Lecture 1: Recap on statistics, assumptions for parametric tests, variance, & probability

Question 1

Types of Descriptive Statistics

Answer 1

Measures of Central Tendency (Mean, Median, Mode)
Measures of Dispersion/Spread (Variance, Standarddeviation, Inter-quartile range)
Measures of Normality of the Distribution (Kurtosis and Skewness)
Graphs

Question 2

Which measure of central tendency and dispersion should I use?

Answer 2

Nominal - Mode
Ordinal - Median
Ratio - Mean

Question 3

Assumptions of Parametric Tests

Answer 3

To use PARAMETRIC tests certain assumptions about your data should be satisfied. Different tests have different assumptions.

Interval or ratio level data

No outliers (extreme or atypical scores –> they affect means and standard deviations)

Normality of distribution (we can check this assumption statistically)

Homogeneity (equality) of variance

Question 4

Assumption: Homogeneity of Variance

Answer 4

The variance in each group or condition should be similar.

Jamovi tests this for you for parametric tests of difference and violations of parametric assumptions can be corrected for: you’ll learn how to do this in the workshops
Variance (and standard deviation) is the basis for parametric tests of difference

Question 5

Assumption: normal distribution

Answer 5

You can ‘eyeball’ a histogram of the distribution and see if it looks like a normal distribution
You can look at skewness or kurtosis statistics
OR
You can see whether the distribution is statistically different from a normal distribution using Shapiro-Wilk goodness of fit tests
These look at the distribution of your data and see if it is statistically different from a normal distribution

Question 6

Graphs for identifying outliers and normality of distribution

Answer 6

Scatterplot: Relationship between variables
Histogram: Frequency distribution
Q-Q Plots: Comparison between real and normally distributed data
Boxplot: Outliers and Dispersion