RELEVANT MAX-DIFF
MaxDiff is a survey research approach forobtaining preference/importance scores for multiple items (brand preferences,brand images, product features, advertising claims, etc.).
Sometimes researchers want to show only theMaxDiff items that are relevant to a given respondent.
Relevant max diff analysis let us to showa customized and dynamic list for each respondent based on a variety ofconditions and logic. This means we can customize the MaxDiff items(potentially a different number of items per respondent) to show respondentsbased on relevance, awareness, importance, preference, etc.
Example link:
https://knowledgeexcelsurvey.com/relevantMaxdiff/cgi-bin/ciwweb.pl?studyname=relevantMaxdiff
Analysis can involve logit, latent class,or HB and you need to specify in the analysis interface how to treat missingitems for the respondent. There arethree methods of treating missing items:
1.Missing items are missing at random andinferred from the population information (default method),
2.Missing items are inferior to includeditems,
3.Missing items are unavailable and shouldhave near-zero probability of choice for a respondent (we insert a largenegative utility such as -20 for the missing utility).
All three methods lead to a full set ofutility scores for all the items (for the group, each latent class, or therespondent if using HB estimation). Ifusing HB estimation, you can also export the scores with missing items perrespondent replaced with blanks, such that the missing items have missingutility scores.
There are two ways to create thecombinations of items to show respondents (the design) for MaxDiffquestionnaires: Pre-generated and On-the-fly.
On-The-Fly Design Generation
On-the-fly (during survey) MaxDiff designscontain the combinations of items to show each respondent and are generated inthe moment that the respondent takes the survey. On-the-fly designs are required when the numberof items in the constructed (dynamic) MaxDiff list is different acrossrespondents for the same MaxDiff exercise.
We should take special care thatrespondents always get a reasonable number of items in their considerationlist. If too many items are included,each item might not be shown enough times for stable HB estimation. If too few items are included, there may betoo few items to populate even one MaxDiff question and the entire MaxDiffexercise would be skipped for the respondent. We should ensure that our consideration set has a minimum number ofitems per respondent.
Researchers also recommend to add a fewitems to the consideration set that are not relevant/important to therespondent. However, given the method weuse for augmenting each respondent's choice data with full information whenmissing items are treated as inferior, this may not be necessary
Pre-Generated Designs
When the number of items to show eachrespondent is the same, we can generate the combinations of items to showrespondents prior to data collection. This is called a Pre-Generated design. The default is to generate 300 versions (blocks) of the design prior tofielding and respondents are assigned to different versions (often times withmultiple respondents receiving the same versions, which is just fine from astatistical standpoint).
There are at least two MaxDiff approachesthat involve customizing a subset of items to show each respondent that couldinvolve pre-generated designs, because the subset of items always has the sameconstructed list length: Express MaxDiff and Bandit MaxDiff. For example, 30 items could be drawn from amaster list of 100 items for each respondent to evaluate, so a pre-generateddesign accommodating 30 item slots could be employed. We should note, however, that prohibitionsmay not be specified if using pre-generated designs for Express MaxDiff orBandit MaxDiff. Rather, on-the-flydesign may be used with Bandit and Express MaxDiff for implementingprohibitions between items.
Visit this website - https://knowledgeexcel.com/maxdiff/
