Research Design and Analysis Ch 7
interactions and simple main effects
- Analysis of variance is used to assess the statistical significance of main effects and the interaction in a factorial design. - Simple main effect analysis examines mean differences at each level of the independent variable.
interpretation of factorial designs
- Main effects of an independent variable - Interaction between the independent variables
Determining number of interactions
1. In a study examining the effect of room illumination (low, medium, high) and room temperature (cold, warm, hot) on test performance, how many interactions are possible?- 1 interaction (between the two independent variables, 3x3 design?) 2. A researcher manipulates a defendant's appearance (attractive, average, or unattractive) and gender (male or female) to study how these variables affect judgments of criminal behavior. How many interactions are possible in this design?- 1 interaction
picture has how many main effects
2 main effects
Determining number of main effects
2x2 design: 1 interaction (2 main effects) -AxB -A researcher studies the effect of room music (fast, slow) and room cleanliness (tidy, untidy) on people's dining experience at a restaurant. How many main effects are possible?- 2 main effects ----- 2x2x2 design: 4 interactions (3 main effects) -AxBxC (aka 3 way interaction) -AxB -BxC -AxC
increasing the number of levels of an independent variable
2x2 is the simplest factorial design. With this basic design, researchers can arrange experiments that are more and more complex. Can increase complexity by increasing number of levels of one or more of the independent variables. Ex: 2x3 contains 2 independent variables. IV A has 2 levels and IV B has 3 levels. Thus, the 2x3 design has 6 conditions.
factorial designs with three or more independent variables
2x2x2 has 8 conditions 2x2x3 has 12 conditions 2x2x2x2 has 16 conditions Problem with having too many independent variables?- the design may become needlessly complex and require enormous numbers of participants
independent groups (between-subjects) design
An experiment in which different subjects are assigned to each group. Also called between-subjects design or between-persons design. Ex: if you have planned a 2x2 design and want 10 participants in each condition, you will need a total of 40 different participants
repeated measures (within-subjects) design
An experiment in which the same subjects are assigned to each group. Also called within-subjects design or within-persons design. -If you want 10 participants in each condition, only 10 would be needed. -This design offers considerable savings in the number of participants required.
factorial designs with manipulated and non manipulated variables
IV x PV design: A factorial design that includes both an experimental independent variable (IV) and a nonexperimental participant variable (PV). Allow researchers to investigate how different types of individuals respond to the same manipulated variable. - If a researcher has male and female participants drive a course under dry or wet road conditions, what kind of research design does he or she have?- IVxPV design - An educational researcher examines the effect of speaker credibility on attitude change in university and community college students. The PV in this design is the- type of student
assignment procedures and factorial designs
Independent groups design Repeated measures design Mixed factorial design using combined assignment
outcomes of a 2x2 factorial design
Possibilities 1. There may or may not be a significant main effect for independent variable A 2. There may or may not be a significant main effect for independent variable B 3. There may or may not be a significant interaction between the independent variables Example of an experiment with 2 IV: Effect of sex and violence on recall of advertising. IV A: exposure to violence- nonviolent vs violent video IV B: participant sex- male vs female DV: number of ads recalled (range from 0 to 8)
interaction
Situation in which the effect of one independent variable on the dependent variable changes, depending on the level of another independent variable. - Can be seen in M&M study. When confederate is unsociable, subjects consume more M&Ms when food intake is high. But when confederate is sociable, food intake has little effect, which is opposite of what is expected based on modeling. - interactions can be seen easily when the means for all conditions are presented in a graph (picture attached) - Example: friend asks if you want to go to a movie. Whether you want to go may reflect an interaction between two variables. 1) is an exam coming up?, 2) who stars in it?. If an exam is coming up you won't go, if you don't have an exam to worry about, your decision depends on whether you like the actors in the movie. Much more likely to go if favorite actor is in it. - Example 2: A researcher finds that for female applicants, the likelihood of being hired for a job increases as their work experience increases. However, for male applicants, the likelihood of being hired decreases as their work experience increases. This finding suggests: an interaction between gender and work experience
curvilinear relations
U shape. Design with only 2 levels of independent variable cannot detect a curvilinear relationship. Needs 3 levels minimum. If only levels 1 and 3 used, no relationship will be detected (may look like a flat line)
picture has
an interaction and 2 main effects
simple main effect of sociability
compare the sociable vs unsociable conditions when the food intake is low and then when food intake is high. The simple main effect you will be most interested in will depend on the predictions that you made when you designed the study.
complex factorial design (terms to understand)
concerned with more than one factor
increasing the number of independent variables: factorial designs
factorial design: A design in which all levels of each independent variable are combined with all levels of the other independent variables. A factorial design allows investigation of the separate main effects and interactions of two or more independent variables. Simplest is a 2x2, which has 2 independent variables, each having two levels
main/simple main effect
main effect: The direct effect of an independent variable on a dependent variable. simple main effect: In a factorial design, the effect of one independent variable at a particular level on another independent variable. Effect of one independent variable averaged over the levels of the other independent variables. Effect each independent variable has by itself. ---- --In order to ascertain the main effect of an independent variable, a researcher must use a factorial design. --Which of the following is the general format for describing factorial designs?- Number of levels of first IV × Number of levels of second IV
mixed factorial design using combined assignment
mixed factorial design: A design that includes both independent groups (between-subjects) and repeated measures (within-subjects) variables. -Ex: study on distraction and extraversion. Participant variable: extraversion, is an independent groups variable. Distraction is a repeated measures variable, all participants studied with both distraction and silence -10 participants assigned to level 1 and another 10 assigned to level 2. IV B is a repeated measures variable, so the 10 assigned to A1 receive both levels of B, and the 10 assigned to A2 also receive both levels of the B variable. So a total of 20 participants are required.
what is going on in this study
one main effect and an interaction. Interaction because the lines will cross eventually. one main effect is because: top line across: (80+60)/2= 70 bottom line across: (30+50)/2= 40 mismatch so 1 main effect^ but first top bottom points: (80+30)/2= 55 second top bottom points: (60+50)/2= 55 match so no main effect^
simple main effect of confederate food intake
simple main effect of confederate food intake is significant when the confederate is unsociable (means of 2.14 vs 10.63), but the simple main effect of confederate food intake is not significant when the confederate is sociable (means of 6.58 and 5.68)
explain why we use factorial designs
to manipulate two or more independent variables
increasing the number of levels of an independent variable (textbook outline)
why increase levels? 1- only two levels of one independent variable cannot provide very much information about the exact form of the relationship between the independent and dependent variables. (ex: including results of other #s of practice throws can give more info. A simple positive relationship can be elaborated as a positive monotonic relationship- practice increases performance up to a point, after which further practice is not helpful). 2- only two levels cannot detect curvilinear relationships between variables. If a curvilinear (nonmonotonic) relationship is predicted, at least three levels must be used. 3- researchers frequently are interested in comparing more than two groups. Ex: seeing effect of different animals on nursing home residents rather than just dogs.
