Psychology Statistics

Independent Variable

Measures that you think cause changes in the outcome variable

Dependent Variable

The outcome variable of interest

Continious

Ranked along a scale with infinite points possible

Nominal

Arbitrary, unranked categories

Ordinal

Ranked along a dimension

Within-Subject Design

Same participants test all conditions

Between-Subject Design

Different participants assigned to different conditions

Null Hypothesis

Treatment has no effect; all groups are equal to one another; means are equal across groups; one variable does not rely on the other

Alternative Effect

Treatment has effect; at least one group is different from the others; means are different across groups; one variable relies on the other

Non-directional

2-tailed test

Directional

1-tailed test

Type 1 Error

Rejecting null when it is really true, or inferring a difference when there is none

Type 2 Error

Accepting null when it is actually false, or inferring no difference when there is one

Statistical Significance

Effect did not occur by chance

Clinical Significance

Effect is meaningful

Correlation

Two variables are related

Association

Two variables provide information about one another

Association Types

Posative, negative, -1 to 0 to 1

Regression

Line fit to the data in a plot, predicts what the likely value is for a given point

Standard Error of the Estimate

Like standard error of the mean but for regression where observed and predicted scores don’t always match

ANOVA

Analysis of Variance, used for comparing 2 or more means

Why use ANOVA instead of a T-Test for multiple comparisons

Greater risk of Type 1 error

F Statistic

If there is no effect it equals one, it cannot be negative, and is associated with p-value

Linear Regression

Association, 2+ continuous variables, 1 or more predictors

Correlation

Association, 2 continuous variables, association between variables

Paired Samples T-Test

Difference, 1 categorical IV, multiple continuous DV, 1 group

Independent Samples T-Test

Difference, 1 categorical IV, 1 continuous DV with 2 groups, independent groups

One-Way ANOVA

Difference, 1 categorical IV, 1 continuous DV, IV with 3+ groups, 1 group

Two-Way ANOVA

Difference, 2 categorical IV, 1 continuous DV

One Sample T-Test

Difference, 1 categorical IV, 1 continuous DV

Factorial ANOVA

Difference, 2 IV, 1 DV, 2+ groups

Mixed ANOVA

Difference, 2 IV, 2 DV, 2+ groups

Central Limit Theorem

For any population distribution with a well-defined mean and variance, the distribution of the means of samples of size n will be mostly normal

Cohen’s d

Effect size, uninfluenced by sample size, independent from statistical significance, .1 = small, .3 = medium, .5 = large

Main Effect

Mean differences among levels of one factor; effect of one factor collapsing across levels of the other factors

Interaction

Variability not explained by main effects; effect of one factor depends on the other factor; non-parallel lines

Partitioning Variance (Factorial ANOVA)

Divided into within treatments, group 1, group 2, and group 1 x group 2

Repeated Measures

Same subjects tested in different conditions; within subjects compared to between subjects

Advantages of Repeated Measures

Quantify and remove individual differences, less error, more statistical power, fewer subjects needed, more likely to get significant effects

p-value

Probability that the observed difference in a test occurred by random chance, assuming the null hypothesis is true