One might reasonably ask why we bother with correlations at all if causal relationships are what researchers really are seeking to address?
Correlation does not equal causation! (Huck, 2012; Statistics of Doom, 2016; Lumen, n.d.). This rule is clearly understood, but correlational research is quite useful. First, a brief review of what correlation is. Correlation shows a relationship between two or more variables. As one variable changes, the other variable will change as well. ((Huck, 2012; Statistics of Doom, 2016; Lumen, n.d.).
An example of this was working in performance improvement and measuring waiting time and patient satisfaction. The assumption that patients were happier when the waiting time was shorter was commonly accepted. While these two variables were related, other variables could have been responsible for the satisfaction of the patients. , such as the reception the patients received when they first came in, the engagement of the provider they saw, and the comfort of the waiting area to name a few. But because there was a relationship between waiting and satisfaction, both variables were measured continuously and always; shorter waiting showed a positive relationship with patient satisfaction.
Correlational research is useful to look at the relationship between the variables. This type of research is also crucial in providing quantifiable data (Goss-Sampson, 2019; Statistics of Doom, 2016). The correlation coefficient determines the strength of the relationship between both variables. Scientists use this information to make predictions. For example, when waiting time is longer than 20 minutes, patients are less satisfied. Scientists can also study each variable independently to gain more insight. Correlation research is also useful in making hypotheses. As a previous science teacher, I taught the If…Then scientific method for inquiry. For example, if a slice of bread is left in a moist plastic bag for 3 days, then mold will grow. This hypothesis is based on the knowledge that moisture and heat are excellent mediums for bacterial growth.
Lumen (n.d.) provides an interactive website to test one’s understanding of the correlational relationship. This site offered further clarification of the concept of correlation. This topic is essential to my research as I look at the factors affecting the decision of high school students to attend college. It will be important that I focus on correlations and not mistake them for causal factors.
Before this weeks’ reading, I had limited experience with correlation in data and research. In my experience, K-12 administrators would assume students arriving late to school was related to non-engaged parents; not based on actual data. These students would lose valuable instructional time and thus, demonstrate unsatisfactory academic progress; all on the assumption that parents are non-engaged.
The reading from this week helped me to understand the concept of correlation better. Huck (2012) discussed correlation as showing a relationship between variables; specifically, two variables where the strength and the nature of the association are discovered through data analysis. The idea of correlation showing the relationship is supported by the JASP (2016), where the example of holidays and food consumption are variables that are related because it is family gathering time. Another example provided by JASP (2016) showed the relationship between variables motivation, effort, coffee, and chocolate. These examples helped me to understand positive, flat, and negative correlation better; regarding the directional relationship between two variables.
This last example in JASP (2016) provides the reason why correlation is useful in research. In my opinion, the correlation analysis allows researchers to evaluate the nature of the relationship and other interpretations as to why the link exists. Calculating correlations and analyzing scatter plots assist researchers in interpreting data to support the strength of a relationship. In research, I feel this support inquiry and investigations for developing authentic research findings. For example, the mathematical computations will help researchers express inferences in their correlation statements; and not make the mistake of stating correlation equals cause (JASP, 2016).