Research Methods, Statistics, and Applications
Chapter 8: Examining Relationships among Your Variables
Journal Article 1: Conn, V. S. (2017). Don’t rock the analytical boat: Correlation is not causation. Western Journal of Nursing Research, 39, 731–732. doi:10.1177/0193945917701090
Learning Objective: 8-1: The advantages and limits of correlational designs.
Summary: This is an editorial about mistaking correlation and causation.
Questions to Consider:
- According to the author, what are some of the reasons that researchers might misinterpret correlational results as causal results?
- What are some of the benefits of correlational research that the author identifies?
- What are some solutions to address the problem of mistaking correlation and causation in published research and reviews?
Journal Article 2: Eakman, A. M., Schmid, A. A., Henry, K. L., Rolle, N. R., Schelly, C., Pott, C. E., & Burns, J. E. (2017). Restoring effective sleep tranquility (REST): A feasibility and pilot study. British Journal of Occupational Therapy, 80, 350–360. doi:10.1177/0308022617691538
Learning Objectives: 8-1: The advantages and limits of correlational designs. | 8-2: How to distinguish between correlational design and correlation as a statistic.
Summary: This is a pilot correlational study on a sleep intervention.
Questions to Consider:
- Why did the authors conduct this pilot correlational study?
- Did the authors use any correlational statistics (Pearson’s r or point biserial r) identified in chapter 8 of the text? Why might this be?
- What are the results and implications of this study?
Journal Article 3: Johnson, R. R. (2012). Police officer job satisfaction: A multidimensional analysis. Police Quarterly, 15, 157–176. doi:10.1177/1098611112442809
Learning Objectives: 8-1: The advantages and limits of correlational designs. | 8-3: How to compute and interpret the statistics assessing correlations between variables of different measurement scales. | 8-4: How to predict scores of one variable based on another.
Summary: This is a study examining predictors of police officer job satisfaction.
Questions to Consider:
- Why is a correlational design the right design choice for this study?
- In Table 2, the authors report Pearson’s r correlation coefficients. In which cases is this the correct analysis, and in which cases should they have referred to the analysis as a point-biserial correlation?
- What was the purpose of conducting the regression analyses?