Survey research has been a part of everyday life for over 70 years. Understanding data, the way questions are asked, and the selection of respondents have significantly evolved during this period. However, every research still has to address two fundamental issues: a) how to reach respondents and why they would be interested in answering questions that are important to us, and b) how to ensure that the answers provided by respondents are truthful.
We have addressed the first issue in more detail in recent months. The appearance of health problems has seriously questioned the most accurate data collection technique, face-to-face interviews. Of course, the pandemic has not only affected our industry, but answers to specific questions related to business transformation must be sought individually. Therefore, we were forced to test some new models of communication and recruitment of respondents, using the new resources of MSSLab, which provided us with answers that greatly helped us adapt to market changes. In this blog, we want to share those answers with you.
The first question we have been thinking about for a long time is the potential of moving the business completely online. Recruiting respondents is the basis of our business, so the main question we have been thinking about is how much potential there is for alternative ways of recruiting respondents. Can we rely on Viber, Facebook, or other forms of online recruitment in the long term? No matter how well thought-out our strategies are, we can only get answers to these questions empirically.
In May of this year, we attempted to conduct a complete survey by recruiting respondents through Viber messages. The pilot study went very well, and we extracted some excellent guidelines from it. Two questions interested us here – how many people can we recruit on weekdays compared to weekends, and is it better to send messages in the morning or in the afternoon? During four waves of recruitment, we sent 12,500 messages per wave, randomly selecting which potential respondents received messages in which wave. The idea behind this experimental design is to obtain roughly the same groups of people so that we can determine whether the day and time of recruitment affect the final response rate we can expect. To our surprise, we could not conclusively conclude that it is better to contact respondents on weekends rather than weekdays. The number of respondents who responded to our call to complete the survey did not significantly differ between Monday and Wednesday, compared to Saturday and Sunday.
However, the second variation in our experiment was very significant. Contacts we made in the morning (11:00 a.m.) achieved a response rate 25% higher than contacts made in the afternoon (6:00 p.m.). The results we obtained guided us in further planning of online recruitment and ultimately helped us optimize time and cost savings in recruitment. In other words, it helped us improve our business.
This brief demonstration is an example of the careful approach we employ in solving all research problems. Using the capabilities of MSSLab, we can also help you get answers to various questions you encounter in your everyday business.