Using mobile technology to collect data

A few snippets from the Mobile Technology and New Media: Trends and Opportunities panel at the September 23 Tech@State event.

Wayan Vota (Moderator)
Joel Selanikio, Data Dyne
Oscar Salazar, Citivox
Matt Berg, Millennium Village
Prasanna Lal Das, World Bank.

No analysis here. Just stuff I wanted to remember.

Wayan: There is a huge cost to inputting data. Example of teacher overburdened by participating in research — having to document when students arrive, what they eat, how they performed, etc. — in addition to teaching. How do we get folks motivated and excited to input data on day two? Because it’s boring and there’s a huge incentive to fake the data.

Joel: In global health the data have is a subset of what we need. EpiSurveyor: Mobile data collection on a phone. When people want to collect data they will use whatever is easiest. We have motivated people who don’t have access to data-collection tools. We use the web to get it out as software. Now so many apps we use are online, no longer installed on our computer. But in international development we don’t use this approach as much. Why? EpiSurveyor is easy. People can download it and start using without our help.

Prasanna: At the World Bank we don’t have data gathering challenges. We work with data already collected. We want to get our data out there, we want people to use it (slice and dice!) and give us feedback.

Matt: Data is the incentive. Healthcare workers want to know. Data collection can’t be a burden. Can’t ask people to repeat data collection processes. Data must be actionable (people won’t report that the well is broken if they know it will never get fixed). Data has to be your own. Have to give the information back in a way that’s understandable and valuable to the community. This creates incentives to use systems.

Data input process has to be easier than paper — and should give something of value back.

Joel: Making something easier than paper is done. We can do it on mobile phones. Not difficult at all.

[Note: Not difficult with closed-ended questions with numerical responses. So depends on the research design.]

Multiple levels: I put in data. It gets analyzed. I review that analysis and add to it. I share my analysis with broader community, then they build on that, etc. Take, digest, bring to next level.

Joel: Affordable distance communications has revolutionized healthcare. Most stock outages are reported orally. We want to help engineer your process with structured information exchange.

Matt: Calling [voice] is great. But informatics comes into play when you’re talking about scale. Often just giving back information along with context of what is normal [so folks have something to compare their results to] improves outcomes.