Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
13 Cards in this Set
- Front
- Back
|
How do you do a One-Way ANOVA? |
Analysis>Compare Means>One Way ANOVA Grouping variable and other variable Options>Descriptives, homogeneity of variance, Welch and means plot Post-hoc>Scheffe & Games-Howell |
|
|
How do you write up a One-Way ANOVA |
-Sample minimum of 30, group 10, balanced -Levene's test (significant=heterogeneity=bad) -Comparing the group means + meaning -Examine pairwise comparisons + meaning -Conclusion |
|
|
How do you do a Two-Way ANOVA? |
Analyse>General Linear Model>Univariate IVs into fixed factors, DV into DV Plots Options>all into box, descriptives, homoegenity, effect size Post hoc>all into box, Scheffe Interaction: Data>split file by 3-level, t-test |
|
|
How do you write up a Two-Way ANOVA? |
-Sample minimum of 30, group 6, balanced -Levene's (significant=heterogeneity=bad) -Main effects + meanings -Interaction + meanings -Post-hoc Scheffe tests + meanings -Interaction effects by t-tests + meanings -Conclusion |
|
|
How do you do a Factor Analysis? |
Dimension Reduction>Factor All variables into variables box Extraction>Scree plot Descriptives>coefficients, significance levels, determinant, KMO and Bartlett's Rotation>Varimax Select>save as variables, regression Options>sorted by size, suppress small coefficients (.33) |
|
|
How do you write up a Factor Analysis? |
-PCA with Varimax rotation -10:1 case to variable -Determinant (greater than .00001=no multicollinearity) -KMO >.7, significant Bartlett's (data correlation matrix is significantly different to the identity matrix so sufficient significant correlations) -Communalities greater than .2=accounts for variance of variables -K1, scree plot and parallel analysis -Factors -Any follow-ups eg t-tests -Conclusion |
|
|
How do you do a Regression? |
Analyse>Regression>Linear Predictors in Independents, outcome in Dependent Enter or stepwise Stats>descriptives, collinearity diagnostics, casewise diagnostics Plots>Zpred and Zresid Save>Standardised residuals To check residuals: Analyse>Desc Stats>Explore, normality plots with tests and Shapiro-Wilk |
|
|
How do you write up a Regression? |
-40:1 ratio in stepwise, 10:1 in enter -Variables used or ANOVA table -R, Rsquared, adj Rsquared, percentage of variance explained -Tolerances (.1 serious multicollinearity, .2 potential) -Equation and explanation -Residuals plot distribution -Shapiro-Wilks -Outliers -Take out outliers and report Rsquared and W |
|
|
How do you do a Cross-Validation (Regression)? |
Compute variable function Paired t-test with significant correlation but non-significant t-test |
|
|
How do you do a Repeated Measures ANOVA? |
Analyse>General Linear Model>Repeated Measures Replace factor 1 with factor name and put in number of repeats, take variables ove Plots>put repeated on Hz axis Options>everything in, Bonferroni, desc stats, effect size and homogeneity tests |
|
|
How do you write up a Repeated Measures ANOVA? |
-Mauchly's test of sphericity (significant=bad=corrections) -Effect -Pairwise comparisons with Bonferroni correction -Conclusion |
|
|
How do you do a Mixed Model ANOVA? |
Analyse>General Linear Model>Repeated Measures Replace factor 1 with factor name and put in levels RM variable into within, other into between Plot Options>everything in, Bonferroni, desc stats, effect, homogeneity Follow ups with t-tests (individual samples) |
|
|
How do you write up a Mixed Model ANOVA? |
-Minimum sample of 30, groups 10, balanced -Mauchly's test of sphericity (significant=bad=corrections) -Main effects + meanings -Interaction effect + meaning -Post-hoc tests -T-tests -Conclusion |
