Some of my favorite Qualtrics functions are the things that have an obvious purpose, but which you can use for all sorts of non-obvious ways. One of those features is the Translate tool, which you’d assume is only useful for displaying your survey in different languages, but which can actually be used to display your survey with different flavors of text in the same language depending upon what you know about your participants already or ask them during the screener.
For example, let’s say you’re surveying educators and you know you’re going to have two groups to whom you can easily customize the survey: one group preferring he or she pronouns and another preferring to be called either a teacher or a professor. Using display logic and piped text, you can certainly do this with some embedded data, but you may wind up with fragmented data across several questions that you’ll have to reassemble in analysis. Wouldn’t it be nice to just ask the same questions of everyone, but customize how those questions appear depending upon their characteristics?
Let’s consider the following survey for a moment, which we’re sending to a list of people about whom we know their preferred gender pronoun and their preferred educator title of “teacher” or “professor.”
Q1 How are you doing today, sir? o Pretty well (1) o OK (2) o Fine (3) o I'm not sure (4) o Other (5) ____________________________________ Q2 I can see from our records that you're a public school teacher. Which of the following statements best describe the role you play in your school? ▢ I teach primary school (K-5) (1) (teachers only) ▢ I teach middle school (6-8) (2) (teachers only) ▢ I tech high school (9-12) (3) (teachers only) ▢ I am an administrator (4) ▢ I'm in a support staff role (5) ▢ I teach introductory undergraduate courses (6) (professors only) ▢ I teach higher-level undergraduate courses (7) (professors only) ▢ I teach graduate courses (8) (professors only) ▢ I am tenured or tenure-track (9) (professors only)
We can customize Q2 with display logic to ensure the choices are only shown to the appropriate person. But how do we fix that “sir” and “public school teacher” to reflect each respondent’s actual characteristics? It’s as easy as going to “Tools” and hitting the “Translate Survey” button.
From here, we can set “English (US)” as the standard language and then create variations based upon it that are also in English. The variations will use English (US) as the default language unless we make a change. So, for example, if we want to change the survey to refer to female teachers as “ma’am” instead of “sir,” we can create a variation that we call “English (Female, Teacher)” and then edit the question text accordingly:
That’s fine for this question, but for Q2, we actually need to know the respondent’s title: are they a teacher or professor? So, we can create two other variations for males and females who are college professors, which I’ve named “English (Male, College)” and “English (Female, College)” and then changed the question text to reflect that they’re a college professor:
Easy! Now all I have to do is set the survey to display the correct version to the correct respondent. I can do this in two different ways:
If I already know these characteristics and I’m using a survey that can reference a contact panel or custom URL with this information hardcoded, I just need to use the Survey Flow to direct Qualtrics as to which language to use. When I created each of my language variations, I gave each a short name: En is my default for male teacher while E-F-T is my variation for English (Female, Teacher). So I can set up branches in my survey flow to assign a language to each group:
That’s the easy way. But what if I don’t know that information ahead of time?
If I am missing these characteristics, I can include two screening questions and then use them to program my survey flow branches. This is as easy as setting up an initial block of questions with a screener for gender and another for type (teacher or college) and then having the survey branch for language after the screener block.
Why go to all this trouble? Here are just a few of the advantages of using this technique:
- Your survey will seem much more personalized to your respondents, which will make them feel more engaged by it. When surveys are asking respondents questions that seem to be smart and well-targeted to them, they tend to take the surveys much more seriously and are more likely to complete them. A little bit of effort can go a long way towards improving your data quality.
- You will spend a lot less time programming redundant questions to account for branches or complex display logic. One of the worst habits you can get into when programming surveys on Qualtrics is creating multiple versions of the same question which are designed to be slightly different for subgroups within your study. This not only makes your survey harder to test and revise, but also can make it an analytical nightmare if you need to compare data between those questions.
It’s far better practice to simplify and reduce the number of questions you’re asking. - This method provides a very effective way to conduct A/B testing or to tailor questions to particular geographic regions or demographic groups without requiring multiple surveys. The combination of the Translate function, Survey Flow branches and Display Logic can allow you to create quick and easy variants of your survey that you can easily compare because the embedded data labeled “Language” can instead become an easy breakout variable for group comparison in your report.
Since you can customize the questions however you like, you can also include pictures or other rich text embeds within your variants. - You can also use this method to introduce some fun and engaging elements. For example, let’s say you wanted to offer the same survey to the same group repeatedly, but throw in some jokes in between questions to keep things light. (I wouldn’t necessarily recommend this, but hey – let’s allow ourselves to dream!)
You could write some sort of joke generation logic or script and toss in random jokes through piped text, but you could also just construct a few variants of your language which would randomly display to the user. If you created, say, five variants where the questions were the same but the jokes were different, you could simply include an element in your survey flow early on to assign a random language using a randomizer with five branches each leading to “Language” being set to one of your five variants. Your participants would be likely to get a different experience each time, and you could easily refresh the jokes without having to touch the survey logic by simply going in and instead adjusting your Translate Survey tool.
If you wanted to have some real fun with this, you could even have each variant have its own background and color scheme. It wouldn’t be hard, but your participants would certainly feel like you went to a lot of effort to make their survey experience fun and engaging.
Want to play around with what I’ve built? Here’s a survey preview, and here’s the QSF file. Have fun!
If you’d like to discover more Qualtrics tips and tricks, be sure to check out other articles in our series, and if we can help you or someone you know with a Qualtrics question or problem, please contact us!