Prev: Setting global InputAutoReplacements
Next: How to get the output of FindFit from MathLink in
From: Lawrence Teo on 20 Nov 2009 06:43
Hi all,

Summary: StudentTCI function's second parameter expects standard
error, rather than standard deviation as stated in the document. The
document has an error.

I just want to report an error in the documentation of StudentTCI
function in Mathematica 7.0.

The document for StudentTCI function reads
StudentTCI[Mu,Sigma, df] gives a confidence interval {min, max} for
the population mean based on the sample mean Mu and sample standard
deviation Sigma.
(refer to HypothesisTesting/ref/StudentTCI)

The truth is StudentTCI function expects Standard Error in the second
parameter, rather than a Standard Deviation.

I was driven crazy by the funny behavior of StudentTCI function for
hours. To make things worse, I didn't suspect it at all and were busy
looking for other mistakes in a rather complex code. I found it in
HypothesisTesting/tutorial/HypothesisTesting. The documentation is
slightly different and clearly points out the second parameter is a
standard error. It was a great relief to me after knowing this
document mistake.

Hope it helps the rest of the Mathematica users.

From: Darren Glosemeyer on 21 Nov 2009 03:34
Lawrence Teo wrote:
> Hi all,
>
> Summary: StudentTCI function's second parameter expects standard
> error, rather than standard deviation as stated in the document. The
> document has an error.
>
> I just want to report an error in the documentation of StudentTCI
> function in Mathematica 7.0.
>
> The document for StudentTCI function reads
> StudentTCI[Mu,Sigma, df] gives a confidence interval {min, max} for
> the population mean based on the sample mean Mu and sample standard
> deviation Sigma.
> (refer to HypothesisTesting/ref/StudentTCI)
>
> The truth is StudentTCI function expects Standard Error in the second
> parameter, rather than a Standard Deviation.
>
> I was driven crazy by the funny behavior of StudentTCI function for
> hours. To make things worse, I didn't suspect it at all and were busy
> looking for other mistakes in a rather complex code. I found it in
> HypothesisTesting/tutorial/HypothesisTesting. The documentation is
> slightly different and clearly points out the second parameter is a
> standard error. It was a great relief to me after knowing this
> document mistake.
>
> Hope it helps the rest of the Mathematica users.
>

Thanks for pointing this out so we can fix it. As a general rule, bugs
should be reported to our Technical Support department at
support(a)wolfram.com to make sure it goes through the proper channels and
gets addressed. In this case, I have taken care of it so there is no
need for a message to tech support if one has not already been sent.

Darren Glosemeyer
Wolfram Research

 | 
Pages: 1
Prev: Setting global InputAutoReplacements
Next: How to get the output of FindFit from MathLink in