3 colleagues of mine wish to use an online device/strategy called Qualtrics to collect and input data to be analyzed using SPSS. A 4th must collect the same data by pencil & paper, then put it in Excel to be delivered to our person who is using Qualtrics. Can we compare the input of the paper & pencil guy with the data collected through Qualtrics. Assuming no significant differences can we then combine the 2 data sets into one? Or, is there a good methodological reason why that should not be done?
if all the variables are the same and you have the same scoring criteria, for example likert scale, yes no response or a categorical breakdown, I don't see why not, pencil and paper is just slower, but it should be the same format and nothing wrong with the good old pen and paper I think we are too computerised such that things get too complicated
I would also recommend power analysis between the different methodologies before combining. If all data sets have sufficient power it will help "wash out" the potential errors mentioned above without impacting p values. Additionally, you can download qualtrics data into an excel file to make merging the files easier.
Something to consider--I use qualtrics and p/p methods, it is very easy to randomize scales/items in qualtrics but p/p becomes more complicated so be sure to check for testing/fatigue effects if your survey is lengthy or make several versions of the p/p survey to counterbalance for scale placement within the survey packet.
I agree with the responses here. So you need to create a variable that tags the pen/pencil surveys. This way you can do some data analyses (e.g., bivariate comparison) to compare the two subsets before going forward with analysis using the full dataset. Good wishes with this.
I read all above responses. They are all suitable for your research. Herewith i want to suggest to you can select closed ended questionnaire to collect the data.
Thank you to all who replied. We will proceed as planned with one colleague administering with paper and pencil, then testing to see if the differences are significant.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS