Lecture 6D: Parametric VS Non-Parametric Statistics Flashcards
What type of data is required for parametric statistics?
Interval or ratio data (continuous data, e.g. height, weight)
What is the second criterion for parametric statistics?
Normality - must be normally distributed
How can one assess normality in data?
By examining histograms and running statistical tests of normality e.g. Shapiro-Wilk Statistic test
A significant result indicates a deviation from normal distribution.
What is the third criterion for parametric statistics?
Homogeneity of variance - variances being compared should be approx. equal
What test can be used to check for homogeneity of variance?
Levene’s test
This test evaluates if the variances between groups are equal.
What are the consequences of not meeting the homogeneity of variance criterion?
It may increase the actual probability of making a type I error, especially with unequal sample sizes
What makes parametric statistics more powerful than non-parametric statistics?
They are more likely to find a significant difference if it really exists in the population
Parametric tests utilize more information from the data.
What is a key characteristic of non-parametric statistics?
They are used for categorical (nominal and ordinal) data or when criteria for parametric statistics are not met
They do not assume normality or homogeneity.
What is the effect of using non-parametric statistics on data detail?
They rank the data, which may result in losing detail
For example, raw scores are converted to ranks.
What distinguishes independent groups from paired groups?
Independent groups have observations that are not paired, dependent, or correlated
Paired groups involve observations that are related or matched.
What is an example of a situation requiring a two-way ANOVA?
Comparing the effect of two independent variables, such as treatment type and time
This allows for examining interactions between the two independent variables.
Which statistical method can be used for comparing weights between two independent groups?
Independent t-test
This method is appropriate when the criteria for parametric statistics are fulfilled.
What is a non-parametric equivalent to the independent t-test?
Mann-Whitney U Test
This test is used when the assumptions for parametric tests are not met.
List the different statistical tests used for parametric VS non-parametric data