Click on the following links. Please note these will open in a new window.
Correlation: Variables, hypothesis, and non-causation
Your book discusses the conditions under which you would perform a correlation, the hypothesis of a correlation, and the difference between correlation and causation. In this activity, you will apply that understanding.
Go to the following website, read the article, and answer the questions that follow: http://www.everydaysociologyblog.com/2015/04/the-starbucks-effect-correlation-vs-causation.html
- If you were conducting a Pearson’s r correlation using the variables mentioned in the title of the study,
- what would your variables be?
- how would you measure the variables?
- what would your hypothesis be, in words?
- What is a cohort effect?
- What are the two reasons your book provides for why correlation is not the same as causation?
- Think of any two variables (the first two that come to mind).
- Now, pretend you were a researcher and you did find a positive or negative correlation between the two variables you thought of. What might be a third variable that could explain the relation between those two variables?
Your book discusses scatterplots and the information that can be learned from them. In this activity, you will apply that information.
- In a previous chapter, you may have encountered Tyler Vigen’s website on spurious correlations. Go to his website: http://tylervigen.com/discover.
- Using the drop-down menu, select a first IV. Then use the second drop-down menu to select three second IVs. Choose one with a moderate to high positive correlation, one with a moderate to high negative correlation, and one with nearly no correlation. The coefficients are next to the variables. Create charts for each of these. It might be helpful to go to Step 3 and return to Step 2 for each of the variables, or open each chart in a new window or tab.
- Create a data set that includes the four variables of your charts and enter the data.
- Create a scatterplot for each of the variables. Examine each.
- Are there any outliers? (Ones that do not seem to be close to other data points?)
- For the near-zero correlation…
- if you accept the null hypothesis, does that mean there is no relation between the two variables?
- look at the scatterplot. Is there any sort of relation between the two variables?
Interpreting a regression and creating a regression equation using SPSS
Your book discussed how to create a regression equation using SPSS output. In this activity, you will apply this information.
N.B. If you completed Chapter 15 Activity 2, use two of the variables from the data set you created for that activity and skip to step 4.
- In the previous chapter, you may have encountered Tyler Vigen’s website on spurious correlations. Go to his website: http://tylervigen.com/discover.
- Using the drop-down menu, select a first IV. Then use the second drop-down menu to select a second IV. Click Create Chart.
- Create a data set that includes the two variables of your charts and enter the data.
- Run a regression analysis on the data.
- What is the R squared? What does this mean, in words?
- Is the regression significant? What does this mean, in words?
- What is the regression equation? Fill in the two required values from your output.
y = _______ x + ________