Difference between revisions of "Help:RankingList Ranking Strategy"
HELP PAGE DISCUSSION CLOSE
(New page: Survey designers may choose to enforce ranking strategies in order to bring consistency to ranking list responses. For instance, assume you have a question asking for someone to rank 5 ite...)
 
m (1 revision)
 
(No difference)

Latest revision as of 23:21, 24 March 2013

Survey designers may choose to enforce ranking strategies in order to bring consistency to ranking list responses. For instance, assume you have a question asking for someone to rank 5 items on a scale from 1 to 5 in increments of 1. One respondent may enter 1,2,2,3,4 while another enters 1,2,2,4,5. Without enforcing a ranking strategy, those are perfectly valid responses to the question. However, although the respondents are likely describing the same feelings in their rankings, the actual numbers they use differ and may throw off the results. It may be nicer to standardize their responses to 1,2,2,4,5 or 1,3,3,4,5.


There are currently two different ranking strategies that you can enforce on your respondents. They are very similar and differ only in how they handle ties. You can also select whether you wish to automatically adjust responses to conform to the selected ranking strategy.


Enforce a standard competition ranking (e.g. 1,2,2,4)

The standard competition ranking, items that are ranked equally receive the same number and a gap is left following the tied items. The size of the gap is one less than the number of items that are tied. You can also view this ranking strategy as assigning each item a rank equal to the number if items ranked above it plus one. Therefore, if you had five items, three of which were tied and one that was ranked better than the tied items and one that ranked worse, the valid standard competition ranking for these items (assuming a Ranking Scale from 1 to 5 with a step of 1) would be:

1, 2, 2, 2, 5


Enforce a modified competition ranking (e.g. 1,3,3,4)

This ranking strategy is very similar to the standard competition ranking, except that the gaps are placed prior to the ties. In this way, each item is ranked equal to the number of items ranked less than or equal to it. Therefore, if you had five items, three of which were tied and one that was ranked better than the tied items and one that ranked worse, the valid modified competition ranking for these items (assuming a Ranking Scale from 1 to 5 in increments of 1) would be:

1, 4, 4, 4, 5


Automatically adjust responses to conform to ranking strategy

When you choose to enforce a ranking strategy, the default behavior if a user enters valid ranks, but the ranking sequence they enter does not satisfy the strategy, is for the survey to generate an error letting them know the problem. This isn't necessary, and can be a burden to the user to try and tailor the responses to some specific strategy.


If you check the option to automatically adjust their responses to conform to the ranking strategy, instead of generating an error, the software will just adjust their ranks to the proper strategy. For instance, let's say that you have asked your users to rank five items along a scale from 1 to 5 in increments of 1. They supply a ranking of 1, 2, 2, 3, 4. Obviously, this does not correctly confirm to either of the ranking strategies that are available. If you choose to automatically adjust the responses, this is how they will appear on the respondent's results:


standard competition ranking:

1, 2, 2, 4, 5


modified competition ranking:

1, 3, 3, 4, 5