There have been arguments about the validity ofidentification based on DNA match
ID: 2951506 • Letter: T
Question
There have been arguments about the validity ofidentification based on DNA matching in criminal cases. One problemis that different subgroups may have different frequencies ofalleles. What's rare in one group may be common in another. Someempirical work has been done, to measure differences amongsubgroups. One geneticist said: Statistical significance is an objective,unambiguous, universally accepted standard of scientificproof. When differences in allele frequencies among ethnicgroups are statistically significant, it meansthat they are real--the hypothesis that geneticdifferences among ethnic groups are negligible cannotbe supported.Comment briefly on this interpretation ofstatistical significance.
I thought that maybe this geneticist iswrong in saying that statistical significance is not completelyuniversally accepted because some statisticians have differentstandards of exactly what they consider statisticallysignificant.
Any help would be greatly appreciated! Thanks!
One geneticist said: Statistical significance is an objective,unambiguous, universally accepted standard of scientificproof. When differences in allele frequencies among ethnicgroups are statistically significant, it meansthat they are real--the hypothesis that geneticdifferences among ethnic groups are negligible cannotbe supported.
Comment briefly on this interpretation ofstatistical significance.
I thought that maybe this geneticist iswrong in saying that statistical significance is not completelyuniversally accepted because some statisticians have differentstandards of exactly what they consider statisticallysignificant.
Any help would be greatly appreciated! Thanks!
Explanation / Answer
Yes, to some extent, statisticians will differ, although it really depends on the numbers. For example, one statistician may use = .05, while another may use =.1. If, however, the numbers clearly show that the null hypothesis is supported, regardless of the choice of (within reasonable limits), then it can be assumed that the null hypothesis is correct. Statistics is not a 100% exact standard of scientific proof, but it is a useful tool to show with reasonable certainty that something is correct or incorrect. What's nice about statistics is that you can also calculate how certain you are about your conclusion. If, for example, you are 99% sure about your conclusion, then you can pretty much say with reasonable certainty that your conclusion is correct.
Related Questions
drjack9650@gmail.com
Navigate
Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.