How many people will be in the top 10% in at least one examinations

  • Context: MHB 
  • Thread starter Thread starter Tomacork
  • Start date Start date
Click For Summary
SUMMARY

In a scenario where 100 individuals take two correlated exams with a correlation coefficient of 0.6, 10 individuals will be in the top 10% for each exam, resulting in an expected 19 individuals in the top 10% for at least one exam. The calculation for those in the top 10% for both exams is derived from multiplying the correlation coefficient by the number of individuals in the top 10%, yielding 6 individuals. For multiple exams with varying correlations, further calculations are necessary to determine the total number of individuals in the top 10% across all exams.

PREREQUISITES
  • Understanding of correlation coefficients in statistics
  • Basic knowledge of percentile rankings
  • Familiarity with probability theory
  • Experience with statistical analysis tools (e.g., R, Python)
NEXT STEPS
  • Research how to calculate expected values in correlated distributions
  • Learn about the implications of correlation on statistical outcomes
  • Explore methods for calculating probabilities across multiple variables
  • Investigate statistical software packages for advanced correlation analysis
USEFUL FOR

Statisticians, data analysts, educators, and anyone involved in evaluating examination results and understanding the impact of correlation on performance metrics.

Tomacork
Messages
2
Reaction score
0
If 100 people take two exams, the results of which correlate at .6, how many people will be in the top 10% in at least one examinations? I understand that if the exams were unrelated then we would expect one person to be in the top 10% for both groups,and therefore 19 people would be in in the top 10% in at least one exam. However, I don't know how to calculate this when the exams are correlated, nor how to calculate it if there is more than 2 examinations. Any help appreciated
 
Physics news on Phys.org
There will be, of course, 10% of 100= 10 people in the "top 10%" of each exam. If the tests "correlate at 0.6" we would expect 0.6 (10)= 6 people to be in the top 10% in both exams.
 
HallsofIvy said:
There will be, of course, 10% of 100= 10 people in the "top 10%" of each exam. If the tests "correlate at 0.6" we would expect 0.6 (10)= 6 people to be in the top 10% in both exams.

Thanks for that. It is simpler than I thought. I thought that the correlation between the examinations would need to be squared. How would I calculate this for more than 2 exams and with different correlations? For instance, how many people would be in the top 10% in three tests(A, B & C) which correlate at .7 for A/C, .6 for A/B and .5 for B/C?
 

Similar threads

  • · Replies 2 ·
Replies
2
Views
1K
Replies
1
Views
3K
Replies
2
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 4 ·
Replies
4
Views
4K
  • · Replies 18 ·
Replies
18
Views
1K
  • · Replies 7 ·
Replies
7
Views
3K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 5 ·
Replies
5
Views
2K