Originally posted July 20, 2014 on VOTO Mobile. Reposted and adapted for Viamo.
By Levi Goertz
I noticed that my colleague Sean would eat the same thing for dinner almost every night. After neither of us could agree on how often he ate Jollof rice, we decided to gather some data on his eating habits, and come back to the conversation with a little fact base.
With Viamo mobile surveys it’s effective to set up a recurring survey or message that goes out at a regular interval.
Think of daily surveys to field agents to check in on key tasks performed and issues encountered; a regular message to a health clinic about supply stocks to manage inventory, or a daily survey to ask you if you did your morning sit ups.
I made a little SMS survey asking Sean what he ate for breakfast, lunch and dinner and how satisfied he was. After a brief consultation with him, we set on a schedule where it would go out every day at 9 p.m. for a month.
We learned that Sean eats Jollof rice for 39 percent of his dinners and 31 percent of his lunches. The man clearly loves Jollof rice – check out how his consumption of Jollof looks over each day of the survey.
My view on Sean’s consistency was cemented when I saw he also eats corn flakes for 41 percent of his breakfasts.
Not only is Sean a frequent Jollof rice consumer, but he takes in big portions. In the Viamo office most people would eat about 2 cedis ($0.24 US) worth of rice. But look at the size of the plates Sean is eating.
We know Sean’s behavior, but is this really making him happy and satisfied? At first glance he seems to be making good choices, because he is pretty satisfied with what he’s eating. Take a look at his daily satisfaction ratings.
We might think that Sean just loves Jollof rice and corn flakes that much. But I did a few different regressions to test Jollof rice’s impact on Sean’s satisfaction. I found absolutely no statistically significant impact of whether or not that is what he ate.
I tried the same regressions with corn flakes; no dice there either. Even the amazing house dinner option – where someone else cooks Sean a great meal – didn’t have any impact on his level of satisfaction. There is no statistical relation between what Sean eats and how satisfied he is with his food. I guess the man is just happy.
Simple daily surveys can yield interesting and counterintuitive results. When you have good data, you may learn that you have an independent variable where you had assumed there was cause and effect. You may learn that you could try eating a few other things and be just as satisfied.