Section 8.3 Investigation 2.3: Readability of Cancer Pamphlets
Exercises 8.3.1 The Study
Researchers in Philadelphia investigated whether pamphlets containing information for cancer patients are written at a level that the cancer patients can comprehend. They applied tests to measure the reading levels of 63 cancer patients and also the readability levels of 30 cancer pamphlets (based on such factors as sentence length and number of polysyllabic words). These numbers correspond to grade levels, but cancer patient reading levels below grade 3 and above grade 12 were not determined exactly.
The following tables indicate the number of patients at each reading level and the number of pamphlets at each readability level:
Patient’s Reading Level
| Patient’s reading level | < 3 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | > 12 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Count | 6 | 4 | 4 | 3 | 3 | 2 | 6 | 5 | 4 | 7 | 2 | 17 |
Pamphlet’s Readability Level
| Pamphlet’s readability level | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Count | 3 | 3 | 8 | 4 | 1 | 1 | 4 | 2 | 1 | 2 | 1 |
2. Limitation of calculating mean.
Explain why the form of the data does not allow you to calculate the mean reading level of these cancer patients.
3. Median reading level for patients.
4. Median readability level for pamphlets.
5. Comparing medians.
How do these medians compare?
- The medians are the same (both 9th grade), so the values are very close.
- Correct! The medians are identical.
- The patient median is higher than the pamphlet median.
- Incorrect. Check your calculations for both medians.
- The pamphlet median is higher than the patient median.
- Incorrect. Check your calculations for both medians.
- The medians are several grades apart.
- Incorrect. Calculate both medians carefully.
6. Interpreting closeness of medians.
Discussion.
Keep in mind when examining quantitative data, that you should start by constructing some simple graphs to explore the data. In fact, you should use technology to explore a couple of different graphs (e.g., change the bin width in a histogram) to help reveal hidden patterns and unusual observations.
Study Conclusions.
When these researchers came to the statistical consultant, they wanted a p-value for comparing the mean reading level of the patients to the mean reading level of the pamphlets. You will learn about such a test in Chapter 4. However, the statistical consultant replied that not only couldn’t means be calculated, but that a simple look at the data revealed a substantial percentage of the patients with a reading level below that of the simplest pamphlet. This "descriptive" analysis was sufficient, rather than looking for a more complicated "inferential" analysis involving p-values and confidence intervals.
Subsection 8.3.2 Practice Problem 2.3
Checkpoint 8.3.1. Estimating tax revenue.
If city officials wanted to estimate total tax revenue for a city (how much money people will pay in based on their incomes), would you rather know the mean income or the median income? Explain.
Checkpoint 8.3.2. Estimating housing cost.
If you want to estimate the typical housing cost for a new city, would it be more helpful to know the mean housing cost or the median housing cost? Explain.
Checkpoint 8.3.3. Research question focus.
In the Cancer Pamphlet study, explain why a comparison of the centers of the distributions does not match the research question of interest.
You have attempted of activities on this page.
