The t-distribution table - how to use

How to use the t-table

What is the t-distribution table?

The t-distribution table lists the point on the x-axis of the t-distribution beyond which 5% of the surface area of the curve lie. We use it every time we run a t-test (also called "Student's t-test").

What is the t-distribution?

The t-distribution is a curve like the normal curve or the F-curve. It is actually a modified t-curve that takes into account how many subjects we have in our experiment.



A hypothetical experiment.
An experiment has 2 groups of subjects, 6 subjects in each group, 12 subjects total. Group 1 received a placebo (an inert substance). Group 2 received a drug. The temperature of each subject was recorded. We have 12 scores, 6 scores in each group.

Step-by-step guide

Step 1. We calculate the means, mean 1 and mean 2.

Step 2. We calculate the degrees of freedom, df, df=10. How do we calculate df? df=total number of scores minus the number of means. In this case 12-2=10.

Step 3. We run the t-test for independent groups and, suppose, we find that t=4.52. We call this "the calculated t".

Step 4. Now we go to the t-table and enter the left column at df 10 We now slide our finger to the right as far as column with the heading 95% two-sided (also called two-tailed). The number we see here is 2.228. We call this "the required t" (also called "critical value").

Step 5. We now compare our calculated t to the required t, i.e. the t in the table. Calculated t=4.53. Required t=2.22.
The calculated t is greater than the required t, i.e. the table t.

We conclude that we have significance. This means that the difference between mean 1 and mean 2 is significant (loosely speaking that it is real and not a chance event). We report this as follows:
"p<0.05" . P less than point o five.

What does "p<0.05" mean?
It means that the probability that your finding is due to chance (and not to what you did to the subjects of your experiment) is less tan 5%, 5 in a hundred.

What does "p<0.01" mean?
It means that the probability that your finding is due to chance (and not to what you did to the subjects of your experiment) is less tan 1%, 1 in a hundred.

What does "p<0.001" mean? It means that the probability that your finding is due to chance (and not to what you did to the subjects of your experiment) is less tan 1 in a thousand.