FINAL REVIEW

1
FINAL REVIEW!
2
CONDUCTING EXPERIMENTSCh 9
3
Experiments Causation Sensations!

• What is the function of a true experiment?
• How does a manipulation achieve this?

4
Extraneous vs Confounding Var

• Extraneous variable- general term, variables
  other than DV and IV
• (extra variable)
• Confounding variable- a type of extraneous
  variable that systematically varies with IV
• (confusing extra variable!)

5
Taking Care of Extraneous Var

• Control by holding constant
• Control by matching
• Control by randomization
• Pros and Cons?

6
Solomon Design

• Pretest-posttest Posttest-only!
• Half of the participants receive only the
  posttest
• Other half receive pretest and posttest
• Why use Solomon?

7
Two Blind Mice

• Single Blind Procedure
• Double-Procedure

8
Practice Questions!

• Which of the following studies of aggression
  exemplify mundane realism?
• Participants subtract points from an online
  competitor
• Participants fight in a boxing ring
• Participants hit pillows shaped as people
• B and C

9
Practice Questions!

• What is the key problem(s) with Prettest-Posttest
  Designs
• a) Mortality effects
• b) High cost
• c) Sensitization
• d) A and C

10

Between Subjects Independent Groups
Independent groups receive unique
treatments, so there are differences between
groups
Within Subjects Repeated Measures
_________________________________________________
___________________________________________
Within all the subjects every treatment is
repeated
11
Confounding Variables in Between Subjects Design

• Individual differences
• Participants differ on some key characteristic by
  group
• Environmental differences
• Environment experiment administered in differs
  for each group

12
Variability

• Variability between conditions? YAY!
• Variability within conditions? OH NO!

13
Disadvantages of Within Subjects
14
Counterbalancing!

• Latin Square Design
• Control for order effects without having all
  possible orders
• A B C D
• D A B C
• C D A B
• B C D A

15
Practice Question

• Which of the following eliminates order effects?
• Structured debriefing
• Control group
• Counterbalancing
• None of the above

16
Practice Question

• Which of the following is a threat to internal
  validity found in between subject designs
• The level of the IV varies by condition
• The effect of the IV in one group spreads to the
  other group
• Participants are equally likely to withdraw from
  both experimental conditions
• There is too much variability
• B and D

17
Complex Designs!
18
Factorial Designs

• 2 or more independent variables
• Each level of IV 1 is paired with each level of
  IV 2
• IV 1 electric shock
• IV 2 ice-cream
• DV- reported happiness

19
Main Effects

• The overall effect of one independent variable,
  averaging across all levels of other IV
• People are less happy when they receive an
  electric shock

20
Interaction Time!

• The effect of one independent variable depends on
  the level of the other independent variable

21
(No Transcript)
22
(No Transcript)
23
Practice Question

• Standing on its own, which of the following is a
  valid interpretation of the graph?
• There is a main effect for ice-cream flavor
• There is a main effect for electric shock
• There is an interaction between flavor and
  electric shock
• All of the above

24
QUASI EXPERIMENTALSINGLE CASEDEVELOPMENTAL
25
Quasi-Experiment

• IV not manipulated (gender, age)
• Potential confounding var
• Why?
• Person by treatment design
• One IV is measured (gender)
• One is manipulated (electric shock)

26
Nonequivalent Designs

• Nonequivalent posttest-only control
• Nonequivalent pretest-posttest control

27
Single Case Experimental

• ABAB Reversal
• Baseline, treatment, baseline, treatment
• Multiple baseline
• Treatment administered under multiple
  circumstances to see if there is a change

28
Cross Sectional vs Longitudinal

• Cohort Effects
• Participant Attrition
• Practice effects

29
Practice Questions

• Which of the following is a disadvantage specific
  to the ABAB design?
• Only using one subject
• Ethical considerations for withdrawing treatment
• Not representative of population
• Time consuming and expensive

30
Practice Question

• You bring depressed and non-depressed people into
  the lab and expose them to either white fluffy
  dogs or large growling dogs, you then measure
  their heart rate. This is an example of a
• Quasi-experimental design
• Between subjects design
• Person by treatment design
• All of the above
• A and C
• B and C
• A and B

31
STATISTICS!
32
Inferential Statistics

• Use information about a sample to draw
  conclusions about the population
• Sampling error- discrepancy between information
  from the sample and what really goes on the
  population

33
Hypothesis Testing

• Determine the null and alternative hypothesis
• Determine the alpha level
• Compute the statistical test and compare
  significance to pre-determined alpha

34
Alpha Level

• The probability that the research results was
  obtained by chance
• Will only reject the null hypothesis if its very
  unlikely that it was due to chance
• If alpha .05, then we reject the null
  hypothesis when there is only a 5 likelihood
  that results due to chance

35
ERROR!
36
T-Test

• Compare two means
• Group Differences
• __________________________________________________
  _________
• Within Group Variability
• -Differences between groups large
• -Differences within groups small

37
F-Test (ANOVA)

• Comparing 3 or more groups to see if at least one
  is different from another
• F variance between groups
• _____________________________
  ____________________________
• variance expected by chance

38
Chi-Squared Test

• Use when DV is categorical/nominal
• DV Do you like tests? Yes or No
• IV Professors vs Students
• Chi-Squared will determine whether students are
  more likely than professors to say they dont
  like tests

39
Practice Questions

• You hypothesize that men are taller than women.
  You measure the heights of men and women. What
  kind of statistical test should you use?
• Two tailed t-test
• One tailed t-test
• Two tailed F-test
• One tailed F-test

40
Practice Question

• Which of the following measures of central
  tendency is affected by extreme scores
• Median
• Mode
• Mean
• B and C