Hello all,

I'm implementing a CBC survey on LimeSurvey - so far, so good, with more or less difficulty I'm achieving what I need, after some layout issues I'm getting to the questions' content. What I'd like to do know (since I did not found anyone with similar doubts or anything related in the manual), is if it is possible to use a subset of predefined items (in this case, I have 4 profiles - i.e. 4 answer options - displayed in each array question, from a total of 18 - please see the image attached) in each question. Is this possible? If so, it would be much more easy & professional to edit each profile, instead of hard-writing it on a separate label set for each question.

I hope I could make myself clear, if not please say so.

Best regards & thanks in advance,
João

P.S.: For instance, in the example attached, I have displayed respectively profiles 15, 18, 10 and 3. This was not randomized, it is a predefined combination of 4 of the total 18 profiles. The ideal solution would let me also edit each profiles' item, but that would be another level of complexity. For now, I just would like to know if there is a feasible solution for what I'm looking for.

Perhaps I haven't made myself clear yesterday. All I wanna know is if it is possible to have a list of items (a label set), and then, from that list, pick some of them as answer options for each question. Currently I have to edit all of this answer options manually in each question, which is, at least, unprofessional and inefficient, since the items are repeated many times, and every time I have to change one of them I have to go through all questions and edit them accordingly. These two screenshots may underline what I'm asking here.

I don't see another way then you preparing each "product profile" (and nothing else are those "answer options" you are presenting) separately according to your CBC design.

This design will be done offline anyway and then only once the design is finalized you will upload those profiles in the respective combinations to the respective questions.

Thanks for the answer, holch. Can you be more specific? In fact, I've already have an .xls file with all the 4-profile combinations I need, I'm just copy-pasting it to the label set and adding the respective HTML code. From your answer I'm guessing you're suggesting something else, so if you could elaborate a little more on what you mean by "(...) and then only once the design is finalized you will upload those profiles in the respective combinations to the respective questions" it would be awesome.

No, I think this is the way you will need to do it.

You might be able to make your life a little easier by using Excel to insert the HTML code already and join the text of the cells (em Portgues a formula em exel seria concatenar). this might make things a little quicker and you have the whole concept ready to copy in one cell. But I think this is all you can do.

Ok, holch. I really suspected that LimeSurvey had some kind of "list" feature related to this, thanks again for pointing out how wrong I was, the world won't end anyway. I think it's a good suggestion to implement it on future versions ;]

Yes, it would be great to have a Conjoint specific function, but honestly even big commercial packages are still struggling with this.

Maybe with 2.05 where due to the API it should be easier to implement addons someone will sponsor this.

I know that quite a few universities are users of Limesurvey. Unfortunately I see little coming back from there (maybe just my impression). I think that improving could be a good task for IT students, but I guess so far we have not found a professor to lobby this.

I was referring to the possibility of having lists and (subset lists) of items for subsequent use. The CBC feature would be also great, but that is something even more ambitious.

Regarding your last paragraph, I must confess I know little on that (the "retribution" from universities). I'm a researcher and this is only my second (short) experimentation with LimeSurvey - the tip you mentioned would be of course highly valuable. One thing is certain: it is a great tool, and it is getting better & better everyday!

I am currently not sure how a list feature would actually help you, as each of your profile is somehow unique, but maybe I am just not seeing it.

And nothing against the use of LS in universities. I think it is great that so many universities use it.

However, I would have expected that there would be a little bit more interest in helping to move Limesurvey forward. Especially government funded universities should be able to help here quite well, not necessarily with money, but with assigning tasks that can contribute to Limesurvey to students studying something IT related. Would be not only a practical task that actually make sense, but would help the university overall, with an improved Limesurvey tool. I am sure the developers would be more than happy to implement to the good solutions (of course not all of them would make it to the core) to the LS core.

1. I really think I'm not being able to simplify my explanations, sorry for that. Recapitulating:

1.1. I have 20 items that basically are the different levels of 8 different factors;

1.2. From that 20 items, I have chosen 18 profiles (predefined combinations of 8 different factors' levels) that I find interesting to present to respondents;

1.3. In each screen I will present 4 of these profiles to the respondents, as the previous screenshots demonstrate. These 4-profile combinations are also predefined, as I pointed out before.

As you can see, this would be much more clean, tidy and professional if I had different lists/arrays for each type of combinations: a first list with the 20 items, another one with the 18 profiles (combinations) based on the first one, and a third one with the 4-profile combinations (based on the second list). That way, e. g. if I have to change the phrasing for some items, that would be automatically reproduced in the questions' presented to the respondents, and of course it would save a lot of time and be much more error proof. The way I'm currently implementing this is much more primitive and "handmade", and also easily inconsistent, because I have to change in each question all the items I use there, and they are repeated a lot of times among questions. If I was not clear enough, I think I can make some kind of scheme to illustrate it ;]

2. I understood from the beginning your point on universities, many of them are using LS but few or none are actually investing on using their computer science students to help expanding LS, which would benefit everyone, LS of course, the universities and also the students with more knowledge and a practical way of putting their knowledge "in the field". Even (web) design students could contribute with new templates, etc.