A continuing debate in public health circles concerns the dangers of being exposed to dioxin, an environmental contaminant. A German study published in the October 19, 1991, issue of The Lancet, a British medical journal, considered records of workers at an herbicide manufacturing plant that made use of dioxin. A control group consisted of workers at a nearby gas supply company who had similar medical profiles.
The following data relating to the number of workers who had died from cancer were obtained.
Control group Dioxin-exposed group
Sample size 1583 1242
Number dying from cancer 113 123
Discuss the question: Is there a statistically significant difference in the rate of cancer deaths in the two groups?
Include the following points in your discussion:
- What does “statistical significance” mean in relation to this question
- Why isn’t it enough to say that “since there are more cancer deaths in the Dioxin group with a smaller sample, it is clear that exposure to Dioxin causes a higher probability of dying from cancer”
- Does the word “rate” indicate sample means or sample proportions?
- Construct a hypothesis test for this question
- Perform the hypothesis test analysis for a significance level of 1%
- What is the p-value for this test?