Suppose you want to know if there is any association between water pollution and gastritis.
Think about how you would go about organizing this statistical study between water pollution
and gastritis. What steps would you have to take/what items would you need to include in
order to get it off the ground?
1) In your statistical study what is your:
Population - specific geographic region or community where you suspect water pollution
might be an issue and where gastritis cases are prevalent.
Parameter - could be the strength or nature of the association between water pollution and
gastritis, which might be quantified as a correlation or odds ratio.
Sample-
individuals or locations that have varying levels of exposure to water pollution.
Sample Statistics - data on both water pollution levels and gastritis cases in the sample
Sampling Methods:
There are various sampling methods to choose from, and the choice depends on the specific
research question and practical considerations. In this case, the most appropriate sampling
method may be Stratified Sampling or Cluster Sampling:
2) Given the 4 sampling methods that we have discussed, which one makes the most
sense in your study?
Stratified sampling
3) Why did you choose this method?
This method is a better choice if you suspect that certain subgroups within the population
may be more or less affected by water pollution
4) Let’s say that polls show that 52% of voters plan to vote for Candidate A in the
upcoming election with a margin of error of 4%. Can you safely predict that Candidate
A will win the election? Explain your answer.
No, you cannot safely predict that Candidate A will win the election based solely on the poll
showing 52% support with a margin of error of 4%. The margin of error represents the range
within which the true proportion of voters who support Candidate A is likely to fall.
In this case, the margin of error is 4%, so the true proportion of support for Candidate A is
likely between 48% and 56%. There is still a chance that Candidate B could win the election
if the actual voter support falls within this range.
