This is your Week 6 research log activity for submission.
This week’s SPSS activities have you practicing how to run and interpret correlational
analyses. In Activity #1 (this activity), we look at Pearson’s r (Pearson’s Product-Moment
Correlation) and Spearman’s rho. We are interested in examining the strength and direction
of the relationship between two continuous (usually interval-level) variables.
PAUSE & CHECK UNDERSTANDING
What is a spurious correlation? Why do we care?
HINT: “causation”
1. Using the class dataset, we will:
o run histograms and a scatterplot
o calculate Pearson’s r
o interpret the results.
2. Open your dataset (make sure it contains the AUDIT variable that you created in Week
2).
The variables you will need are:
o AUDIT variable (created in week 2)
o two sub-scales of motivations (social and coping) for drinking from the items
in Q25 (see Week 4 lecture for an explanation of these scales).
In the SPSS data file supplied with these instructions, we have created the three scales for
you.
3. Your first tasks is to MERGE this file with your class data file. To do this, you will need to:
o Click on Data menu
o Select Merge Files, then Add Variables
o Select the other data file
o Choose ONE-TO-ONE MERGE BASED ON KEY VALUES in Merge Method
o Select ID as the key variable
o Go to the variables tab. Move the AUDIT variable in key variables back to the
left. (This variable is already in your orginal data set, so you do not need to
add it to the data).
If you like to see this visually, watch https://www.youtube.com/watch?v=yACxsqiMAQo. It
goes through adding cases as well as adding variables. You will be adding variables.
1
WEEK 6 ACTIVITY #1
Remember to check that your merge was successful. We also strongly recommend that you
save your file under a different name (e.g. add a number to the end of the file name). This
way if it goes wrong, you do not have to re-do everything that you may have already done.
4. Now use graphs to check the assumptions for Pearson’s r. (As well as this week’s
materials, you may wish to go back and look at the materials for Week 4.)
REMEMBER: the interpretation involves judgement about whether any violations are
sufficient enough to not use Pearson’s r. We are looking for distributions and patterns that
are approximately the right shape.
So decide whether the assumptions are violated. (HINT: one relationship will approximately
fit the assumptions).
5. Next we will calculate Pearson’s r for that relationship by:
o Click on Analyze menu
o Select Correlate
o Select Bivariate
o Select and move the relevant variables
o Under Correlation Coefficients, tick the box next to Pearson
o Under Test of Significance, choose Two-tailed
o Ensure that the box next to Flag significant correlations is ticked. Significant
results are flagged with one asterisk * at the p < .05 level, two asterisks ** at
the p < .01 level, and three asterisks *** at the p < .001 level). (Note: APA
formatting requirements require you to report the actual p-value.)
o Click OK
Answer the following questions:
(i) What is the dependent variable? What are the independent variables?
(ii) Is Pearson’s r the appropriate test for all the relationships of interest? If not, why
not?
2
PAUSE & CHECK UNDERSTANDING
Looking at these graphs is one way of checking some of the
assumptions for using Pearson’s r. What are these
assumptions?
PAUSE & CHECK UNDERSTANDING
What are the hypotheses for a two-tailed test?
HINT: presence/absence
WEEK 6 ACTIVITY #1
(iii) For the relationship where Pearson’s r is appropriate to use, is the relationship
statistically significant?
(iv) What is the direction of the relationship?
(v) What is the value of Pearson’s r for this relationship? Is this considered weak,
moderate or strong relationship?
(vi) What is the percentage of shared variance between the two variables? (HINT:
square Pearson’s r).
(vii) Write the result in a sentence.
Use the template supplied with this activity to record your answers to these questions.
6. Let’s know look at how to calculate Spearman’s rho. Spearman’s rho is used when
Pearson’s r is not appropriate. We have one relationship where there was a violation of
Pearson’s r. Using those variables, we will calculate Spearman’s rho to estimate the
strength and direction of the relationship.
7. To calculate Spearman’s rho, we do the same actions as for Pearson’s r, EXCEPT under
Correlation Coefficients you tick SPEARMAN. Again, select two-tailed, non-directional
test of statistical significance.
Answer the following questions:
(i) What assumptions were violated? Explain how?
(ii) Is the relationship statistically significant?
(iii) What is the direction of the relationship?
(iv) What is the value of Spearman’s rho for the relationship?
(v) Write the result in one sentence.
Use the template supplied with this activity to record your answers to these questions.