I administered the same questionnaire to a sample of 800 but for each item, the wording presented to a participant was one of two variations based on whether they were told that their responses would would be anonymous or public (e.g. "do you feel obligated to sign your name in support of the petition?" v.s. "do you feel obligated to anonymously support the petition?"). This means that in SPSS, there are data for 2 versions of the same item where participants have a response value for one and are missing for the other (depending on their assigned condition).
Using the above example, I would like to consolidate the data for both versions of the same item so that i can analyse "obligation to support petition" as one variable, regardless of the wording. At the moment, "Obligation anonymous version" has 400 responses and 400 missing values, where every case with a missing value has a response on "Obligation public version" instead.
Could anyone please help me with this task? I have a total of 31 items, each with 2 versions of itself.
It seems as though your independent variable for this analysis would be the wording of the question. You have to different groups within this independent variable (being told that their answer is private vs being told that their answer is public). Owing to this, you should create a new variable on SPSS and you can call it something along the lines of 'type of question'. Then you can give each participant an answer of either 1 (public) or 2 (private).
This will then allow you to analyse your data through these nominal variables.
Hi Dara,
Thank you for your help. Wording of the question is one independent variable true but it isn't my focal independent variable at the moment. I have another, which is my focal IV: people were exposed to an anti-protest or pro-protest message. I would like to compare the effect of an anti-protest message to a pro-protest message on participants' responses to items in a questionnaire (including the "obligation to support" example item with a 7-point response scale). But before i can do that I need to have a single "obligation to support" variable and currently my data is split across two because i had two differently worded versions.
Do you know how to merge my data for these two versions in SPSS so that i can treat "obligation to support" as a single variable, regardless of the fact that participants received slightly different wordings? (I will be exploring the effect of the wording in later analyses)
....I have attached a section of my SPSS dataset to illustrate my problem. 888 refers to missing values. I would like to create a 3rd variable ("obligation to support") that merges the two variables present so that (effectively) there are no 888s. I need to do this for a lot of other variables too.
Thank you,
Nathan
Hi Nathan, I think i know what you are talking about now. This shouldn't be hard to resolve, however, it would be easier if I could see the screenshot, unfortunately I don't think it has been uploaded correctly as it will not load.
I have attached a screenshot! I have only taken a small section so its easier to understand/talk about/illustrate
Responses are on a 7-point scale, 1=strongly disagree and 7=strongly agree.... 888=missing
You can combine the two variables together, but i would recommend changing them from categorical to numerical values. To do this you will have to transform each answer into a number between 1-7 (as with your scale). The numbers can then be labeled in variable view using the 'values' box. Given that you have 800 data sets, this would take a lot of time to do manually but you can give SPSS instructions to do this for you:
Simply click on Transform > Recode into different variables> click the variable you want to change and move it over> and then click on 'old and New Variables'> Then you can type in the name of the variable you want to change (e.g. strongly agree) and type in what you want to change it to (e.g. 7 ).
Once you have done this for both sets of variables, you can merge them into one:
Click Transaform>Compute Variabl..
In the top left box, type the name of the new variable (e.g. total obligation). Then you can tell SPSS to add each variable together to give you a total score. Alternatively you can use the calculator to make SPSS give you an average overall obligation score for each participant by adding the variables together (in brackets) and dividing them by 2
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS