Hello

You can, however, it is better to include open questions and report the participant's answers

Why do you visit this particular sites

Hyder Mirghani Hello. I understand what you are trying to say but in my research design, these 9 reasons that I have so far were derived from interviews in the earlier stage of the research.

I am thinking between asking participants to rate the 9 reasons and other reasons they may suggest OR asking participants to rate the 9 reasons, but only filling in other reasons that they may have. What do you think is the best course of action here?

Thank you, if it is like that, I do not think adding other will add to your questionnaire

Yeah having an 'other' option would always be better because you never know your respondents may come up with a reason which is never reported before or you didn't anticipate. In this ever changing world, a new impacting factor may have been arisen to persuade people to visit place 'X'. Therefore, it is imperative to leave an extra space for the respondents to express their own ideas, suggestions and thoughts in addition to what you've outlined for them to choose.

Mohammed A. Mohammed how would including an 'other' option affect my ranking process of all the reasons after data collection. Since not all participants would have ranked the 'other' reason.

Evan Toh Of course you may expect that only few respondents would choose the 'other' option. If that's the case, obviously your ranking process would not be affected significantly but you are expected to tell readers what the respondents have told you in 'other' option. However, if, in case, considerable number of respondents have chosen the 'other' option, you may establish derived variables from open-ended responses. You may consult this blog if you seek to learn how to analyze open-ended response answers. https://www.snapsurveys.com/blog/give-survey-respondents-option-choosing-other-response/

Hi! using other option depends on the type of questionnaire. In certain cases If the question is very subjective then researchers need to interpret and elaborate separately regarding the information collected under "others" option. Thanks

No. What you want is this type of question https://help.surveymonkey.com/articles/en_US/kb/How-do-I-create-a-Ranking-type-question You're asking them to rank the reasons they go to X.

You can then have the next question being something like "Thinking about X, are there other reasons that you like going there?" (edited for grammar -its too early in Ireland at the moment for my brain to work ;)

If you use the format of the question outlined in your question, you are going to end up with 3 groupings -people who LOVE X and 'strongly agree' to all the reasons, people who like it for random reasons (little if any pattern unless your sample is big) and those who like it for some reason that you didn't list.

By using a ranking question you force them to pick 1 top reason, 1 second to top etc.

Hope that helps,

Ingrid

Ingrid Holme is correct if Even Toh's purpose is to have each individual rank his or her reason for attending a place. However, he is the one doing the ranking based on the total responses based on the overall distribution: Evan wrote that "My intention is to compute the means of each reason and rank them based on the highest to the lowest mean value." Ignoring the issue of transforming an ordinal response scale into the assumption of an underlying rational response scale (by assigning discrete numerical values to the text responses and then calculating a mean), Evan's approach would be to rank the strength of agreement with the various reasons from strongest agreement that this was a reason to visit to least strong. That would be an okay approach with the specific reasons given. While an "other" question with an open-ended responses might be added, one should NOT treat that question the same as the specific reasons (e.g., "....it is in a good location") since the "other" responses will have a variety of open-ended responses and those should not be "lumped" together as though each "other" response meant the same thing. Just treat the "other" responses as separate, listing each open-ended response to "other" and the strength of agreement with that response. Also, the ranking should be done separately for each location -- don't try to rank reasons for going there across locations but one could discuss differences in reasons for visiting different locations. Finally, the prompt "...because it is consistent" is poorly written -- consistent how, in what way?

Stephen Maack I agree the issue is not clear cut, however, my suggestion is to bypass the issue of calculating means from Likert scales...which shouldn't really be done unless there isn't another option. ie. https://achilleaskostoulas.com/2013/02/13/on-likert-scales-ordinal-data-and-mean-values/ The main issue is that 'no option' counts for '3' and that we need to assume that the psychological distance between each option is the same (ie. that ‘strongly disagree' and 'somewhat disagree ’ is the same as ‘somewhat disagree' and 'neither disagree nor agree'. We aren't getting into what amount 'somewhat' relates to... ; ) One strong suggestion that I would make is that Evan Toh pilot the questions -and ask people how they found answering them. This will make sure that they are using the same type of 'ranking system' as you expect.

Ingrid Holme, I agree with you on not calculating means from Likert scales, although it is frequently done. While calculating the means would allow the ranking the interpretation of differences among means can lead to specious conclusions. As far as I could tell, Even Toh did not have a "no option" response for each specific category -- which if you mean "no answer" could be included as a missing value and completely excluded from the calculation of the mean. He does have a "neither agree nor disagree" choice if that is what you mean by "no option" and yes, it would be coded as a "3" and that could cause interpretation problems when the mean is calculated. If everyone used the "neither agree nor disagree" response the mean would honestly be "3" -- neither agree nor disagree. If no one chose "3" and there were an equal number of responses of "strongly agree" or "agree" as of "disagree" and "strongly disagree" the mean would move toward whichever was given the value of "5" and "4" -- either the "strongly agree" or "agree" side or the "strongly disagree" or "agree" side. For these reasons, I prefer to sort the grouped responses by the percent who "strongly agree," then the percent who "agree" and discuss the resulting pattern. (That can get tricky if there are different numbers of respondents to each reason for visiting...). Alternatively, if collapsing results, look at the percentage of responses who "strongly agree" or "agree," compared to the percentage who used the neutral "neither agree nor disagree" and the percentage who "disagree" or "strongly disagree." That will give you a better sense than the mean of the strength of agreement that this factor is an important consideration, or disagreement that it was an important consideration. You could discuss the overall results of the sorting of the nine categories of reasons (if that is how any you have) in relation to the strength of agreement, disagreement, or "neither agree nor disagree" (which could be interpreted as either "not important" or as "I don't care about" that factor or as "that factor wasn't a consideration of why I chose to visit that place" -- still ambiguous but perhaps giving more insight into the distribution?). Hope that helps.