Hello,

I'm doing a survey for a class and now that I have all the information I need to create a t-test.

Background: Participants used two products and I evaluate them on how long it took them to complete a certain task, and # of errors. There are 5 participants.

How would I create a t-test from this data?

Thanks

 

A t-test for what?

Whether the mean time (for the population) to complete the task is above or below (or equal to) a given value?

Whether the mean number of errors (for the population) is above or below (or equal to) a given value?

It isn't clear what you mean by "# of how many completed". Do you mean that not everyone completed the given task?

You cannot develop a t-test - or, for that matter, any meaningful statistical test - without a clear idea of the hypothesis you're testing.

 

Yes, that's what I'm trying to find out.

Yes. But I'm not going to need a t-test for this.

 

did you check your textbook??

 

What do you know, in toto, about hypothesis testing using t-statistics?

 

Not that much... this isn't a statistics class so I don't have a text book. I need to show a graph that will represent the subjects and their performance.

 

If this isn't a statistics class, why, pray tell, are you expected to perform a t-test? Do they have reason to believe that you've already had a statistics class that covered hypothesis testing?

 

And what would possess you to ask for statistics help in a car forum?

 

Perhaps a feeling similar to those experienced by people who come here seeking legal advice, medical advice, career advice, romantic advice, . . . .

 

Don't do 2 T tests for each variable. You're going to want to subtract the data from one product from the data from the other (for each person). Then, do a T Interval test and create a range of 2-3 standard deviations in both directions for the difference. If 0 isn't included in the range, one product is clearly better than the other. If it is, you can't conclude anything meaningful from the data.

Example:

If one person did it in :22 with product one and :28 with product two, the difference would be :06. You'd do that for each person, and then do a T Interval test.

All of this can be done on any graphing calculator, by the way.