Psychology - Statistics 2Nd Year

Front 1

Bonferroni Correction

Back 1

- Corrected level of significance when performing multiple comparisons (simple effects)
- .05/number of tests

Front 2

3 Decimal Places

Back 2

- P values
- R values
- Partial n2 values

Front 3

Reporting T-test Results

Back 3

- t value, p value, 95% Confidence Interval
- Cohens d (effect size)

Front 4

Reporting One-Way ANOVA Results

Back 4

- Significance: f value, p value
- Partial n2 (effect size) + variance explained (%)
- Post hoc test: p and d values for each comparison

Front 5

Reporting Two-Way ANOVA Results

Back 5

For both Main Effects and Interaction:
- Significance: f value, p value
- Partial n2 (effect size)

- Post hoc test: p values

Front 6

Reporting Correlation Results

Back 6

> Pearsons r + p value
> Strength of relationship (weak/moderate/strong)
> Percentage of shared variance*

Front 7

Reporting Simple Effects

Back 7

- Number and type used
- Bonferroni adjusted criterion
- t-test results + cohens d

Front 8

Reporting Linear Regression

Back 8

- R2 (squared)
- ANOVA Result: F + p
- Slope + 95% CI

Front 9

Reporting Multiple Linear Regression

Back 9

- Variance explained % = (R2*100)
- ANOVA Result: F + p
- T statistics for each predictor
- Slope + 95% CI
- Most influential predictor (based on Beta)

Front 10

Confounding Variable

Back 10

- Hidden extraneous variable
- Correlates with IV and DV
- Can cause us to see a causal relationship where there is none

Front 11

Quasi-Experiment

Back 11

- Unable to randomly allocate participants to conditions
- e.g. age or gender

Front 12

Kurtosis

Back 12

- Peakedness of population distribution
- i.e. steep or flat

Front 13

Skewness

Back 13

- Asymmetry of a probability distribution
- i.e. how far left/right distribution is from centre (mean)
- Negative = Right
- Positive = Left

Front 14

Interval Scale

Back 14

- True numerical relationships
- Equal intervals
- No true zero point
- e.g. Temperature

Front 15

Nominal Scale

Back 15

- No numerical relationship
- Categorical
- e.g. Gender, Ethnicity, Job

Front 16

Ordinal Scale

Back 16

- Data organised by rank
- True numerical relationships
- BUT intervals between values not equal
- e.g. Likert Scale

Front 17

Ratio Scale

Back 17

- True numerical relationships
- Equal intervals
- True zero point
- e.g. Speed

Front 18

Type 1 Error

Back 18

- Reject null hypothesis when it is in fact TRUE
- Believing there's a relationship when there isn't

Front 19

Type 2 Error

Back 19

- Fail to reject null hypothesis when it is in fact FALSE
- Believing there's no relationship when there is

Front 20

Residual Variance

Back 20

- Variance not explained by regression line
- Distance between each data point and the regression line
- SSr

Front 21

Total Variance

Back 21

- Linear regression
- Variance not explained by the mean
- Distance between each data point and the mean (horizontal line: simplest model)
- SSt