# StatsTTest Values?

What is the P parameter returned by StatsTTest?

I am designing a demo for a class. It shows a "theoretical" normal distribution. It adds "raw data" to the distribution (using gnoise()). I want to return a report on the Accuracy of the raw data to represent the theory.

I understand the hypotheses tests based on t and tcrit. What is confusing me is the P value. Is this a probability of the null hypothesis being true? Or vice-versa? Or something else?

Alternatively, what might be a better message to report on this demo about the accuracy of the "raw data" relative to "theory" (an infinite population size).
jjweimer wrote:
Is this a probability of the null hypothesis being true?

Yes. As for a better message: there's a nice video by Geoff Cummings called Dance of The P-Values, which illustrates the point you are (I think) trying to make. It's on youtube and highly recommended.
Wonderful! Thank you.

--
J. J. Weimer
Chemistry / Chemical & Materials Engineering, UAH
Jeff,

The null hypothesis in this case is that the means are equal. The P value is the probability that 't' is outside the critical values by chance. This is the area under the t-distribution curve away from the +-critical value (for two tails).

The "Dance of the P-Values" is indeed a very nice example that should not be difficult to reproduce in IGOR.

A.G.
WaveMetrics, Inc.
Igor wrote:
...
The "Dance of the P-Values" is indeed a very nice example that should not be difficult to reproduce in IGOR.

Thanks. I am heading that way. My latest iteration shows as in this figure.

I hope to post the demo here soon. Including a way to "build the stats in steps" may have to be something I do as a summer project.

--
J. J. Weimer
Chemistry / Chemical & Materials Engineering, UAH
statsdemo2.png (117.08 KB)
Version 1 of the demonstration has just been posted. See the recent forum posts to find it. The package includes a screen cast of the experiment in action.

I hope folks in education might find this tool can help them explain the concepts of precision and accuracy to students. Indeed, the driving force for me to make it was a realization that students in undergraduate chemistry labs seemed to have no clue what the concepts really mean. Even I learned some important things by the time I was done.

Enjoy!

--
J. J. Weimer
Chemistry / Chemical & Materials Engineering, UAH