü Any effect, however strong, is influenced by the myriad of other variables that comprise the natural world; ü A research study addresses a model of the world: a simplified representation of how the world works; ü Simple experiments necessarily ignore many influences on the phenomenon we are trying to explain. ü As research becomes more complex, we are able to capture more diverse explanations. ü An an example, click the box to see the many variables that influence human obesity…

Psychology 242 Introduction to Research Main effects 6 The examples we have been using with the t-tested a single Main effect: n An experimental design with a Single Independent Variable [IV]. Experimental groups n n n [true experiments] Manipulated by experimenter, to Test an a priori hypothesis, e. g. ; n Placebo v. active drug n Intervention v. distraction group …etc. Or: A Quasi-experiment representing a simple group contrast n Naturally occurring groups [quasi-experiments] Both Predictor and Criterion (outcome) are measured, e. g. ; n n Male v. female School 1 v. school 2, …etc.

Psychology 242 Introduction to Research n 7 Example of a main effect Do stressful life events lead to more depression? E X A M P L E n n Men were sorted into 5 groups, corresponding to # major life stressors they experienced from age 21 to 26. At age 26 men in groups 3 & 4 were significantly more likely to have lifetime major depression episode than groups 0 2 Looking only at stress as an Independent Variable, there is a main effect of stress Number of major stressful events on depression. (Ages 21 to 26) Partial data from Avshalom C. , (2003), Science Magazine. Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research This tests a relatively simple theory: Ø It links one hypothetical construct to one outcome Ø Stress depression Ø Arousal performance Ø Gender sex-role attitudes Ø Assumes the main effect is independent of other key variables. Ø Of course in reality is it rare (impossible? ) for an important outcome to be under the control of only one variable. 8

Psychology 242 Introduction to Research Multiple variables in psychological research Multiple Independent variables allow us to test more complex theories & hypotheses: Ø Link two or more hypothetical constructs to an outcome. § Stress and genetic vulnerability depression § Arousal and gender may affect performance § Drugs and expectations sexual risk Ø Test if an effect depends upon other key variable(s). § One Independent Variable may affect the outcome only at one level of a second IV. Maltreatment affects depression (a main effect), but only for some groups (the interaction). Psychology 242, Dr. Mc. Kirnan Multiple independent variables 9

Psychology 242 Introduction to Research Multiple variables in psychological research Multiple Independent variables allow us to test more complex theories & hypotheses: Ø Alcohol use leads to more Link two or more hypothetical constructs to an outcome. Ø Test if an effect depends upon other key variable(s). emotional arousal (a main effect), but men get stimulated and women get sedated (the interaction). § One Independent Variable may affect the outcome only at one level of a second IV. § The first IV may have a different effect on the outcome at different levels of IV 2. Psychology 242, Dr. Mc. Kirnan Multiple independent variables 10

Psychology 242 Introduction to Research Two uses of multiple Independent Variables A. Including a ‘control’ variable as an I. V. ü E. g. , gender, age, race, etc. ü Test if the I. V. has the same effect within both groups B. Testing hypotheses about 2 or more I. V. s 1. Separate, ‘main effects’ of each I. V. (Do each of these variables significantly affect the outcome? ) 2. ‘Additive’ effects of > 1 I. V. s simultaneously (What is the combined effect of these variables? ) 3. Interaction of 2 or more I. V. s (Does the effect of one I. V. on the outcome depend upon a second variable. . . ? ) Psychology 242, Dr. Mc. Kirnan Multiple independent variables 11

E X A M P L E Psychology 242 Introduction to Research A. Including a ‘control’ variable as an I. V. 1. Block the data by gender, age, race, attitudes, etc. 2. Test if the main Independent Variable has the same effect within both groups. Stress leads to depression. Is this process the same for men and women? Psychology 242, Dr. Mc. Kirnan 12

Psychology 242 Introduction to Research A. Including a ‘control’ variable as an I. V. 13 ü All adolescents tend to become depressed in response to either interpersonal or chronic stress. ü Girls respond much more strongly to specific episodes of E X A M P L E interpersonal stress (e. g. , argument with friend…). ü Boys respond more strongly to chronic stress (“not belonging”). ü Adding gender to our analysis allows us to better understand stress among adolescents. Shih, J. H, Eberhart et al. (2006) Differential Exposure and Reactivity to Interpersonal Stress Predict Sex Differences in Adolescent Depression. Jr. Clin. Child & Adol Psy, 35 (1), p 103115. Here. Click for a general, integrative overview of gender differences in depression.

Psychology 242 Introduction to Research B. Testing hypotheses about 2 or more I. V. s 14 E X A M P L E Examining both trauma and genetic vulnerability allows us to better understand the onset of depression. Psychology 242, Dr. Mc. Kirnan There is a general (main) effect whereby more trauma leads to greater likelihood of adult depression

Psychology 242 Introduction to Research B. Testing hypotheses about 2 or more I. V. s However … the effect of trauma interacts with genetics E X A M P L E Childhood trauma has no effect in people who have no genetic vulnerability. This constitutes a more complex theory of depression… …depression results from the interaction of several variables. Psychology 242, Dr. Mc. Kirnan With increasing genetic vulnerability, increasing trauma increases the likelihood of depression 15

Psychology 242 Introduction to Research Multiple independent variables q Testing hypotheses about > 1 independent variable q Factorial Designs: Main effects, Additive Effects, Interactions Examples of complex experiments q The interaction of drug use & attitudes on sex risk among gay men q The interaction of “nature” and “nurture”: Genetics & stress and depression q 3 Independent variables: alcohol and behavioral disinhibition Psychology 242, Dr. Mc. Kirnan Multiple independent variables 16

Psychology 242 Introduction to Research 17 Factorial Designs We use a Factorial Design to test hypotheses about 2 or more Independent Variables The first IV is “crossed” with the second IV. The Dependent Variable (Depression) is assessed for each combination of the IVs IV #1; Gender IV #2; Type of stressor DV: Depression Female Psychology 242, Dr. Mc. Kirnan Male Interpersonal DV DV Chronic DV DV Multiple independent variables

Psychology 242 Introduction to Research 18 Factorial Designs This allows us to test: ü The Main Effects of each IV; ü The Additive Effect of the two IVs together ü The Interaction of IV #1 by IV #2. IV #1; Gender DV: Depression Female IV #2; Type of stressor Psychology 242, Dr. Mc. Kirnan Male Interpersonal DV DV Chronic DV DV Multiple independent variables

Psychology 242 Introduction to Research 19 Factorial design ü Girls respond much more strongly to specific episodes of interpersonal stress (e. g. , argument with friend…). ü Boys respond more strongly to chronic stress (“not belonging”). ü A factorial design (2 IVs) tells us much more than a main effects (1 IV). Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research Example: Factorial design testing 2 IVs Hypothesis: Coping skills delivered by a peer helps diabetics maintain blood sugar. Independent Variables: Skills training Trainer Values: None (Placebo / Distraction grp) High (experimental group) Nurse (“Standard of care”) Peer (experimental group) NB: This example & the data are completely made up 20

Psychology 242 Introduction to Research n n 21 Example: Factorial design testing 2 IVs The hypothesis rests on the interaction of two variables Adding a second IV of ‘training condition’ tests a more complex theory of skills training. Independent Variable 1 Experimental v. control groups IV #2 Training condition 4 -group factorial Design Skills training Distraction Peer trainer DV = M glucose control Nurse DV = M glucose control Dependent Variable: Glucose control (assessed for every combination of IV 1 and IV 2) Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research 22 Example: Factorial design testing 2 IVs Each “cell” of the design represents both IVs: Independent variable 1 No skills (Distraction) M peer, no skills Nurse trainer M Nurse, no skills Peer trainer Independent variable 2 Skills Training M M peer, skills Nurse, skills Each cell contains data for a specific combination of conditions Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research 23 Example: Factorial design testing 2 IVs This is a 2 x 2 factorial design: Independent variable 1 ü 2 independent variables ü 4 data cells, each with a M Data Table No skills Skills value for the D. V. Training Row Ms (Distraction) Independent variable 2 Peer trainer M peer, no M peer, skills M Peer trainer Nurse trainer M nurse, no M nurse, M Nurse trainer Column Ms M skills no skills M skills The “Marginals” show overall Ms for each I. V. : Skills v. no skills (a. k. a. main effect for skills…) Peer v. nurse trainers (trainer main effect…) Contrasts among individual cells show any interaction effects. Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research Here is a (completely made up) example of a main effect. No skills (Distraction) Skills Row Ms Training Peer 14 60 37 Nurse 12 13 58 59 35 36 Column Ms ü Glucose control is enhanced by skills training ü The amount of change is the same for both training groups. Psychology 242, Dr. Mc. Kirnan 24 Testing Main Effects (hypothetical data) Multiple independent variables

Psychology 242 Introduction to Research Alternate display of two Main Effects 25 (hypothetical data) An alternate display of the same main effect data Skills training helps, by about the same amount no matter who it is delivered by. These data would support a simple Main Effect of skills training Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research 26 Two (additive) Main Effects Here is an example of an Additive Effect. The data show a main effect of trainer (Peers do better than nurses). Data set 2. No skills (Distraction) Skills training Row Ms Peer 40 80 60 Nurse 12 40 26 Column Ms 26 60 43 And of skills Training generally helps training: Getting skills training helps about the same in both groups …but contact with a peer is generally more helpful

Psychology 242 Introduction to Research Additive Main Effects (hypothetical data) “Adding” these effects together shows a very high value for patients who get skills training by a peer… Psychology 242, Dr. Mc. Kirnan Multiple independent variables 27

Psychology 242 Introduction to Research 28 Additive main effects: alternate display Here is an alternate display of the additive effect. Main effect of trainer: peers do better than nurses no matter what the intervention… …and a main effect of skills: skills training helps, whether delivered by a nurse or a peer… Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research 29 Alternate display of two additive main effects Alternate display of additive effect of two variables These two effects “add up”: Skills delivered by a Peer have the best results. Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research Here is an example Data set 3. of an Interaction Effect of trainer by Peer training condition. Nurse trainer Column Ms ü Skills training made a difference ü But only among patients trained by a peer. ü For patients trained by a Nurse, training had little effect Psychology 242, Dr. Mc. Kirnan 30 Interaction Effects (hypothetical data) Multiple independent variables No skills Skills (Distraction) Row Ms training 12 80 46 12 20 16 12 50 31

Psychology 242 Introduction to Research 31 Interaction Effects (hypothetical data) Here is an example of an Interaction Effect of trainer by training condition. Data set 3. No skills Skills (Distraction) ü Since these data Row Ms training Peer trainer 12 80 46 Nurse trainer 12 20 16 Column Ms 12 50 31 show an interaction, our graph shows the 4 individual cell Ms. ü The group that improved was the specific combination of: ☞ Skills ☞ Delivered by a peer Multiple independent variables

Psychology 242 Introduction to Research Alternate Display of interaction effect (hypothetical data) 32 Here is an alternate display of an interaction between two variables There is a large overall effect of training versus distraction, but only for the peer trainer, 90 80 70 Overall M for skills training not the nurse. 60 Distraction 50 40 Overall M for distraction (placebo) 30 20 Skills training 10 0 Nurse trainer Peer trainer Interaction: the effect of the 1 st Independent Variable (training) depends upon the 2 nd IV (peer v. nurse). Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research 33 Click – 1 What is a main effect? a. The simple effect of one I. V. on the D. V. b. One IV combines with another IV to produce an effect on the DV. c. One IV has an effect on the DV only at one level of another IV. d. An IV has a different effect on the DV at different levels of a second IV. Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research 34 Click – 2 What is an additive effect? a. The simple effect of one I. V. on the D. V. b. One IV combines with another IV to produce an effect on the DV. c. One IV has an effect on the DV only at one level of another IV. d. An IV has a different effect on the DV at different levels of a second IV. Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research 35 Click – 3 What is an interaction? a. The simple effect of one I. V. on the D. V. b. One IV combines with another IV to produce an effect on the DV. c. One IV has an effect on the DV only at one level of another IV. d. An IV has a different effect on the DV at different levels of a second IV. Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research Multiple independent variables q Testing hypotheses about > 1 independent variable q Factorial Designs: Main effects, Additive Effects, Interactions Examples of complex experiments q The interaction of drug use & attitudes on sex risk among gay men q The interaction of “nature” and “nurture”: Genetics & stress and depression q 3 Independent variables: alcohol and behavioral disinhibition Psychology 242, Dr. Mc. Kirnan Multiple independent variables 36

Psychology 242 Introduction to Research Example of interaction effect; AIM study Sexual Risk among gay & bisexual men who combine alcohol and drugs with sex. ü People who use drugs during sex are more likely to have unsafe (as well as more) sex. ü What causes that… ü The drugs themselves (“disinhibition”) ü Some characteristics of people who use them? ü These are data from the Awareness Intervention for Men study of interventions for unsafe sex among MSM who use drugs. Psychology 242, Dr. Mc. Kirnan Multiple independent variables 37

Psychology 242 Introduction to Research Example of interaction effect; AIM study, 2 38 Mc. Kirnan, D. J, Vanable, P. , Ostrow, D. , & Hope, B. (2001). Expectancies of sexual “escape” and sexual risk among drug and alcohol-Involved gay and bisexual men. Journal of Substance Abuse, 137 -154. Paper here. Two “main effect” hypotheses: ü Drug use: Bio-behavioral hypothesis: drug use has brain-based disinhibiting effects, can lead to sexual (and other) risks. ü Attitudes: Cognitive “escape” hypothesis: using drugs to avoid thinking about personal problems or risks both more drugs and drugs + risk. Interaction hypothesis: ü Drugs make people more risky, but primarily if they also have “high risk” (“escape”) attitudes. Implications for theory: Interventions should focus on attitudes & motivations as well as simple drug use. Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research Example of interaction effect; AIM study, 3 Sexual Risk… Two Main effects: Independent of drug use, men with higher cognitive escape motives reported more risk. Psychology 242, Dr. Mc. Kirnan Multiple independent variables Independent of motives, higher drug use led to more risk. 39

40 Psychology 242 Introduction to Research With the factorial design, we summarize all 4 means in our table. The effect of Cognitive Escape motivation DV: risk levels Low High drug use Low escape motive 17 High escape motive 33

41 Psychology 242 Introduction to Research With the factorial design, we summarize all 4 means in our table. And the effect of Drug Use. DV: risk levels Low High drug use Low escape motive 17 High escape motive 33 12 28

42 Psychology 242 Introduction to Research With the factorial design, we summarize all 4 means in our table. Filling in the cells of the design shows us the interaction effect. DV: risk levels Low High drug use Low escape motive 12 18 17 High escape motive 15 40 33 12 28 As we can see, the Highest risk is among men with both high drug use and strong cognitive escape motives.

Psychology 242 Introduction to Research 43 Sexual Risk… Overall / Interaction finding: Drug users with strong escape attitudes were very risky. Drug users without escape attitudes were less risky. For men who did not use drugs, attitudes did not affect risk. Drugs & attitudes interact: Drugs lead to risk primarily in the “escape” group. Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research Alternate display of interaction effect Drug users: risk is high, but primarily among those with strong expectancies No drug use: risk stays low for all participants Psychology 242, Dr. Mc. Kirnan Multiple independent variables 44

Psychology 242 Introduction to Research Multiple independent variables q Testing hypotheses about > 1 independent variable q Factorial Designs: Main effects, Additive Effects, Interactions Examples of complex experiments q The interaction of drug use & attitudes on sex risk among gay men The interaction of “nature” q and “nurture”: Genetics & stress and depression q 3 Independent variables: alcohol and behavioral disinhibition Psychology 242, Dr. Mc. Kirnan Multiple independent variables 45

Psychology 242 Introduction to Research Interaction of genetics & stress on depression. Overall hypothesis: Consistent with epigenetics perspective, stress may “switch on” genes that confer vulnerability to depression Independent variables: n n Variations in a gene that controls serotonin production in the brain [a measured variable]. The number of “serious” stressful life events between ages 21 and 26 [also a measured variable] Outcome variables: n Symptom counts n Major depression episode n Suicide attempt n Others’ reports of depression Psychology 242, Dr. Mc. Kirnan Avshalom C. , et al. (2003). Influence of Life Stress on Depression: Moderation by a Polymorphism in the 5 -HTT Gene. SCIENCE, 301 (July 18), 386 -389 [www. sciencemag. org]. [See readings: summary or the actual article. ] 46

Psychology 242 Introduction to Research Interaction of genetics & stress on depression. 4 outcome measures (DVs). 3 levels of genetic disposition. Psychology 242, Dr. Mc. Kirnan 5 levels of stress. 47

Psychology 242 Introduction to Research A main effect of stress on More stress = more depression on all measures depression. Psychology 242, Dr. Mc. Kirnan Interaction of genetics & stress on depression. Multiple independent variables 48

Psychology 242 Introduction to Research An interaction of stress by …but only among genetically vulnerable people genetic vulnerability. Psychology 242, Dr. Mc. Kirnan Interaction of genetics & stress on depression. Multiple independent variables 49

Psychology 242 Introduction to Research 50 Key genotypes and stress Stress leads to depression only for those who are genetically vulnerable. This is illustrated by different stress depression effects for two key genotypes. High vulnerability Psychology 242, Dr. Mc. Kirnan Multiple independent variables Low vulnerability

Psychology 242 Introduction to Research Interaction of genetics & stress on depression; Trauma. 51 Interaction is very strong in an analysis of childhood trauma and depression. People with no genetic vulnerability: childhood trauma has no effect on depression People with increasing genetic vulnerability: More trauma greater likelihood of depression. Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research 52 Multiple independent variables q Testing hypotheses about > 1 independent variable q Factorial Designs: Main effects, Additive Effects, Interactions Examples of complex experiments q The interaction of drug use & attitudes on sex risk among gay men q The interaction of “nature” and “nurture”: Genetics & stress and depression 3 Independent variables: q alcohol and behavioral disinhibition Fillmore, M. T. , & Weafer, J. (2004). Alcohol impairment of behavior in men and women. Addiction, 99 (10), 1237 -1246. Article here. Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research Example: 3 independent variables 53 The study addressed two phenomena: gender differences in: 1. Emotional arousal after consuming alcohol (or a placebo). 2. Disinhibition; loss of the ability to inhibit a response after being primed to do so. Two Core Hypotheses: 1. Men get more aroused by alcohol, women get less aroused. 2. Men are less able to inhibit behavior in response to alcohol than are women. Theories: Social learning: Men are socialized to lessen behavioral control in alcohol-related situations, women socialized to increase caution. Bio-behavioral: Basic inhibitory mechanisms in men are more reactive to alcohol and other drugs than in women. Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research 54 Experimental Design 2 Studies, each with 3 Independent Variables. Both studies provided alcohol v. placebo beverages (IV #1) to men v. women (IV #2). Study 1: Questionnaires addressing: ü Subjective stimulation ü Subjective sedation. (IV #3) Study 2: In a simple reaction time task participants told to: a. Press a button quickly after a “go” stimulus (a green light). b. Do not press with a “no go” stimulus (red light). c. They are “primed” to expect one or the other of the stimuli. d. With the “Go” prime v. “No-Go” prime (IV #3) can they inhibit pressing the button? Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research 55 Design Dependent Variables: Study 1: Arousal level ü Subjective arousal / stimulation 2 questionnaires: ü Subjective sedation Study 2: Behavioral (dis)inhibition Under different priming conditions how well can the participant keep from pressing the button. ü Expect to “go” (press the button). ü Expect “no-go” (do not press). Both DVs are repeated measures: All participants get Psychology 242, Dr. Mc. Kirnan ü both questionnaires ü both go & no-go primes.

Psychology 242 Introduction to Research 56 Study 1 Hypotheses: 1. Alcohol leads to significant mood changes v. a placebo beverage a. Stimulation & arousal b. Sedation 2. Mood changes vary according to participant gender a. Men stimulation b. Women sedation Statistical test: 3 -way interaction… Alcohol v. Placebo Psychology 242, Dr. Mc. Kirnan X Male v. Female Multiple independent variables X Stimulation v. Sedation

Psychology 242 Introduction to Research 57 Study 1 Participant variables Questionnaires Alcohol: Yes No Subjective sense of stimulation Gender: Male Female Subjective sense of sedation First 2 Independent Variables Third IV Repeated measure: Stimulation v. sedation 1 manipulated, 1 measured Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research 58 Study 1 Full factorial design: IVs: This creates a 2 x 2 set of participants, or 4 cells: Gender Alcohol Consumption Male Alcohol ½ participants are men, ½ women Female Men who get real alcohol. Women who get real alcohol Men who get placebo alcohol Women who get placebo alcohol ½ get alcohol, ½ get placebo No alcohol

Psychology 242 Introduction to Research 59 Study 1 Full factorial design with a repeated measure: IVs: Gender Each participant is given two questionnaires: perceived arousal, and sedation. Alcohol Consumption Type of arousal Male Alcohol Stimulation questionnaire Sedation questionnaire Each group of participants get both questionnaires No alcohol Stimulation questionnaire Sedation questionnaire Each group of participants get both questionnaires Female Stimulation questionnaire Sedation questionnaire Each group of participants get both questionnaires Stimulation Sedation questionnaire Each group of participants get both questionnaires

Psychology 242 Introduction to Research 60 Study 1 Full factorial design with a repeated measure: IVs: Gender Alcohol Consumption Type of arousal Male Alcohol Stimulation questionnaire Sedation questionnaire Each group of participants get both questionnaires No alcohol Stimulation questionnaire Sedation questionnaire Each group of participants get both questionnaires Three IVs create a 2 x 2 design for 8 cells, using 4 groups of participants. Female Stimulation questionnaire Sedation questionnaire Each group of participants get both questionnaires Stimulation Sedation questionnaire Each group of participants get both questionnaires

Psychology 242 Introduction to Research 1 st DV: 2 -way interaction for stimulation 61 There was a 2 -way interaction of gender (male v. female) by alcohol level (alcohol v. placebo) on stimulation. Alcohol led to more stimulation than did the placebo… But primarily among men rather than women Alcohol use (IV # 1) led to stimulation, but only at the “male” level of IV #2 (gender). Multiple independent variables Figure 3 Mean ratings of subjective stimulation on the BAES under 0. 65 g/kg alcohol and placebo in women and men. Psychology 242, Dr. Mc. Kirnan

Psychology 242 Introduction to Research 1 st DV: 2 -way interaction for sedation There was also a 2 -way interaction of gender (male v. female) by alcohol level (alcohol v. placebo) on sedation. Alcohol led to more sedation than did the placebo… But primarily among women rather than men Alcohol use (IV # 1) led to sedation, only at the “female” level of IV #2 (gender). Multiple independent variables Figure 3 Mean ratings of subjective sedation on the BAES under 0. 65 g/kg alcohol and placebo in women and men. Psychology 242, Dr. Mc. Kirnan 62

Psychology 242 Introduction to Research 1 st DV: show contrast condition There was a 3 -way interaction of gender (male v. female) by alcohol level (alcohol v. placebo) by arousal (stimulation v. sedation). Alcohol (v. placebo) made men more stimulated. Alcohol made women more sedated Figure 3 Mean ratings of subjective stimulation and sedation on the BAES under 0. 65 g/kg alcohol and placebo in women and men. Psychology 242, Dr. Mc. Kirnan Multiple independent variables 63

Psychology 242 Introduction to Research Alternate portrayal of 3 -way mood interaction Placebo conditions do not show much effect The alcohol conditions show a classic “cross -over” effect for gender & mood; Men get aroused M BAES subscale scores Psychology 242, Dr. Mc. Kirnan 64 Women get sedated Multiple independent variables

Psychology 242 Introduction to Research 65 External validity: 3 -way interaction How much external validity does this finding have? Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research Study 2: Disinhibition Hypotheses: A. Men & women differ in their ability to inhibit behavior. B. Particularly after consuming alcohol. How do we take this general question into the lab to test these processes in a true experiment? In a simple reaction time task participants told to: a. …press a button quickly in response to a “go” stimulus, b. …do not press with a “no go” stimulus. Participants are first: a. …primed to expect a “go” stimulus, b. …primed to expect a “no-go” stimulus. Psychology 242, Dr. Mc. Kirnan Multiple independent variables 66

Psychology 242 Introduction to Research 67 Study 2 Full factorial design with a repeated measure: IVs: Gender 2 x 2 = 8 cells, using 4 sets of participants. Alcohol Consumption Priming condition (repeated measure) Male Alcohol “go” prime ½ participants are men, ½ women “no-go” prime (repeated measure) ½ get alcohol, ½ get placebo No alcohol “go” prime “no-go” prime (repeated measure) ¼ ¼ of participants in each major cell “go” prime “no-go” prime (repeated measure) ¼ Psychology 242, Dr. Mc. Kirnan Female “go” prime “no-go” prime (repeated measure) Each participant gets two conditions Multiple independent variables ¼

Psychology 242 Introduction to Research Design: 2 nd Dependent Variable: Disinhibition Prime Stimulus Told there is a 80% likelihood of a “No. Behavioral inhibition Go” stimulus condition; Can you not press when expecting ‘go’, but Told suddenly told ‘no-go’. 80% like -lihood of a “Go” Nostimulus Go Psychology 242, Dr. Mc. Kirnan Multiple independent variables Go 68

Psychology 242 Introduction to Research 69 Design: 2 nd Dependent Variable: Disinhibition Participant variables Participant priming Alcohol: Yes No Expect “go” stimulus Gender: Male Female Expect “no -go” stimulus First 2 I. V. s Third I. V. Repeated measure: each person gets both conditions Psychology 242, Dr. Mc. Kirnan Multiple independent variables Actual target stimulus Button press? “go” no “no-go” yes Common procedure Assess “no-go” condition only Dependent Variable: Pressing in the “no-go” condition = disinhibition

Psychology 242 Introduction to Research Dependent variable 2: “disinhibition” of button press Alcohol & gender results 1 Alcohol “disinhbition” is much stronger for men than for women But only with a “go” prime. …not for the “no-go” prime. Figure 1 Mean proportion of failures to inhibit responses to no-go targets following go and no-go cues under 0. 65 g/kg alcohol and placebo in women and men. Multiple independent variables Psychology 242, Dr. Mc. Kirnan 70

Psychology 242 Introduction to Research Summary of the 3 -way interaction 1. Alcohol (vs. no alcohol) makes it difficult to inhibit behavior 2. …primarily among men (v. women) 3. …when they are primed to act, (vs. when they are primed to keep from acting). How much external validity does this finding have? Psychology 242, Dr. Mc. Kirnan Multiple independent variables 71

Psychology 242 Introduction to Research 72 Multiple IVs; summary 1 Multiple Independent Variables / Predictors tell us much more than simple main effects. S U M M A R Y Main effect: n Test how one IV in isolation affects the DV # of major stressful events (Ages 21 to 26) Psychology 242, Dr. Mc. Kirnan Multiple independent variables

Psychology 242 Introduction to Research Multiple IVs; summary 1 Multiple Independent Variables / Predictors tell us much more than simple main effects. Main effect: One IV one DV S U M M A R Y Additive effects: n n Two IVs each have a main effect One combination has a particularly strong effect on the DV Psychology 242, Dr. Mc. Kirnan Multiple independent variables 73

Psychology 242 Introduction to Research Multiple IVs; summary 1 Multiple Independent Variables / Predictors tell us much more than simple main effects. Additive effects: Some variables may combine with others to produce very strong effects Interaction effects: n Emotional arousal S U M M A R Y Main effect: One IV one DV One IV has a different effect on the DV depending upon another IV: The effect of alcohol on emotional arousal depends upon gender. IV 1: Emotional arousal IV 2: Gender Interaction: Psychology 242, Dr. Mc. Kirnan ü Stimulation goes up if you are male. ü Sedation goes up if you are female. Multiple independent variables 74

Psychology 242 Introduction to Research Multiple IVs; summary 1 Multiple Independent Variables / Predictors tell us much more than simple main effects. S U M M A R Y Main effect: One IV one DV Increasing vulnerability: More trauma greater likelihood of depression. Additive effects: Some variables may combine with others to produce very strong effects Interaction effects: n n One IV has a different effect on the DV depending upon another IV: One IV has an effect only at one level of a 2 nd IV: No genetic vulnerability: childhood trauma has no effect on depression The effect of childhood abuse on risk for depression depends upon a genetic disposition. IV 1: Level of trauma Psychology 242, Dr. Mc. Kirnan IV 2: 3 forms of 5 -HTT genotype Multiple independent variables 75

Psychology 242 Introduction to Research Multiple IVs; summary Multiple Independent Variables / Predictors: S U M M A R Y n Are critical to theory development and testing: Changing sexual risk reduction requires that we understand both peoples’ psychological dispositions and their drug use patterns. Stress or other environmental events can “switch on” genes that create psychological or other problems; genetic dispositions and environment are not separate processes. n Establish key “boundary conditions” to theory: when and among whom does a basic psychological process operate? Alcohol makes it more difficult to inhibit behavior, but primarily among men. Psychology 242, Dr. Mc. Kirnan Multiple independent variables 76

Psychology 242 Introduction to Research Key terms: S U M M A R Y n 77 Summary Psychology 242, Dr. Mc. Kirnan n Main effect n Additive effect n Interaction n Cross-over interaction n Factorial design n Repeated measure Multiple independent variables