How to randomize items at Limesurvey

Limesurvey provides various randomization features at different levels and you should make use of it for various reasons. Especially at market research surveys randomizing items is a must-have feature in order to collect data properly. Another use case is A/B testing where you randomly show version A or B or a question for comparing the outcome of e.g. different question texts.
There are some more reasons we have outlined below when introducing the different Limesurvey randomization features.

Randomizing the order of answers/items

Research shows that people tend to pick one of the first answer options and often do not even check all available options. This can result in misleading survey results. So what can we do?
Well, that is rather simple: Most Limesurvey question types provide a setting Random order which can be found at the Display section below the General options. Simply turn it on and that’s it:

Limesurvey "Random order" setting
Limesurvey “Random order” setting

Want to check how this looks like at a real survey? Then check out the test survey we created for you.

Randomly show 1 of X questions (can be used for A/B testing)

There already is a special blog post dealing with this topic so we will just link to it: https://survey-consulting.com/how-to-randomly-show-1-of-x-questions-at-a-limesurvey-survey/

You can check this randomization feature at this test survey.

For creating an A/B test the approach is the same as explained at the example linked above. Simply created two copies of the same question and set conditions based on the random number to either show question A or question B.

Randomly show questions on page

Besides randomizing items within a question you can also randomize the questions displayed on a single page or within the whole survey.
For setting this up there is an option Randomization group name at the Logic section when editing a question. Enter the same randomization group name at all questions which should show up in random order:

Adding a randomization group name to a question
Adding a randomization group name to a question

So if you added rand1 as randomization group name to questions Q1, Q2 and Q3 these will show up in random order when loading the survey like Q3, Q2, Q1 or Q2, Q1, Q3.

We have provided a sample survey at this link.

Randomly show question groups or survey pages

The group by group display mode is the survey format used at most surveys. This means that you put questions dealing with the same topic into the same question group like questions asking for demographic details. If your survey deals with e.g. tech companies you could then have another set of questions at the follow up question group dealing with question related to Amazon, then have another question group asking details for tech company 1, another one for tech company 2 and so on…

Since the drop off rate would usually lead to receiving many opinions about Amazon but just some very few responses related to the tech companies listed towards the end of the survey, your data will be less reliable when comparing the response details for different companies.
To work around that you could simply show the survey groups related to the different tech companies in random order. This can be achieved by entering the same Randomization group when editing a question group:

Adding a randomization group title
Adding a randomization group title

By adding randomizecompanies to all your to be rated companies these pages will show up in random order within the survey.

We have set up an example with Amazon, Google and Facebook for you showing the question groups for rating these companies in random order.

Randomly show 1 of X follow up surveys

Sometimes users are redirected to a certain follow up survey once they have completed the main survey. In certain cases there is not just one survey but you may want to pick 1 of X surveys randomly and redirect users to them.
This is doable by following the approach of randomly showing a question. You create a question of type equation with code randnumber and as question text you simply add:

{if(is_empty(randnumber.NAOK), rand(1,3), randnumber.NAOK)}

This will create a random number between 1 and 3 if there was no such random number created so far. The newly created number can then be used for defining which URL to call.
You can have another question of type equation using question code redirecturl and based on the random number you can define URLs for 3 different surveys.

{if(randnumber.NAOK == 1, "https://survey.mydomain.com/index.php/111111", 
if(randnumber.NAOK == 2, "https://survey.mydomain.com/index.php/222222", 
https://survey.mydomain.com/index.php/333333"))}

At the survey settings you can optionally enable Automatically call end URL and at the End URL text field of the setting just add {redirecturl.NAOK}. Depending on the randomly created number and the end URL set for it Limesurvey will call one of the 3 surveys randomly.

We added the random redirect to the sample survey created for the Randomly redirect or show 1 of X questionsuse case. Click here to start that survey and you should then see different end URLs at the final “Thank you” page pointing to different surveys.

Sample survey showing all randomization features

All the different features outlined above can be tested at this Limesurvey randomization sample survey.

How to calculate the number of days between a date inputted by the user and the current date

Sometimes you want to find out how many days or years have passed between today’s date and a date inputted by the user at a Limesurvey survey. The process is a little complex but it can be done using the Limesurvey Expression Manager.

Calculating the number of days

Let’s assume there is a question of type “date” at your survey which uses question code “q2”. The user can select any date using the date picker and at the help text you can show a message like “That was X days ago.”.
To calculate the value for X you can use this expression:

{floor((time() - strtotime(q2)) / (60 * 60 * 24))}

This looks a little complex, so let’s have a look at the details:

  • To get the current time stamp we use time().
  • Using the strtotime() Expression Manager function we create a time stamp based on the date inputted by the user at question “q2”: strtotime(q2)
  • That time stamp gets subtracted from the current time: (time() – strtotime(q2))
  • Since we are not interested in the time difference in seconds but want to know about the exact number of days we divide the result by (60 * 60 * 24) because one day has 24 hours with 60 minutes and 60 seconds.
  • Finally, the result is rounded down using the floor() function.

Calculating the number of years

Similar to the above approach you can also calculate the number of years between a certain date inputted by the user and today’s date. The expression for this is:
{floor((time() - strtotime(q2)) / (60 * 60 * 24 * 365))}

Using calculated result within conditions/relevance equations

You can not only use this expression to output the results of the calculation within any (follow up) question but you can also use this code for conditions by entering it at the “relevance equation” field (without “{” and “}”). E.g. you can ask participants for their day of birth and then calculate if they are already 18 years old. If not, you can use conditions to hide certain follow up questions or show them a warning.

Storing calculated result

If you want to store the calculated result all you need to do as adding the expression (with “{” and “}”) at an equation question. You can even hide that question (edit question -> advanced question settings -> “Always hide this question”) and the result will still be stored at the database.

Limesurvey – Conditions based on token attributes

Sometimes you want to show certain questions of a survey to a specific user group only. With Limesurvey this is no problem. If your survey is set to be non-anonymous and uses tokens, you can use user defined attributes for conditions. Here we describe the necessary steps.

1. Adding additional attributes

Set the survey to be non-anonymous:
Limesurvey survey settings

Enable tokens:
Enable tokens at Limesurvey

Add a certain number of user defined attribute fields:
Add token attributes at Limesurvey

Assign attribute names after having created the new fields:
Label additional attributes

Add some dummy tokens with user defined values

Add dummy tokens

Set conditions for a certain question

Go to the question and call the condition designer:
Limesurvey condition editor

Select the token field on which the condition should be based on (e. g. “Department”) and define the constant which the condition should match (e. g. “Management”):
Condition based on token attribute

Check condition

You can use the group preview for a first check if the condition was stored successfully,
Limesurvey group preview

because in this case the question will not be shown at the group preview (because the conditions isn’t met). The arrow indicates where the question was previously shown before setting the condition:

Test survey at runtime

Once tokens were assigned when adding users or by using the “Generate tokens” feature, the “Display tokens” screen contains an icon to test each token entry:
Limesurvey token overview

Of the above users, only “Jane” belongs to the department “Management”. Let’s check the different survey behavior to check if the condition works as expected:

Survbey at runtime Survey at runtime

More information

 

How to conditionally hide/show groups in Limesurvey

When first working with Limesurvey’s condition designer, you might miss a feature to show/hide whole question groups. Doing this is quite obvious though:

  1. Set a condition on the first question of the group like “Show question X if question Y was answered ‘>=3’ “.
  2. Click the “copy conditions” icon at the condition editor.
  3. Select the previously set condition on top and at the list of following questions select all questions of the according group:

    Show/Hide groups at Limesurvey by using conditions
    Show/Hide groups at Limesurvey by using conditions
  4. Click the “copy conditions” button
  5. Done 🙂

Note: At Limesurvey 1.92 you can use the Expression Manager for hiding groups.