Parametric and Non-Parametric Tests Flashcards
Parametric vs Non-Parametric: Distribution
Parametric: Equal
Non Parametric: Unequal Distribution
Parametric vs Non-Parametric: Randomization
Parametric: Randomization of Sample
Non-Parametric: No Randomization
Parametric vs Non-Parametric: Curve Characteristics
Parametric: Bell Shaped Curve: Normal Distribution
Non-Parametric: Skewed Curve
Parametric vs Non-Parametric: Data Types
Parametric: Interval or Ratio
Non-Parametric: Nominal or Ordinal
What type of research is considered to be more “powerful” parametric or non-parametric?
Parametric
What is a T-Test? Parametric or Non
Parametric
- Compares the difference between 2 independent groups
- Equal sized groups
- Different Characteristics
e.g: 30 athletes vs 30 non-athletes
What is a paired T-Test?
Compares difference between 2 matched samples
- Paired; sample characteristics are the same
e.g: Comparing 30 of the same athletes pre and post training
Are Tailed Tests exclusive to either paired or independent t-tests?
No, tailed tests can be used for both types
Why would we implement a 1 tailed test?
If we have an hypothesis and we are inferring/know the direction of the hypothesis/tail.
- Directional Hypothesis
e.g: comparing pre and post test, we can infer the post test scores would be better
Why would we implement a 2 tailed test?
- Non-direction
- 2 ends of distribution positive or negative
Done when we cannot assume which group would be better or worse based on our hypothesis
When would we use an ANOVA?
When we have greater than 2 groups with equal sample size in each group
- Parametric
When to use a One Way Anova
More than 2 independent groups compared on 1 intervention/variable
e.g Comparing 30 Gymnasts to 30 Football players to 30 baseball players with an exercise
When to use a 2-Way ANOVA
More than 2 independent groups compared on 2 interventions/variables
e.g Comparing 30 Gymnasts to 30 Football players to 30 baseball players with an exercise and gender
When to use a repeated measures ANOVA?
When we are trying to measure individuals over time
- Repeatedly looking at the same group
- At baseline
- 2 weeks
- 4 weeks
What is an ANCOVA?
Comparing more than 2 groups while controlling for covariance.
- Trying to control this to eliminate bias
- Eliminated the covariance