Chapter 17 Testing for Difference Between Two Samples Flashcards
Parametric Test (4)
- Relatively powerful significance test
- Uses estimations of population parameters
- Data tested use usually therefore satisfy certain assumptions
- Also known as a distribution dependent test
Distribution Dependent Test (2)
- Significance test
2. Using estimations of population parameters
t-Test (2)
- Parametric difference test
2. For data at interval level or above
Distribution Free Test (2)
- Significance test
2. Does not depend on estimated parameters of an underlying distribution
Related t Test (2)
- Parametric difference test
2. For related data at interval level or above
Unrelated t Test (2)
- Parametric difference test
2. For unrelated data at interval level or above
Difference Mean (3)
- Mean of differences
- Between pairs of scores
- In a related design
Cohen’s d (1)
- Measure of effect size
Pooled Variance (3)
- Combination of two sample variances
- Into an average
- In order to estimate population variance
Non-Parametric Test (3)
- Significance test
- Does not make estimations of parameters of an underlying distribution
- Also known as a distribution free test
Data Checking (3)
- Checking that data are suitable for a parametric test
- Including checking normality
- And testing for homogeneity of variance
Transformation of Data (3)
- Performed in order to remove skew from a data set
- So that it conforms to a normal distribution
- Thus enabling the use of a parametric test
Robustness (3)
- Tendency of test
- To give satisfactory probability estimates
- Even when data assumptions are violated
Mann-Whitney U Test (2)
- Ordinal-level significance test
2. For differences between two sets of unrelated data
Power Efficiency (2)
- Comparison of the power
2. Of two different tests of significance
Wilcoxon’s T - Matched Pairs Signed Ranks (2)
- Ordinal-level significance test
2. For differences between two related sets of data
T (1)
- See Wilcoxon test