I am not a statistics expert, so forgive me if this question is too simplistic.
I have the following salary survey for a position:
What is the easiest way to calculate the percentile of the salary within the salary survey for the position?
$\begingroup$
$\endgroup$
2
I am not a statistics expert, so forgive me if this question is too simplistic.
I have the following salary survey for a position:
What is the easiest way to calculate the percentile of the salary within the salary survey for the position?
-
1$\begingroup$ A "simplistic" approach will work well: plot the (value, percent) pairs in Cartesian coordinates, connect them with a nondecreasing fit in any way consistent with your expectations about the distribution of salaries, locate 78,656.06 on the first axis, and read the answer from your curve. Consider drawing several plausible curves so you can assess the accuracy of your estimate from the range of answers those curves give you. If the survey is a random sample, then the curve you draw doesn't have to pass through the data exactly, but it should be within errors expected from the sample. $\endgroup$whuber– whuber ♦2024-02-02 14:51:38 +00:00Commented Feb 2, 2024 at 14:51
-
$\begingroup$ The salary is close to halfway between the 75th and 90th percentiles, so its percentile should be roughly between 80 and 85. The data is not good enough for more precision than that, especially because salary like any surveyed quantity may be defined differently by different survey respondents. $\endgroup$user225256– user2252562024-02-05 20:24:02 +00:00Commented Feb 5, 2024 at 20:24
Add a comment
|
$\begingroup$
$\endgroup$
A percentile basically tells you the % of measures that are below said value. With this definition, the 50th percentile is of course the median (50% of measures lie below; 50% of measures lie above).
To find the percentile of that value you need either the distribution of your data or the entire ordered dataset to estimate where that value lies with respect to the other points. I assume the percentages in the table are percentiles. Then, the only thing you can say so far is that your value is between the 75th and 95th percentile. But how close or how far from these upper and lower limits your value is will depend on how your data is distributed.