Assumptions Flashcards
Chi-Square Test
Data values that are a simple random sample from the population of interest
2 categorical or nominal variables
For each combo of the levels of the two variables, we need at least 5 expected values
Independent samples T-Test
DV is continuous
IV is categorical (2 levels)
Independent observations (data come from ind. groups)
No significant outliers
DV is normally distributed (Shapiro-Wilk Test)
Homogeneity of variances (Levene’s Test)
Dependent samples t-test
DV is continuous
IV is categorical (2 levels)
Related groups/matched pairs (data comes from the same participants)
No significant outliers
DV is normally distributed (Shapiro-Wilk Test)
One -sample T-Test
DV is continuous
Independence (data are not correlated/related)
No significant outliers
DV is normally distributed (Shapiro-Wilk Test)
One-way ANOVA
DV is continuous
IV is categorical (2 levels)
Independent observations
No significant outliers
DV is normally distributed for each category of the IV (Shapiro-Wilk)
Homogeneity of variances (Levenes)
Two way ANOVA
DV is continuous
IV is categorical (2 levels)
Independent observations
No significant outliers
DV is normally distributed for each category of the IV
Homogeneity of variances for each combination of the groups of the two IV’s
Repeated measures ANOVA
DV is continuous
IV is categorical (2+ IV’s with two or more levels)
Related groups/matched pairs
No significant outliers
DV is normally distributed for each category for the IV
Sphericity (variance)(Mauchly’s test)
Mixed ANOVA
DV is continuous
IV is categorical (2 levels)
Within-subjects factors AND between-subjects factor (2+ levels)(at least one variable should contain data from separate groups and at least one other should contain data from the same group)
No significant outliers
DV is normally distributed for each category of the IV
Homogeneity of variance
Sphericity(variance)
One-way ANCOVA
DV is continuous
IV is categorical (2 or more levels)
Independent observations
No significant outliers
Residuals normally distributed for each category of the IV (Shapiro-Wilk and Q-Q plots)
Homogeneity of variances
Covariate linearly related to DV for each level of IV (seen by plotting a scatterplot of the covariate, DV, and IV)
Homoscedasticity (seen by plotting a scatterplot of the standardised residuals against predicted values)
Two way ANCOVA
DV is continuous
IV is categorical (2 IV’s with 2 or more levels)
Covariates (1 or more covariates, continuous IV)
Independent observations
No significant outliers
Residuals normally distributed for each category of the IV (Shapiro-Wilk Test and Q-Q plots)
Homogeneity of variances
Covariate linearly related to DV for each level of IV (seen by plotting a scatterplot of the covariate, DV, and IV)
Homoscedasticity (seen by plotting a scatterplot of the standardised residuals against predicted values)
Linear Regression
DV is continuous
IV is continuous
Independent observations (Durbin-Watson statistic)
Linear relationship between 2 variables (seen by creating a scatterplot of the DV against the IV)
Homoscedasticity (seen by plotting a scatterplot of the variances along the line of best fit)
Residuals (errors) of the regression line are normally distributed (use of a histogram or normal p-p plot)
Multiple Regression
DV is continuous
2 or more IVs (can be continuous or categorical)
Independent observations (Durbin-Watson statistic)
Linear relationship between the DV and each IV
No significant outliers
Homoscedasticity (seen by plotting a scatterplot of the variances along the line of best fit)
Multicollinearity (Variance Inflation Factor)
Residuals (errors) of the regression line are normally distributed (use of a histogram or normal p-p plot)