The interviewees swore they preferred the pill style button over the rectangle and that the colour red was nice. They also said the flow of steps was fine and smooth, the product site page they liked too. They said the experience was good. The beta trial showed this wasn’t exactly the case and the launch showed mediocre results. The launch site page with the main CTA failed to get more than a 23% read and CTR on the offer link was below 1%.
The interviewees? They lied. On the surveys and in the interviews. Straight up fibbed. Not on all the answers, but some key ones.
This is cultural bias at work. Yea-saying, nay-saying, socially desirable responses. All are aspects of how different cultures respond to questions. This applies to both different cultures in countries as well as market cultures (an often overlooked aspect.)
Controlling for various biases in survey questions is difficult at the best of times. If you’re UX research involves blasting out an email link to a survey to as many people as you can and from anyone who signed up, this is even harder. Even if you ask screening questions at the front for age, location and perhaps income brackets, not all answers will be true. confidence percentages in this case are almost impossible to predict. And understanding any cultural bias is almost impossible. Surveys often suffer from acquiescence bias, where people will agree with the question being asked.
From acquiescence bias to nay/yea saying much is impacted by the culture of the individual that they were born into or raised in, or perhaps, have just lived in for the majority of their lives. Cultural factors always influence responses.
In many Asian and Scandinavian cultures for example, acquiescence bias and yea-saying is much more likely because their cultures are more conservative and place an emphasis on conformity and the “self” is seen as part of the greater society. In American and French cultures, the “self” is more independent and there’s a tendency to think more of ones own views as being ahead of the collective.
This applies to corporate cultures as well. In businesses where employees are given more latitude and empowerment to express their opinions and make autonomous decisions, you get less acquiescence bias as opposed to corporate cultures that are very hierarchical and collaboration is encouraged less.
Taking cultural biases into account in both qualitative and quantitative UX research is critical. Otherwise you won’t really have the insights you expect.
To deal with these biases, questions need to be short and easy and use intentional language. Adding in free-form boxes in online surveys is helpful as well. But keep in mind, long-form responses don’t often come from cultures where conformity and a stronger sense of being part of something bigger than the self comes into play. And of course, always avoid leading questions. You’re really not doing yourself any favours with those kind of questions, especially with SaaS and mobile apps.
We often see startups and even more so large enterprise clients come to us with their cultural biases baked into their research scope already, often without realizing it. Culture is the first thing we address in UX research and strategy scoping.