The Student Room Group

Inputting data to SPSS

This is the point where I really wish our course had done more on statistics instead of the one semester in the 4 years that we actually got :frown:

Ok so I now have all my questionnaires back and I am ready to input the data into SPSS, and this is where I have hit a brick wall!

I have 4 different questionnaires and 160 participants, who completed all questionnaires.

Do I 'score' the participants questionnaires first and then only input the scores into SPSS or do I need to put in all the raw data too? If the raw then how do I organise this within SPSS? Some of the questionnaires for example the Social Desirability scale are just True / False options whereas the others are scale ratings from 1-9.
Reply 1
It would probs be easier to score the questionnaire separately, so you'd have columns

Participants | Questionnaire 1 | Questionnaire 2 | Questionnaire 3 | Questionnaire 4

then in the variable view for the questionnaire 1 -if that had true/false answers you would code 0 as true 1 as false (or whatever), then input either 0 or 1 for what answer they gave.

For the questionnaire with scale ratings, you'd do 1=not ever, 2= sometimes, 3=unsure, 4=often, etc for whatever the scale is, then put either 1-9 for whichever answer they gave.

There is a tutorial for inputting and giving variable options.
Reply 2
Thanks, I'll have a look at the tutorial.

The bit thats confusing me though is that obv each questionnaire has a number of individual questions within it. So if I have columns as you suggested and coding then each questionnaire only has one column so how do I put in the answer to each individual question?
Reply 3
Oh right, for some reason I was thinking you were putting the total scores from each questionnaire in for each participant.

Separate data sheets for each questionnaire? (or put them all on one) with the same idea as before but instead of columns for questionnaire 1,2,3,4, have question numbers?

I think what you're getting at is this:

Participant_Question 1____Question 2____Question 3 __Total
1,_________True (1)______False (0)_____True (1)_____2
2._________False (0)_____False (0)_____False (0)____0


etc? That's the only way I can think of it atm.
Reply 4
squeak
Oh right, for some reason I was thinking you were putting the total scores from each questionnaire in for each participant..


Well, yes I was thinking of doing that. But I dont know if that is what I should do :s-smilie:

squeak

Separate data sheets for each questionnaire? (or put them all on one) with the same idea as before but instead of columns for questionnaire 1,2,3,4, have question numbers?

I think what you're getting at is this:

Participant_Question 1____Question 2____Question 3 __Total
1,_________True (1)______False (0)_____True (1)_____2
2._________False (0)_____False (0)_____False (0)____0


etc? That's the only way I can think of it atm..


I think I understand that. But that would only work for the one questionnaire, right? Because the questionnaires dont all mean the same thing, and are asking about different issues :s-smilie:

Basically I have 160 responses. I sent it out as one big questionnaire but in actual fact it contains 4 seperate measures - one for emotional eating, one for perceived stress, one for defence mechanisms and one for Social Desirability. At first I thought I would go through each participants questionnaires and score them on each seperate measure then input each score / result in SPSS... but now I'm thinking thats NOT what I should do and I need to put each participants 'response' to each question on each questionnaire into SPSS.

So what I have at the moment is something that needs to look a bit like this

Participant / Question / EES / PSS / DSQ / SD
1_______ / 1______ /_____ /____ /_____ / true (1)
1_______ / 2 ______ / _____/____/ _____/ false (0)
1_______ / 3 _____ /______/____/_____ / false (0)
2_______ / 1 _____ /______/____/_____ / false (0)
2_______ / 2 _____ /______/____/_____ / true (1)

But then that doesn't make sense either because the questionnaires dont all have the same number of questions and their totals dont mean the same things obv so it may look like missing data...

Oh god, I've just looked at that my mind is boggling again. I think I'll go to bed and try again in the morning! Thanks for your help, I'm sure its jus me being very dim!!
Reply 5
Ok, thought about it again and its more likely to be

Participant / EQ1 / EQ2 / EQ3 / EQ4 etc / PSS.Q1 / PSS.Q2 / etc / DSQ.1 / DSQ.2 / ....

??
Reply 6
Yeah I was saying you could do a spread sheet for each questionnaire.

I recently used the OLife questionnaire which assesses schizotypy but consists of 4 categories unusual experiences, cognitive disorganisation, introvertive anhedonia, and impulse non conformity.

I had a scoring spread sheet on excel (though sure you could do it on SPSS) where you inputted a participants scores and it gave you a total for each category and a total score, then I inputted the total scores for each category into SPSS so I had:

Participant___UE___CD___IA___INC__Schizotypy Score
1.__________5_____6____2____8______21
2.__________6_____1____1____4______12
3.__________5_____3____1____6______15

I'm not sure if you do have four questionnaires or you're just counting each measure on the one questionnaire.
If it is only the measures, and you only need to know the totals of each measure and not what was scored on each question I would only input the totals.

The example you gave it difficult to understand and wouldn't work, as it makes out that you have 3 participant 1's. If you had to set it out like this, it would be better to do:

Participant__Question 1__Question 2__Question 3__EES__PSS__DSQ__SD
1.
2.
3.

But I don't really understand what you have, do you have one questionnaire with four categories that are scored using different scales?
Reply 7
_amanda86
Ok, thought about it again and its more likely to be

Participant / EQ1 / EQ2 / EQ3 / EQ4 etc / PSS.Q1 / PSS.Q2 / etc / DSQ.1 / DSQ.2 / ....

??


Yeah that makes more sense.
Reply 8
Two choices, which you use depends on what you want to do with the data.

You could do what you've outlined (i.e., input scores for each item/question, then a total score for each questionnaire) or you could just enter the totals alone.

Knowing you want to perform multiple regression (or so you suggested in t'other thread) you could just enter the totals - would do the job. If you want to run reliability analyses on the questionnaires, then enter the individual items and totals (which would be wise).
Reply 9
Thank you!

I've went for the individual items and totals. Set it up... now just need to enter all the data! (and hope its right / it works).

Fab way to spend a Saturday night, eh? :p:

Latest