Stuff From Readings Flashcards
Cluster sampling
Randomly selecting pre-existing groups from the population, not individuals who belong to those groups.
Pre-existing, not pre-defined.
Counterbalancing
Can control potential carryover effects by presenting the levels of the IV to different participants in a different order. E.g. can work well in a repeated measures design
Interaction Effect
The effects of one variable are contingent on the level of the second variable, or different at different levels of another variable.
Increasing alpha
Will increase the probability of a Type I error and decrease the probability of a Type II error. Increasing the magnitude of alpha makes it easier to reject the null (type I error, also more power) and decreases the chances of retaining a false null hypothesis (type II error)
Central Limit Theorem
Predicts that the sampling distribution increasingly approaches normal as the sample size increases (not the number of samples, which is considered infinite)
Chi square
Best for comparing frequencies, or the number of observations in each category
F-ratio
Divide “mean square between” (a measure of variability due to the effects of error plus the independent variable) by “mean square within” (a measure of variability due to the effects of error only). If there is no independent variable, F-ratio will be 1.
T test for dependent samples
Always involves pairs. Df is number of pairs minus one
Item Discrimination
The extent to which a test item discriminates between examinees who obtain high scores vs low scores on the entire test or an external criterion. D= U - L. Index ranges from -1 to +1. Usually .35 is considered acceptable.
Item response theory
Linked with item characteristic curve
Spearman Brown Prophecy Formula
Provides an estimate of what the reliability coefficient would have been had it been based on the full length of the test. Used with split-half reliability.
Coefficient alpha
The average reliability for all splits of a test. Linked with KR-20 for dichotomous variables.
Interrater reliability
Measured by either a kappa statistic or by determining % agreement.
Communality
Each test’s communality indicates it’s “common variance” or the amount of variability in test scores that is due to the factors that the test shares in common with the other tests included in the analysis. It indicates the total amount of variability in test scores that is explained by the identified factors.
