dataSpring’s research on research suggests mobile respondents need to be a consideration in every online research study.

Consumers’ use of mobile to access the internet overtook desktop computers, accounting for 70% of internet use. So it is no wonder that researchers and consultants are eager to conduct more mobile studies. One aspect of the growth in mobile research that is often overlooked, however, is its effect on 'non-mobile' studies and the potential to introduce bias into the research results.

Whether intentional or not, your online studies are more than likely already including 'mobile' respondents. It is estimated that 68% of email is opened on a mobile device, so many online respondents are first seeing study invites on their mobile device. Further, according to research on research reported in the ESOMAR/GRBN Guideline for Online Sample Quality report, 20% to 30% of research respondents are already responding via smartphone or another mobile device. Given this, it is imperative to review all research methods and consider the mobile respondents' potential implication on both sample representation and response rate, and questionnaire design.

Survey Sampling in Mobile Surveys

Certain global regions (Asia, in particular) and demographics (younger consumers) are more effectively reached via mobile. In building dataSpring proprietary mobile research panels in Taiwan and Indonesia we found mobile respondents tend to be newer to survey participation and participation rates are generally higher. This contributes to better, and often more thoughtful, data.

In addition, as part of a recent qualitative study on these proprietary mobile panels, some interesting insights emerged:

Taiwan

• A higher response rate is seen in mobile surveys among panelists under 30 years, particularly in the crucial first 24 hours of fieldwork.
• Panelists aged 30 and above have almost the same response rate whether they take the survey via desktop or mobile.

Indonesia

• Desktop response rate is higher than mobile, but when it comes to the rate of increase of responses over time, mobile is higher.
• Overall, however, there are minimal gender and age differences between desktop and mobile response.

As Semee Park, dataSpring Product Manager explains, “The difference between these two markets is significant in understanding the data. While smartphones are the primary device used to access the internet in both markets, respondents from Taiwan are more informed on answering online surveys. Indonesia is still developing as an online survey market as the majority of market research methodologies used in Indonesia are still offline.”

So be sure to consider the potential of including a mobile sample in your sample plan and supplier selection when looking to target younger demographics and certain regions of the globe.

Questionnaire Design for Mobile Devices

Comprehension and clarity are imperative to good survey design. Given the limitations of the mobile interface, it is important to review all online questionnaires in an attempt to find a middle-ground where survey usability and readability is appropriate for both desktop and mobile devices.

As discussed in the blog entry How to Adapt your Online Study to Mobile, the key factors to review include:

• Text: Readability for any survey is key, so review the length of all text and be sure instructions are brief and to the point. Also, consider editing more formal ‘research’ language to be more conversational.

• Question Types: Large blocks of attributes and grids are difficult to comprehend on the smaller interface and anything that requires excessive scrolling, or pinch and expand may result in respondents speeding or straight lining these questions. Consider breaking up grids into separate questions.
• User Interface: Keep the design as clean as possible by avoiding unnecessary graphic clutter. Logos, headers/footers, and the like should be eliminated or minimized. Drop-downs are difficult on mobile so consider alternatives that also work efficiently on desktops.

Lastly, as a general best practice, consider including a quality check question or two in order to flag suspect data. This can include a single question instructing the respondent to select a certain response, or a grid attribute directing the respondent to click a specific column.

Final Thoughts

Whether intentional or not, mobile is becoming a factor in online research and does have the potential to create bias in the results. This fact requires that researchers and consultants review their sampling approach and questionnaire design to ensure data quality in mobile research is not compromised. Mobile respondents shouldn’t have an adverse effect on data quality, and in fact can, bring added benefit by enhancing survey response and more effectively reaching certain respondent groups.

Due to advances in mobile device technology and widespread adaption of mobile phones for online access, especially in Asia, mobile research has become a powerful tool for market researchers to harness. If you want to know more about mobile research, how to enhance your methodology toolbox, and Asia mobile panels, check out our Mobile Research Essentials page.