Article published in:How to do Linguistics with R: Data exploration and statistical analysis
[Not in series 195] 2015
► pp. 139–170
Chapter 7. More on frequencies and reaction times
After the previous chapter has introduced some basic elements of regression analysis, this chapter will provide a more thorough discussion of linear regression. This method enables one to model and explain the relationships between one or more explanatory variables at any level of measurement, on the one hand, and one ratio- or interval-scaled response variable, on the other hand. In addition, one can investigate interactions between explanatory variables. You will learn how to fit a multiple linear regression model, to perform its diagnostics and to interpret the results. You will also learn how to carry out non-parametric linear regression with the help of bootstrap. The case study investigates the relationship between reaction times in a lexical decision task, and such factors as word length, corpus frequency and part of speech of lexical stimuli. In contrast with the previous case studies, all these factors are tested here simultaneously in a multiple linear regression model.