2

There are two groups of people, target and neutral

There are two group of events X and others

We're making assumption that people from target group react on event X very different to others.

We have some sets of reaction numeric values on different type of events by those two groups. And after running Student's T-Test in excel I get 0.032 for target group and 0.55 for others so I can tell that target group react on X different but is it possible to measure how much different in 0-100 range?

example of data set (not real data set) for both groups looks like

30, 20, 45, 15, 26 ... - reaction on X event
10, 14, 22, 8, 13 ... - reactions on other events

What can I use to measure how much reaction X different (alike in T-Test) to reaction on other events in 0-100 range and is it possible in general?

cnd
  • 123
  • You could probably look at the sample means and conclude that the difference is statistically significant based on t-test with, i.e., 5% significance. – runr Oct 26 '18 at 07:31

1 Answers1

1

If the data are approximately normal then use a Welch 2-sample t test. This test does not assume that the two groups are sampled from populations with equal variances.

When you mention the 0-100 range, perhaps you mean the P-value which gives the probability of a more extreme difference between groups than observed in your data, if the two groups are really equal (null hypothesis). A common standard is to reject the null hypothesis that the groups are equal if the P-value is less than 5%. Most statistical software computes the P-value of the test. Here is one description of the Welch test in Excel.

BruceET
  • 51,500