A repeated/related/within subjects design is the type of design when repeated observations on the same subject take place (often, it is a pretest-posttest combination).An independent/unrelated/between subjects design is the design when the subjects of the experiment are randomly selected from the population, and randomly assigned to some groups; the participants take part in the experiment (or are observed in a treatment combination) only once.

A quasi design is the design when the researchers follow within subjects or between subjects design procedures, but do not select the participants using random assignment.

## Define nominal, ordinal, interval and ratio data

Nominal data are represented by categorical unordered variables, e.g. by the variables which are characterized by labels, the numeric values of which are only used in text sense (it is not possible to establish any order within categories). Ordinal data are represented by categories which can be logically ordered, but there is no information on the differences between the values. Interval variables are represented by continuous variables measured by equal intervals, which can be associated with equal differences in the property measured by these variables. Interval scales do not have a fixed zero point, while ratio scale possess a meaningful zero point. Thus, the meanings of ratio variables are valuable along the whole scale (unlike interval variables, where only the values of the variables can be meaningful).

## What do the terms parametric and non-parametric mean?

Parametric statistics and tests relate to the data which are suggested to have normal distribution. These tests are more powerful, but there is a number of assumptions to be held for parametric tests: independence of observations, and the populations used for research should be normally distributed and have the same variances. If some of these assumptions are not true, then it is recommended to use non-parametric statistics, e.g. the methods not based on the properties of normally distributed population.