Hygiene: Separate Measures from Claims

Distinguish a reported behavior, a measured outcome and an inference.

All worksheets

5 questions0m 0s

Learn the skill

An event-rate chart needs an outcome definition, denominator and time period. Self-reported handwashing is not the same measurement as observed illness. Nonresponse and different measurement methods can bias comparisons.

Worked example

In a separate fictional survey, 48 of 60 respondents report a routine, while 100 people were invited. That is 80% of respondents, not proof that 80% of all invitees followed it. No personal reports are collected in this worksheet.

Invented study data show 20 events among 100 people in group A and 10 among 100 in group B over the same period. The difference is 10 percentage points; the relative reduction from A is 50 percent. This is not a product claim.

Would you believe it?

Cleaning removes germs along with dirt. It can help even without killing every germ; cleaning and disinfecting are different tasks.

CDC: cleaning and disinfecting

Question 1 What percentage of the 60 respondents reported the routine? 60% 80% 48%

Question 2 How many invited people did not respond? 40 52 20

Question 3 May we assume the same rate among nonrespondents? Yes, missing means identical Only if we hide the missing count No, they may differ

Question 4 Is reported handwashing the same outcome as illness incidence? Only at a small sample size No, they measure different things Yes, the words are interchangeable

Question 5 Which report is most transparent? State the denominator, missing data and limitations Show only the largest percentage Remove the study period

Review your work

0correct

Complete the questions to review each answer.

View dashboard