Confused about my data file

#1
I have conducted a pilot study on two individuals (Likert data). These individuals completed a survey, every three days for 8 weeks. Their data looks like this:

subject Q1 Q2 Q3 Q4
1 3 3 4 2
2 4 2 4 4
2 5 2 1 2
1 3 2 1 4

edit: these columns lined up when I created the question but dont seem to when I post it.

And so on (hypothetical data here).

What I want to know, among other things, is whether the scores on Q1 differed significantly from those on Q2 (basically I want to compare all four Q's and look for significant differences).

At this stage, I am not interested in gender or even comparing subjects, I just want to know whether the overall scores differed on these questions.

Now, my first choice would be to just do a paired samples t test but that seems wrong, because that is supposed to involve a number of different participants, not just 2 over time.

Any suggestions on how best to tackle this would be appreciated.
 
#2
hope this helps

I think you should do the normality test for each and then if all are normal you can proceed with the t-test ( Anova) and if not the non parmetric test ( Kruskal Wallis ) is a good test for significance.
 
#3
So you are saying, assuming the distributions of each question are normal, I can do paired samples t-tests to compare these 4 questions? Despite the fact that there are only two participants and all the data in say, question 1, is just responses over time from the two participants?
 
#4
I think I have figured it out. Quite simply, I think my data file is set up wrong and I need to change it around. I can't have multiple responses from one subject going vertically down a variable column, as SPSS interprets that as being a separate participant. I need to change my data file so all data from one participant is on one line, with variables called Q1_day1 , Q1_day3 etc. Then I can do repeated measures to compare them. Does this make sense?
 
E

elnaz

Guest
#5
Hello
the first you cant use paired t -test because you have 4 question then you should do repeated measure that you can see it in SPSS but you attention to normality of data
if are normal you use repeated measure if not you use friedman test .
i hope you be succeed
Best regards
Elnaz