SPSS Practice: Correlation Testing
To best prepare for the assignment in this unit, you must become familiar with some basic statistical skills related to correlation testing. Although there are several tests that you could choose from, such as Pearson, Spearman, Kendall, and Biserial, you will only need to understand the basic differences for two slightly different correlation tests: Pearson correlation and Spearman correlation.
Before performing any inferential statistical analysis, it is common for a researcher to look for relationships among the various variables for which data has been collected.
For example, in our scenario, we want to find out if a change to a vegetarian diet from a typical American omnivorous diet will have an effect on emotional well-being. Before we come to that part of the analysis, however, it is a good idea to see if there are other factors that might influence the study results. It is entirely possible that there may be some hidden influence on the outcome that is related to age or BMI. We will explore just one statistical detective method that can be used to address this issue.
Instructions
For this discussion, refer to the helpful links in Resources and use the Alaska study’s Emotional Well-Being Corrected data set to perform the following analyses for only three variables that have interval/ratio data: Age, BMI and Baseline SF-36 Scores:
Pearson Correlation
- Assess the selected variables for outliers and normal distribution and report which type of statistical correlation testing would be the most appropriate.
- Create a scatterplot for each selected combination of the above variables to identity the graphic nature of the relationship.
- Perform a Pearson Correlation test on the following, regardless of whether the data distribution looks normal: relationship between Age and BMI, then relationship between BMI and Baseline SF-36 scores.
- Report the results as the magnitude of the relationship (correlation coefficient) and direction of the relationship (positive or negative).
Spearman Correlation
- Perform a Spearman Correlation test regardless of whether the data distribution looks normal for the same two-variable combinations.
- Report the results.
Comparison
- Explain the differences between the Pearson Correlation and the Spearman Correlation, including when to use each test, advantages, and disadvantages of each.
- Describe one or two of the challenges you found while performing these exercises and how you resolved the issues. Where appropriate, provide the address of any website that helped you.
Remember to refer to the guidelines in the FEM as you prepare your post.
Response Guidelines
Read and respond to the posts of your peers according to the guidelines in the FEM.
Address one of more of the following in your response:
- How do the challenges and resolutions of your peers compare to yours?
- How did the comparison between the Pearson Correlation and the Spearman Correlation of your peers compare to yours?
Learning Components
This activity will help you achieve the following learning components:
- Prepare data for analysis.
- Identify the chi-square test of independence.
- Perform a chi-square test of independence.
- Interpret the overall clinical meaning and limitations of the relationship of two variables, based on a correlation analysis and literature regarding age and stress.
- Write about statistical concepts clearly, accurately, and professionally.