What does reliability deal with?
The consistency of a measure
A test can be reliable without being valid, but cannot be valid without being reliable.
True or false: A real change in a certain trait being measured is a source of unreliability.
FALSE
Fluctuations in scores due to personal circumstances are considered sources of unreliability.
What is a constant error?
An error that leads to a person’s score being systematically low/high
Reliability only deals with unsystematic errors, not constant ones.
What does a bivariate distribution represent?
The relationship between two variables
It is often visualized as a scattergram or scatter plot.
What does the correlation coefficient r provide?
A numerical summary of the relationship between two variables
It only accounts for linear relationships.
What is a regression line used for?
To predict status on Y (dependent variable) from knowledge on X (independent variable)
Assumes that the Ys have a normal distribution with the same standard error of estimate.
What does homoscedasticity refer to?
Equal scatter along the line
The opposite term, heteroscedasticity, refers to different scatter along the line.
What are major sources of unreliability? List them.
- Test scoring
- Test content
- Test administration conditions
- Personal conditions
Each of these can introduce unsystematic errors in test scores.
Define observed score.
A person’s actual score on a test
It is different from the true score, which would be obtained without any unreliability.
What is test-retest reliability?
Administration of the same test to the same individuals on two occasions
The reliability coefficient is the correlation between the first and second scores.
What does inter-scorer reliability assess?
Unsystematic variation due to who scores the test
The coefficient is the correlation between scores obtained by the first and second scorers.
What is alternate form reliability also known as?
Parallel form or equivalent form reliability
It involves administering two forms of a test to the same examinees.
What is internal consistency reliability?
A method of expressing test reliability based on how well items in a test correlate with each other
Common methods include split-half reliability and coefficient alpha (Cronbach’s alpha).
What does the Spearman-Brown correction address?
The reliability for a test half the length than the one we are interested in
It is used in split-half reliability calculations.
What is the rule of thumb for comparing groups regarding reliability (r)?
At least 0.7
For individual diagnostics, r should be at least 0.8 or 0.9.
What does the standard error of measurement (SEM) represent?
The standard deviation of a hypothetically infinite number of obtained scores around the person’s true score
SEM can be used to create a confidence interval.
List the three types of standard errors.
- Standard error of measurement
- Standard error of the mean
- Standard error of estimate
Each type serves a different purpose in statistical analysis.
-
Week 121
-
Week 232
-
Week 336
-
Week 415
-
Week 4 part 215
-
Week 536
-
Week 625
-
Week 710
-
Summary The World of Psychological Testing11
-
Summary Test Norms19
-
Summary Reliability17
-
Summary Test Development, Item Analysis, and Fairness14
-
Summary Validity25
-
Summary Factor Analysis13
-
Summary Intelligence: Theories and Issues19
-
Summary Individual Intelligence Tests13
-
Summary Objective Personality Tests13
-
Summary Clinical Instruments and Methods21
-
Summary Projective Techniques11
-
Summary Interests and Attitudes14
-
Chapter 1528
-
Chapter 1414
-
Chapter 1330
-
Chapter 1226
-
Chapter 828
