Design Equation and Statistics

Abstract

The previous chapter discussed the different types of variables one has to ­understand to design a study. This chapter uses the experimental design equation to show how these variables contribute to variability in results. The independent variable is assumed to affect the scores of the dependent variable. Other variables need to be controlled so that the effect of the independent variable can be clearly seen. Then, an overview of how to test the changes that the different levels of independent variables bring about is provided. Descriptive statistics are introduced first. These statistics describe the results, such as mean and standard deviation. Then inferential statistics or statistical testing follows. Statistical testing allows the researchers to draw conclusions about a population of users based on a study that involves a ­sample. Underlying, essential principles, such as the standard distribution and the central limit theorem, are reviewed, followed by the three tests most commonly performed in informatics: t-test, ANOVA and chi-square.