Talk:Percentile rank

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Statistics (Rated Stub-class, Mid-importance)
WikiProject icon

This article is within the scope of the WikiProject Statistics, a collaborative effort to improve the coverage of statistics on Wikipedia. If you would like to participate, please visit the project page or join the discussion.

Stub-Class article Stub  This article has been rated as Stub-Class on the quality scale.
 Mid  This article has been rated as Mid-importance on the importance scale.
 

merge with percentile
agree

I disagree, keep it as it is. If someone wants to look up 'percentile rank' then that's what they should be able to do. The explanation is clear and lucid and I really don't see what the problem is. Nick mallory (talk) 08:38, 9 May 2009 (UTC)
I also disagree, keep it seperate as it's a seperate concept. 24.222.205.6 (talk) 15:40, 29 May 2009 (UTC)

A link should be set with the term Quantile and cumulative distribution function. Its contrast with percentile and quantile should be explained. I am not familiar with current term, but note that the function 'percentrank' in excel gives 0% for the minimum value and 1% for the maximum value. In contrast, different formulas exist to calculate the quantile of an observed value. Statistica for instance proposes [(i - rankadj)/(n + nadj)]to calculate the quantile ot the i'th rank. With rankadj = .375 and nadj = .25 as default. 13:16, 29 July 2009 (UTC) Just a humble junior statistician —Preceding unsigned comment added by 194.7.152.194 (talk)

Previous version referred to "cumulative frequency" for the count of scores below the item of interest. But it isn't the cumulative freq. It is a simple count. —Preceding unsigned comment added by 204.210.156.95 (talk) 21:21, 23 June 2010 (UTC)

This text is in error in that it describes percentile ranks as normally distributed and NCEs as having rectangular distributions. As I understand them, Normal Curve Equivalents (NCEs) are a type of standardized score with a mean of 50 and a standard deviation of 21.06. NCEs have a range of one to 99 and in many ways look a lot like percentile ranks -- including having a rectangular distribution. PRs are ordinal and have a rectangular distribution, which is the central problem in their analysis. Please correct me if I'm wrong Andrewt99 (talk) 16:23, 3 August 2010 (UTC)Andrew [andrewt@ucla.edu]

The first two sentences are in direct conflict with each other. Either a score at the 75th percentile is "greater than 75% of the scores" or it is "the same or [higher] than" scores in its frequency distribution. I came here to confirm what I thought I knew, only to be confounded by the contradiction of the two explanations. Please fix it. 64.85.248.115 (talk) 20:24, 3 December 2010 (UTC)