Question

Classify the following bivariate relationships as to whether or not there is staistical association. If so, is the relationship likely to be causal one, or is it due to confounding with some other factor or due to bias.

1. During a recent police strike, the reported rate of new crimes was 20 per 1000 population compared to a normal new crime of 40 per 1,000 population.

2. Carriers of matches experience a lung cancer incidence rate of 80 per 100,000 person-years as compared to 20 per 100,000 person-years among those who don't carry matches.

3. In a survey of students, 50% of the males reported drinking in the last 30 days, while 49% of the females reported drinking in the last 30 days.

please answer 1, 2 and 3 and explain your answer.

thank you

Answer 1

1. The cause could be due to some other factor , not a causal relationship as a police strike does not cause new crimes rate to fall.

2. The relationship is only due to bias and not a causal one. The carriers of matches are more likely to be smokers than those who donot and thus their incident of lung cancer is higher compared to those who don't carry matches

3. The is no statistical association between the % of males or females drinking in the last 30 days. There might be factors such as male and female go to drinking together and thus the proportion of the survey is equal between the genders, there could be selection bias in the survey and a number of other factors involved. Thus a causal relationship cannot se ascertained.