PreFinal Exam Review

1
Pre-Final Exam Review

• Stanovich, Ch. 9 - 11
• Pelham Blanton, Ch. 3 12
• All lecture notes since Midterm

2
Outline

• Exam format
• Review presentation
• Questions about things I didnt cover

3
Exam Format

• Entirely multiple choice
• bring pencil and calculator
• Non-cumulative, although there is considerable
  overlap between textbooks
• in case of discrepancies, base your answer on
  material covered after Midterm
• 90 mins max
• December 16th, at Noon, Osborne A

4
Human Behaviour is Complex

• Most behaviours are the result of (caused by)
  many factors
• To say that A causes B, does not mean that A is
  the only thing that causes B
• Most behaviours result from a variety of
  biological, dispositional, social and
  environmental factors.

5
Interaction

• The magnitude of the effect of one independent
  variable depends on the level of another
  independent variable
• Types
• Ordinal / Spreading effect exists at one level
  of the other IV, but is weaker or non-existent at
  the other (Pelham Blanton, p. 231)

6
Example
7

• Types contd
• Disordinal / Crossover NO MAIN EFFECTS. The
  effect of an IV at one level of a second IV is in
  the opposite direction at another level of that
  second IV.
• This can be tricky. Lines can cross in an
  ordinal interaction. Not a disordinal
  interaction if there is a main effect (Pelham
  Blanton, p. 234).

8
Example

• Is this a ordinal/spreading interaction or a
  disordinal/crossover interaction?

ORDINAL
9
Example
Spreading or Cross-over?
Cross-over
No main effect
10
Validity

• The relative accuracy or correctness of a
  psychological statement.
• 4 types
• Internal
• External
• Construct the extent to which operationally
  defined variables map on to the abstract concepts
  we are interested in.
• Conceptual the extent to which the study
  procedures map on to the hypothesis being tested.

11
Measurement scales

• Nominalnamed categories.
• Male/female
• Ordinalordered categories.
• Birth order, ranked likings
• Intervalcertain distance between categories.
• Temp., most psych scales
• Ratiotrue zero, no negative values.
• Weight, magnitude

12
Threats to Validity

• 1. Individual Differences
• 2. Selection Bias
• Self-selection bias people select themselves
  into a study
• Non-response bias people select themselves out
  of a study

13

• 3. History changes occurring in larger group
• 4. Maturation changes occurring in the person
• 5. Regression Toward the Mean
• 6. Testing Effects
• 7. Attrition
• Homogeneous attrition equal level of attrition
  across all experimental conditions
• Heterogeneous attrition attrition is different
  depending on the experimental condition

14

• 8. Participants Reaction Bias
• Try to do what researcher expects (participant
  expectancies)
• Try to do the opposite of what researcher expects
  (participant reactance)
• Try to do whatever makes them look good
  (evaluation apprehension)
• 9. Experimenter Bias
• Researchers make biased observations
• Researchers treat their subjects differently
  depending on what condition they are in
• 10. Confounds systematic variability
• Confounds vs. noise

15
Research Designs

• Between Groups
• Within Groups
• Between-within Groups
• Know advantages and disadvantages of each
• Counterbalancing and Latin Square

16
Statistics

• Descriptive vs. Inferential
• Measures of Central Tendency
• Mean, Median, Mode
• Measures of Variability
• Range, Standard Deviation
• Shape of distribution
• the normal curve

17
Example data

• Measuring depression on a 20-item measure of
  depressive symptoms
• Score lt 7 not depressed
• Score 7 15 mildly depressed
• Score gt 15 severely depressed, suicidal
• 10 patient scores
• 1, 3, 4, 7, 8, 10, 12, 17, 18, 20
• On average, are my patients highly depressed?

18

• Should I worry about suicide?
• Calculate the standard deviation
• A good chunk of my patients score above 15!

- 1
19
Hypothesis Testing

• a probability of making a Type I error
• ß probability of making a Type II error
• 1 ß power to detect a significant difference