Standard Deviation of Discrete Random Variable

[Luke Setzer]

Does the TI-Nspire have a way to compute the standard deviation of a
discrete random variable?

Given a list x and P(x) I can compute mean(x, px) but not any standard
deviation even though the functions have the same syntax.

[Nelson Sousa]

The standard deviation can be computed the same way as the mean, using the functions stDevSamp (sample standard deviation) and stDevPop (population standard deviation).

Nelson

[Luke Setzer]

x:={0,1,2,3}
px:=(.25,.25,.25,.25}
mean(x,px) 1.5
stDevSamp(x,px) "Error:Domain error"

[Nelson Sousa]

px is the list of frequencie. If the variable is discret the frequencies should be integers.

Both stdevsamp and stdevpop only accept integer frequencies.

Just use integer frequencies (25 each in this case) and it will give you the expected result for stdevpop (stdevsamp depends on the size of the sample).

However, if data is inserted onto the Spreadsheet and you calculate the 1-var stats you get the correct stdev, even with non-integer frequencies. Only on the calculator page you have that extra limitation.

Nelson

[Luke Setzer]

On Oct 15, 11:10 am, "Nelson Sousa" <nso...@gmail.com> wrote:

> Just use integer frequencies (25 each in this case) and it will give you the
> expected result for stdevpop (stdevsamp depends on the size of the sample).

This method worked.

> However, if data is inserted onto the Spreadsheet and you calculate the
> 1-var stats you get the correct stdev, even with non-integer frequencies.

It took me a while to figure that I had to use the one list version of
this function with px as the frequency list, but, by heck and by gosh,
it worked!

Thanks!