431 Review

1
431 Review
2
Final Exam

• Section 1 (50)
• 45 MC- 1 point each
• Section 2 (50)
• 25 short answers- 1 point each
• 1 long answer, from a choice of 5- 10 points.
• You will have three hours.

3
General points

• There will be some memorization questions, but
  you will not need to memorize many tiny details
  that you would not know in real life
• I try to make most questions conceptual e.g. I
  want to test whether you can think
  psychometrically
• There may be so-called trick questions
  Questions that require you to understand what I
  am asking

4
Meta-analysis Why no pre/post designs?

• Pre/Post designs confound time with treatment
• We need to compare waiting list to treatment to
  unconfound these two

5
Surveys What is stratified sampling?

• This is what I called cluster sampling stratify
  the population into non-overlapping groups,
  randomizing units at each level

6
Information

• What is information?
• How can it be recognized and measured?
• Where does it come from?
• Do deduction, induction, and abduction all lead
  to new information?

7
Probability

• Know the one-slide summary
• Be prepared for both non-conditional and
  conditional probability problems
• Understand all implications of 2x2
  dsitribution/covariation tables

8
Statistics

• Understand why many things are normally
  distributed- and when they might not be
• Understand that the normal curve is a
  distribution of probabilities (hence the area
  under the curve sums to 1)
• Understand what the binomial theorem tells us
• Know the properties of a correlation

9
Rationality

• Know the cognitive illusions in the course pack
  reading be able to recognize them from
  description or name

10
Psychometrics

• What is a psychometric test?
• Have some knowledge about the most common tests
• Understand clinical vs actuarial strengths
• Know what a linear versus non-linear relation is

11
Validity

• Be familiar with the nomological net
• What turns a law from statistical into causal?
• What is validity?
• Construct? Convergent/Discriminant? Face?
  Criterion-related?
• What impacts validity?
• What are moderator variables?
• Know the Multitrait-multimethod matrix

12
Reliability

• What is reliability?
• Test-retest parallel forms split-half its
  improvements inter-rater intraclass
• Please learn how to spell Cronbachs!
• Know what a poor, reasonable, and excellent
  Cronbachs alpha is
• What is the formal definition? Why cant we use
  it directly?
• How is reliability related to validity? To
  SEmeasurement? To SEmean?
• Why is SE normally distributed?

13
Factors

• Know what a factor is (and isnt!)- both
  informally and more formally
• Understand orthogonality
• Understand axes rotation, and its ontological
  status
• Have some understanding of how to read factor
  analysis output what is a factor table, an
  eigen-value, a scree plot?
• What was Cyril Burts error?

14
Cutting scores

• Understand a cutting score, and the trade-offs
  involved in setting cutting scores
• What is the relation between cutting scores and
  conditional probability/Bayes Theorem?
  Covariance tables? Base rates?
• What is a selection ratio? Sensitivity?
  Specificity?
• You dont need to memorize (but should
  understand) the implications of Bayes Theorem we
  discussed, as laid out in the paper

15
Intelligence

• Know the issues around IQ and its measurement
• Is IQ a construct? Does it have a single factor
  or more? How reliable are IQ tests? What are the
  issues around their validity?
• How is the WAIS (point-scale format) different
  from the Stanford-Binet (age-scale format)? From
  Ravens Matrices?
• How heritable is IQ?

16
MMPI

• What makes the MMPI very good? What are its
  weaknesses?
• What are the three main categories of scales?
• What scales are unique to the MMPI? (Q, K,L,F)
• How valid is the MMPI?

17
Projective Tests

• What is the projective hypothesis?
• Know the main projective tests, and who designed
  them.
• What is useful/bad about projective tests?

18
Personal constructs

• What is constructivism?
• Know something of Kellys theoretical stance
  (Personal construct theory), and its jargon terms
• Understand the repertory grid, with its
  variations, and how it relates to Q-sorting

19
Prediction

• Understand the relation between group differences
  and individual placement
• How do we decide if groups are different? How we
  decide which group an individual comes from? See
  also Neuropsychology
• Understand the issues in inferring from group
  membership Why are psychometricians so insistent
  that this be allowable?
• Why be concerned that similarity and prediction
  is more highly correlated than similarity and
  base rates?
• What do we need to make good predictions?
• What is good/bad about GREs and SATs? Which part
  of the tests predicts future performance best?

20
Neuropsychological Tests

• Know the main tests we discussed
• Why did I argue that neuropsychological tests
  have special validity status?
• The main conceptual meat is in the group versus
  individual classification problem see
  Prediction.

21
Meta-analysis

• Understand the steps and concepts behind
  meta-analysis
• Understand the meaning and interpretive
  limitations of a p-value
• What are fixed versus random effects?
• What is an effect size?
• What is Cohens d?