Step 3: Explore the data
The article actually did not state the sample sizes of the three groups other than they were “approximately equal.” For the purposes of this lab, we will assume the following two-way table of results.
|
Opt in |
Opt out |
Neutral |
Total |
Donor |
23 |
41 |
44 |
108 |
Non-donor |
32 |
9 |
12 |
53 |
Total |
55 |
50 |
56 |
161 |
(e) Calculate and compare the conditional proportions that agreed to be a donor in the three treatment groups (opt in, opt out, neutral). (Use at least three decimal places and remember these are proportions not percentages.)
(f) Open the Multiple Proportions applet in a new window. The data are preloaded. Check the Show Table box and confirm your calculations in (c). Make a screen capture of the segmented bar graph (or toggle to the Mosaic plot) and include in your lab report. Write a brief summary of what you learned about these 161 participants from these numbers and graph.
Thought Question (Think about briefly, discuss with partner, move on)
Suggest a strategy for determining whether the observed association between donation and default option is statistically significant. [Hint: How are you going to get a p-value?]