Please enable JavaScript.
Coggle requires JavaScript to display documents.
Repeated measures (Paired samples t-test (Output (Descriptive Statistics:…
Repeated measures
Paired samples t-test
compare two samples related to the same group of subjects in terms of some numerical dependent variable.
-
Variables
Independent (nominal variable used to distinguish related samples)
Dependent (interval/ ratio scale variable hypothesised to be influenced by the independent variable)
Rappel: Levels of Measurement (NOIR)
Asssumptions
The two samples being compared are random samples.
The dependent variable is measured on an interval or ratio scale.
The dependent variable is normally distributed.
There are similar variances for the dependent variable between the two samples being compared.
How the test works
Example: No significant difference
Therefore, the Paired Samples t-test is concerned with whether the mean difference (pre-post) is large in relation to the standard deviation of the pre-post-difference.
Output
-
Paired Samples t-test results: Confidence Intervals (Lower and Upper); t = t statistic; df = Degrees of freedom; Sig. (2-tailed) = p > 0.05 so there is no significant difference between the two conditions
Anova Repeated Measures
The repeated measures ANOVA test determines whether there are significant differences between 3 or more related samples for some numerical dependent variable.
Asssumptions
The dependent variable is measured on an interval or ratio scale.
The dependent variable is normally distributed.
Sphericity
There is homogeneity of variance and covariance between the repeated measurements of the dependent variable.
Normality is tested using a Kolmogorov-Smirnov test (50 or more values) or a Shapiro-Wilk test (fewer than 50 values)
Homogeneity of variance and covariance (sphericity) is tested using Mauchly’s test.
-
-