- Chapter
Chapter 12: Correlation, Regression, and Non-Parametric Tests
pp. 281-310- Add bookmark
- Cite
- Share
Summary
CHAPTER PREVIEW
In the previous chapter we offered an introduction to survey research. Quite often, data from surveys are analyzed using correlations or non-parametric techniques. We now consider how to analyze the information obtained from surveys using correlation, regression, and Chi-Square techniques.
Psychologists frequently use correlational approaches to investigate the relationships among variables when experimental approaches are not feasible. Correlational studies permit us to find relationships between variables. The simplest correlational technique involves a relationship between two variables, or a bivariate correlation. Researchers typically employ the well-known Pearson product-moment correlation in such circumstances. We can also use the correlation coefficient to develop a regression equation that can be used to predict behavior. A regression or prediction equation is based on the mathematical relationship between variables, and it can aid in prediction of events. Survey data are often used to predict future outcomes or behavior.
Quite often when we conduct a survey, we are interested in examining whether our sample, or respondents, accurately reflects the population. In other words, we must ensure that demographic characteristics of our sample are similar to that of our population. For example, if we conduct a survey of Psi Chi members, we will want to be sure that our results are generalizable to the larger population of Psi Chi members. In order to generalize our survey results to the larger group of Psi Chi members, we will need to make sure that our sample is comparable to the full membership of Psi Chi. For example, we will need to compare the number of women in our sample to that of the population. We can use a non-parametric technique—Chi-Square (Chi sounds like sky with the s missing) goodness-of-fit test—to test this comparison.
In this chapter, we also introduce a second type of Chi-Square analysis, or the test of independence, is used when we wish to determine if two variables are independent. For instance, is there independence between the numbers of women versus the number of men majoring in psychology versus chemistry? In other words, is gender related to choice of major, or are these variables unrelated or independent? We can use a second type of Chi-Square, or a contingency test, to examine this relationship. In both cases, the test of independence and the contingency test, we use nominal, or categorical, data to conduct these analyses.
About the book
- Chapter DOI https://doi.org/10.1017/9781108550734.013
- Book DOI https://doi.org/10.1017/9781108550734
- Subjects Qualitative Research Methods
- Format: Digital
- Publication date: 06 August 2018
- ISBN: 9781108550734
- Find out more details about this book
Access options
Review the options below to login to check your access.
Personal login
Log in with your Cambridge Higher Education account to check access.
Purchase options
If you believe you should have access to this content, please contact your institutional librarian or consult our FAQ page for further information about accessing our content.