1-cdf value vs chi-squared statistic mislabeled on site?

Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
2 replies · 2K views
nomadreid
Gold Member
Messages
1,779
Reaction score
258
I am confused by the labeling of the (only) graph in
http://en.wikipedia.org/wiki/Pearson's_chi-squared_test
The values on the vertical axis appear to be the values of 1-(the chi-squared cumulative probability distribution values*)
as I would expect in finding the p-values.
However, this graph is labeled the "chi-squared distribution" -- which is ambiguous enough, but when one clicks on this term under the graph, one is directed to a site where the term means the chi-squared probability density function. But the graph cannot be the probability density function: the values are different*, and a probability density function has probability densities on the vertical axis, not probabilities.
So, is this mislabeled?

(*The pdf and cdf graphs were taken from http://en.wikipedia.org/wiki/Chi-squared_distribution#Probability_density_function, and individual values, for more precision, from http://www.danielsoper.com/statcalc3/calc.aspx?id=11)
 
Physics news on Phys.org
The graph isn't actually labeled as the chi-squared distribution. It's merely saying it's a variation of the chi-squared distribution. The density function given on the second page is correct. Chi-squared is just a gamma distribution so it's obvious the first page doesn't show a standard chi-squared. The graph on the first page is just the function of a statistic, not a random variable.
 
ahhhhh. That solves that mystery. Thanks, jwatts.