The fallacy of 10% as failure

If 10% of the people who come into a telecenter receive help with computers from the staff person, is that considered sufficient impact?

I just returned from Chile with our infomediary research team and after visiting several telecenters, libraries and cybercafés I feel that the development field under-appreciates the benefits that come from the 10% of internet uses that rank far behind email, social networking, and playing games.

Let’s play this out. Alexi, a manager at one telecenter we visited in Traiguen, says he gets 50-60 people per day, helping on average 6 to 10 of them complete whatever task they came to do. That’s 300 to 500 assists per year in a town of under 20,000 (admittedly many are repeat users). One area where he helps a number of people is registering on the ChileCompra website, a government procurement program aimed at small and medium sized businesses. Alexi says many of the small business owners don’t know how to use computer so he helps them through the process, showing them how to bid for a public contract. He doesn’t track how many succeed, but knows anecdotally that many do because they come back and bid on new projects and tell him about their past successes. Alexi also helps people with their resumes, set up bank accounts, comparison shop, make Skype calls, and a myriad of other tasks, many of which I observed during my two-hour visit (e.g. three mothers came in with their high school aged children to set up bank accounts, a requirement to qualify for a college scholarship). Yes, most visitors of telecenters are young, use the computers for social networking and other tasks we typically place under the category of communications and leisure, and don’t need the type of assistance that Alexi provides, but if we are serious about reaching people in need should we dismiss what happens with the 10% of users and activities just because they show up at the lower end of our frequency graphs?

This tendency to expect more productive uses of ICTs among underserved populations reflects a double standard. In the west we don’t expect the majority of our internet uses to be for so-called productive purposes. We surf, email friends, and download music, and while we could all probably do without the internet for much of this, each of us could point to examples (let’s say 10%) where the absence of the internet would severely impact our lives. So why when we assess an ICT and development project do we look at 10% of uses and conclude that the investment is not being properly utilized and deem the program a failure?

Part of the problem is methodological, for which we researchers are to blame. I believe we are making a mistake by comparing uses that are regular in nature with those that are episodic. A typical survey asks people to state how frequently they use the internet for email, social networking, health, employment, education and so on. When people report (unsurprisingly) that they do email every day and use the internet for employment related activities infrequently, we conclude that the importance of the internet is greater for the former. And, since the everyday activities are typically associated with unproductive or non-developmental purposes, we then conclude that impact is minimal and the program unworthy of public support. This is an erroneous conclusion in my mind because it mixes regular with episodic activities. Thank goodness I don’t need to use the internet weekly for employment or health related purposes. But if or when I need to find a new job or learn about some disease I’ve caught I’ll be glad I have the internet. The methodological remedy I propose is to stop comparing regular with episodic uses. It leads to such conclusions as 10% representing failure.