Considering BMI as our dependent variable, we could easily consider exercise as our independent variable. It is common knowledge that lack of exercise tends to increase an individual’s BMI, whereas someone who exercises moderately to frequently is more likely to have a lower BMI. The independent variable (exercise) increases while the dependent variable (BMI) decreases. This shows a negative correlation between these two variables. Another independent variable we could consider is an individual’s daily caloric intake. The higher their caloric intake combined with lack of exercise, the higher their BMI is expected to be. The lower their caloric intake in combination with exercise, the lower their BMI is expected to be. The independent variable increasing or decreasing (caloric intake) is associated with the dependent variable (BMI) increasing or decreasing. This shows a positive correlation between these two variables. Bennett et. al. (2018) tells us that the correlation coefficient can be used to assess the strength of a correlation as well as the validity of predictions with the best fit lines. Our lesson this week says that the coefficient of determination helps us determine whether or not our predictions given by the best fit line are precise. The higher the coefficient of determination, the closer our values will be to the best fit line; meaning that our predictions are more precise. If our coefficient of determination for these examples is closer to 1, the more precise our predictions will be and our claim that exercise and caloric intake have a correlating relationship with BMI is supported.
A common medical concern we see today is the effects of smoking on an individual’s risk for heart attack or ischemic stroke. I found an article from the International Journal of Environmental Research and Public Health concerning this very topic narrowed down to Korean males. In this topic, our dependent variables are Ischemic stroke and heart attack, so we have two different studies to compare. Both studies use smoking exposure as the dependent variable. The claim is that the more tobacco smoked the higher the risk for stroke or heart attack on the individual. This would display a positive correlation on a scatter plot graph.
A correlation is defined as a relationship between two variables “when higher values of one variable are consistently associated with higher values of another variable or when higher values of one variable are consistently associated with lower values of another variable” (Bennett et. al. 2018). Causality claims that one variable causes a change in the other variable. Although correlations between variables may suggest causality, the correlation alone cannot ascertain causality. Bennet et. al. 2018 tells us the 3 factors that may be responsible for correlation between two variables: coincidence, common underlying cause, or one variable having a direct influence on the other. In order to begin to establish causality you must first rule out the first two options. This article’s data specifically states that smoking exposure is not causative of stroke or myocardial infarction, but rather the longer smoking exposure the higher the risk those idividuals had for those diagnoses. In other words, long term smoking exposure heightens the risk for one to have a stroke or myocardial infarction in combination with other variables such as health history, age, weight, life style practices etc. Smoking in itself does not guarantee that the smoker will have a stroke or heart attack.
Bennett, J., Briggs, W. L., Triola, M.F. (2018). Statistical Reasoning for Everyday Life 5th Edition.Pearson.
Chang, S., Kim, H., Kim, V., Lee, K., Jeong, H., Lee, J. H., Shin, S. A., Shin, E., Park, M., … Ko, E. (2016). Association Between Smoking and Physician-Diagnosed Stroke and Myocardial Infarction in Male Adults in Korea. International journal of environmental research and public health, 13(2), 158. doi:10.3390/ijerph13020158 Retrieved from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4772178/ (Links to an external site.)