DingDongD

5+ Year Member
May 18, 2013
115
6
Status
Medical Student
I have a question on this KA passage:
https://www.khanacademy.org/test-prep/mcat/social-sciences-practice/social-science-practice-tut/e/social-conflict-theory-in-an-historical-context

It mentioned this before:

Table 111 outlines the correlation between social context and positive drug test in arrestees. The odds ratio describes how much more likely a positive cocaine/opiate urinalysis becomes when the social control factor is increased.
Table 111
Social control factor or structural disadvantage Odds Ratio
Male 0.74300
White 0.52690
Full-time employment 0.59390
High school graduate / GED 1.0850


The question is:
According to the data in Table 1, which factor has the highest probability of testing positive for cocaine or opiates?
Please choose from one of the following options.
  • Having graduated from high school / having a GED
  • Being White
  • Being non-White
  • Not having graduated from high school / not having a GED

    Can someone explain why the answer is "being non-white"?
 
Jan 26, 2015
477
460
Status
MD/PhD Student
Check the hints for the question. If being white decreases the probability of testing positive the most, then being non-white would increase the probability the most, if that makes sense.
 

SouthernDoc2Be

5+ Year Member
Jul 18, 2012
167
113
New Orleans, LA
Status
Pre-Medical
I have a question on this KA passage:
https://www.khanacademy.org/test-prep/mcat/social-sciences-practice/social-science-practice-tut/e/social-conflict-theory-in-an-historical-context

It mentioned this before:

Table 111 outlines the correlation between social context and positive drug test in arrestees. The odds ratio describes how much more likely a positive cocaine/opiate urinalysis becomes when the social control factor is increased.
Table 111
Social control factor or structural disadvantage Odds Ratio
Male 0.74300
White 0.52690
Full-time employment 0.59390
High school graduate / GED 1.0850


The question is:
According to the data in Table 1, which factor has the highest probability of testing positive for cocaine or opiates?
Please choose from one of the following options.
  • Having graduated from high school / having a GED
  • Being White
  • Being non-White
  • Not having graduated from high school / not having a GED

    Can someone explain why the answer is "being non-white"?
Check the hints for the question. If being white decreases the probability of testing positive the most, then being non-white would increase the probability the most, if that makes sense.
I'm also stumped on this question. I've looked into understanding odds ratio and understand that it is important to consider the reverse of the factors but I am also reading that if the OR > 1, then that scenario is related to higher outcomes. So, is that saying that high school graduates test positive for cocaine/opiates more than non-graduates? That seems wrong...?

From NCBI:
  • OR=1 Exposure does not affect odds of outcome
  • OR>1 Exposure associated with higher odds of outcome
  • OR<1 Exposure associated with lower odds of outcome
Please explain.
 

aldol16

2+ Year Member
Nov 1, 2015
4,912
3,436
Status
Medical Student
You would have to understand what an odds ratio is, which is a concept that usually is not taught in introductory statistics. The rules you quote are correct - odds ratio of <1 is "good" in that more of X equals less risk of Y. Odds ratio > 1 is the opposite and odds ratio = 1 doesn't matter. So basically, according to the data, graduating from high school or not results in almost no relationship with positive cocaine test (OR = 1). Being white reduces the risk of testing positive for cocaine so of the factors given, only being non-white would increase the risk (the converse of white reducing the risk).
 

SouthernDoc2Be

5+ Year Member
Jul 18, 2012
167
113
New Orleans, LA
Status
Pre-Medical
You would have to understand what an odds ratio is, which is a concept that usually is not taught in introductory statistics. The rules you quote are correct - odds ratio of <1 is "good" in that more of X equals less risk of Y. Odds ratio > 1 is the opposite and odds ratio = 1 doesn't matter. So basically, according to the data, graduating from high school or not results in almost no relationship with positive cocaine test (OR = 1). Being white reduces the risk of testing positive for cocaine so of the factors given, only being non-white would increase the risk (the converse of white reducing the risk).
@aldol16 thanks for the explanation--just a follow up.

I think I'm letting my biases interfere with understanding the math principles but--if I were to consider just the social factor of being a high school grad/having a GED which has an odds ratio of 1.08 and were to compare that to it's inverse which would be not being a high school grad/not having a GED, surprisingly being a high school grad would have a higher probability that one would test positive for drug use. This is true since the OR>1, correct?

Again, I'm just trying to understand the math here.
 

aldol16

2+ Year Member
Nov 1, 2015
4,912
3,436
Status
Medical Student
I think I'm letting my biases interfere with understanding the math principles but--if I were to consider just the social factor of being a high school grad/having a GED which has an odds ratio of 1.08 and were to compare that to it's inverse which would be not being a high school grad/not having a GED, surprisingly being a high school grad would have a higher probability that one would test positive for drug use. This is true since the OR>1, correct?
According to the data given, I believe that is correct if the error is small. However, you should keep in mind that unless you're working in a physical discipline where you are limited only by the accuracy of the machine performing the measurement, you will never get a perfect result. That is, you won't ever get OR = 1. There's just too much noise in a chemical or biological system. So I think 1.08 is within the acceptable bounds (depending on the confidence interval) to say it's "1." In which case it would be no effect either way.