Study Guides (390,000)
CA (150,000)
UW (7,000)
PSYCH (700)

PSYCH291 Study Guide - Midterm Guide: Discriminant Validity, Predictive Validity, Logical Positivism

Course Code
John Holmes
Study Guide

This preview shows pages 1-2. to view the full 6 pages of the document.
Potential Essay Questions
Logical Positivism:
Negative reaction to “all in the head” and untestable theories whose
concepts cannot be reliably measured, must have operations
o Operationalize and definitions
Replicable/ repetitive
Modern mechanism models in scientific psychology:
Cognitive-Affective Personality System- CAPS Model.
There is something between stimulus and response (S & R)
Objective qualities of situationsituation as construedbehaviours
Anything that looks at the stimulus and response and how it is measured
Behaviorist tradition
B.F. Skinner: rejected introspection and studied: consequences shape behaviour
Observable behaviour
Vividness effects
Vivid exp overwhelms any abstract knowledge
Human mind is not designed to think in terms of probabilities/statistics:
statistical baseline (what happens on average)
People hate being categorized colorful exp overwhelms knowledge
Third variable problem
Problem with correlations
o Third variable arises because things in the world are related
must pry apart these naturally occurring relationships
Third variable: Causal variable that makes both variables to increase/decrease
Logic of detecting a spurious association
Third variable makes 2 associations spurious
o Spurious: no correlation
Higher correlation with AC,BC rather than AB
Dealing with the bi-directionality problem
2 variables w arrow going both ways
can’t have temporal precedence you don’t’ know if a causes b or b causes a
Systematic measurement error
Affects validity
Systematic errors are biases in measurement which lead to the situation where
the mean of many separate measurements differs significantly from the actual value
of the measured attribute
o Response biases (e.g. desirability)

Only pages 1-2 are available for preview. Some parts have been intentionally blurred.

o Method variance (e.g. why are you so shy vs, putting room to meet ppl)
o Diagnostic (context) e.g. sampling shyness
“when you’re at a party, do you tend to dance on a table when
midnight approaches?”
think: to measure shyness you need to know when shyness shows
itself - revealing shyness
Reasons for poor inter-observer agreement
small number of judges and not enough to cancel out errors
dissimilar threats : different belief observers have on the concept are inconsistent
imprecise broad coding schemes definitions of what they are looking for isn’t
o test-retest reliability: give the same readings virtually
Reasons for high internal consistency reliability
bicycles I like to ride bicycles
asking a large number of questions and getting the same results within that testing
same theme
Discriminant validity
Divergent or Discriminant validity:
when you measure 2 different things they shouldn’t go together
o if they go together then why have 2 concepts?
Problem: everyone wants to have their own concept but under the same umbrella
Measures of conceptually distinct concepts should not relate strongly
Problems of shared method variance vs. actual similarity
Eg., Self-esteem, social anxiety and shyness
Proliferation of personality scales: An industry
Predictive validity
Measuring something to predict outcomes
the ability to predict something with a valid basis
o gre/lsat and future successes
o psychological vulnerability and later depression
Two five-point questions or one 10-point question will be drawn from Dr.
Wood’s lectures.
1. A researcher has the hypothesis that men’s semen prevents depression in
their female sexual partners (this is a true story; research was described in the
Globe and Mail in 2001). He compares women whose sexual partners use
condoms with women whose partners do not, and finds that those whose
partners use condoms are more depressed than those whose partners do not.
He concludes that semen does indeed prevent depression.
You're Reading a Preview

Unlock to view full version