# Unpaired t-test is ok in this case?

#### AnaHochmano

##### New Member
Good morning,

I have two groups with the same number of participants (50 participants each group).
Each group is invited to talk about subjects of conversation, each group selects 5 subjects, the subjects are different for both groups and each subject has a priority for each group, so group A talks about the following subjects:
Subject 19
Subject 4
Subject 12
Subject 67
Subject 14
The order of the subjects of conversation is important because for group A the most important is Subject 19, this means that Subject 19 is the subject with the higher priority for them, second Subject 4... and the least important subject for group A is Subject 14.

Group B talks about the following other subjects:
Subject 15
Subject 11
Subject 87
Subject 16
Subject 70
So for group B the most important subject is Subject 15, secondly Subject 11.. and the least important subject is Subject 70

When they talk about each subject they use words, we select the 20 words that they use most frequently and assign a value to each word, values ranging from 0 to 10.

I'm trying to compare the values that each group obtain. I'm trying to show that one group gets higher values than the other group.
It doesn't matter that the Subjects they talk about are different for both groups.

First I'm trying to compare the mean values they get for the subject with higher priority, so I have 20 words for Subject 19 from group A and 20 words for Subject 15 from group B. Each of the 20 words have values, so I have two groups of values ranging from 0 to 10. I'm trying to compare the means of those values to try to show that group A gets higher values than group B.

I think that I should use an unpaired t-test because I have distinct test samples.. is this correct?