Hi all, I am doing a survey where I have a two-part question.
In the first part, I ask respondents to select from a list of 10 reasons why they have visited this place. In the second part, I ask respondents to rank the importance of the reasons they have selected in the first part. As such, this will mean partial rank order data is collected as respondents will certainty not choose all 10 reasons. For example, if respondent A chooses 5 reasons in the first part, he/she will only rank the 5 reasons he/she has selected.
Therefore, my question is, what kind of analyses am I able to run on SPSS for such data? My rationale for setting the question is to understand the reasons why people visit the place, and which reasons are the most important. I believed that there is no point ranking the reasons they did not select in the first part of the question.
Appreciate all the help!
Hi! While I'm not able to answer your question specifically, let met throw some (unsolicited, I know) ideas back at you. As far as I understand your research design, respondents can say yes or no to items 1 to 10 , and then rate whichever amount of items they selected. This complicates your study, in my opinion, and I believe you could change these items from dichotomous to scalar.
If every participant was required to rate these "reasons to visit" from, say, 1 to 5, you'd already have a decent amount of information. Not only if this reason is important at all, but how important it is. Also, this "likert-like" outcome is probably better to model in terms of distribution and analysis. Hope this helps someway!
@Breno Marques so based on my understanding of what you suggested, let's do an example. If a respondent chooses 5 reasons out of the 10 listed, you are saying that it would be good for them to rate those reasons from 1 to 5 on a likert scale?
How would this likert scale look like? If it was from 'Not important at all' to 'Extremely important', my concern would be that respondents will most likely choose 3 and above since it is one of the reasons they have selected, making for skewed results.
One example would be: rate the following reasons for visiting place X
A - Because it's beautiful
1 - Not important at all
2 - Some little importance
3 - It's important
4 - it's very important
I'm being extremely generic, but what I'm suggesting is some form of spectrum for the importance/likelihood of these reasons. By forcing 4 items, you're forcing either negative or positive answers in likert-like scale.
One alternative to reduce this response bias towards item selection could be forcing respondents to rate every reason in a spectrum of importance. Following your first question, this would mean your scale would have 10 items with 4 possible answers to each item, in a scale of importance, likelihood or however you'd like to frame it. I believe this way you can reduce the complexity of analysing data from different items in every subject's answer.
- Badges
- Science topic
- Similar topics
- Philosophy
- Philosophy Of Science
- Reasoning