|
<< Click to Display Table of Contents >> Navigation: User Guide > Project definition > Amending entries > Randomize questions and responses |
For CATI and CAPI you may want certain questions or response lists to appear in a random order to the interviewer.
This section describes the ability to ask certain groups of questions and responses in a random order during data input.
Randomising questions and responses does not alter any data locations and the data is stored exactly the same as it would be without randomisation.
Randomised questions can have randomised responses.
A question with responses can have the "Randomise responses" setting on the Entry details window to request that the responses are displayed in a random order during data input.
Individual responses can be marked as "Fixed position" on the "Input settings" tool window so that they are not moved in the response list.
The order that questions are asked in data entry can also be randomised. This is done by adding entries that surround the questions to be asked in a random order using Randomise questions accessed from the Main window menu.
TIP: The randomise question order entries are normally added at the end of the project definition phase, because it adds a lot of entries to the entry list.
The program uses these two terms to describe the way randomisation is implemented.
A set is one or more sequential entries that are to be shuffled with other sets in the same group.
During data entry each group is inspected in turn and all the sets within the group are moved around at random, swapping places with each other.
For example, if there are 5 questions Q1A to Q1E that are to be asked in random order then each of the five questions will be marked as a set by placing a random entry before and after each of the five questions. All these five sets will belong to the same group and be identified with the same group name.
A set is usually a single question but it can be any number of questions with associated variables and filters.
Any number of sets and groups can be used.
Randomisation can be nested - a large set can contain other groups of sets within it.
You are allowed entries between sets in the same group, these entries will not be moved.
You can repeat the same order in different groups or response lists by allocating the same named random entry to them.
A named random entry is a ZZZRAN... entry, for example ZZZRANCARS.
Each group using the same named entry will have sets in the same order provided there are the same number of sets in each group, see Randomise questions.
Each response list using the same named random entry will appear in the same order provided there are the same number of (not fixed) responses in the list, see Entry details question.
The named random entries must be positioned before the first question in the project.