Two Factor Completely Between Subjects Analysis of Variance

1
Two FactorCompletely Between Subjects Analysis
of Variance
2
Hypothetical alertness data from a 2 x 2
completely between subjects factorial experiment
3
Mean alertness scores in the 2x2 completely
between subjects factorial experiment
4
Sources of Information in a 2 x 2 Completely
Between Subjects Factorial Design
Main effect of A
A x B interaction
Main effect of B

• Sampling error estimated by pooled within-group
  variance
• Main effect of treatment A estimated using
  variance of mean values of Factor A collapsed
  (i.e., averaged) over factor B
• Main effect of treatment B estimated using
  variance of mean values of Factor B collapsed
  over factor A
• A x B interaction -- estimated by variance of
  mean values of all cells minus Treatment A minus
  Treatment B

5
Hypotheses Tested in a 2 x 2 completely between
subjects ANOVA
6
Exercise

• 7.1 A researcher tests the influence of
  different dosages of a drug on the detection of
  targets flashed briefly in the visual field.
  Target location is also manipulated For some
  participants targets flash on the periphery of
  the visual field whereas, for other participants,
  targets flash toward the center of the visual
  field. Scores represent the number of targets
  identified in 10 trials. 18 participants are
  assigned randomly to conditions.

(a) Does the evidence indicate that performance
differs according to target location? (?
.05) (b) Does the evidence indicate that drug
dosage and target location interact? (? .05)
7
Mean Alertness Scores as a Function of Target
Location and Drug Dosage
Collapsed
2.33
4.00
8
Preliminary Calculations
9
Summary Table and Hypothesis Tests
10
SPSS Output
11
Hildebrand-Saints, L. Weary, G. (1989).
Depression and social information gathering.
Personality and Social Psychology Bulletin, 15,
150-160.

• Depressed persons are thought to feel a lack of
  control
• They feel chronically deprived of information
  and, therefore, have a heightened tendency to
  seek information, including social information
• e.g., what other people are like

12
Method

• 345 undergraduates
• Screened using the Beck Depression Inventory
  (BDI)
• Final sample -- 24 depressed and 24 not depressed
  participants
• Cover story
• interview another participant (confederate) as
  part of a film for another study
• Task
• select 10 prepared questions from a set of 30 to
  ask the interviewee
• questions varied in how much revealing
  information they would elicit - 8 were highly
  diagnostic
• participants were also informed that they would
  be quizzed about the interview afterward (high
  informational utility) or that they would be free
  to leave (low informational utility)

• Independent variables
• informational utility
• high (quizzed)
• low (free to leave)
• depression
• depressed
• not depressed
• 2 x 2 completely between subjects factorial
  design
• Hypotheses
• nondepressed participants will ask diagnostic
  questions only when information utility is high
• depressed participants will ask diagnostic
  questions whether information utility is high or
  low

13
Mean number of high diagnostic questions selected
by level of depression and utility condition
Only difference to exceed chance
Note Mood x utility interaction, p lt .05.
14
Mean number of high diagnostic questions selected
by depression level and utility condition