I’m a data nerd at heart – always have been.
When I was a kid, I had to do the dishes as my chore. I would have fun timing myself to see if I could beat my previous record of loading or unloading the dishwasher.
Today, I’m still tempted to collect data for household chores, but my attention has been redirected to statistics for the Bible translation movement. In my role at SIL International, half of my job is being the director of ProgressBible. Our aim is to inspire informed action amongst Bible translation organizations, enabling them to be most effective in their work. As a part of ProgressBible, I am heavily involved in the Every Tribe Every Nation (ETEN)  collective impact alliance. We follow Stanford’s model  for collective impact, in which one of the conditions is shared measurement. ProgressBible serves as the engine that collects and analyzes data across all organizations participating in the collective impact alliance.
One of our current areas of focus (and my NPO) is providing an analysis on Scripture usage, and reasons why there is a lack of Scripture usage by some language communities. I’ve had over a hundred conversations with missiologists, sociolinguists, strategists, Scripture engagement specialists, church leaders, field workers, and community members to try to understand the reasons for lack of usage. Unfortunately, this is an exceptionally complex field, and there were not any easy answers (or solutions) to this problem. But as I have been asking around, I have come to the realization that I have to fight the temptation to give into my confirmation bias . I have a hypothesis for the major reasons why Scripture is unused, and my tendency is to be biased towards that hypothesis. So I’ve had to ‘Search my feelings’ as described in rule one of Tim Harford’s How to Make the World Add Up.  How were so many blinded to counterfeit Vermeer paintings? They were blinded because they wanted to believe that the paintings were legitimate.
Another contribution that ProgressBible makes to the Bible translation movement is to provide a forecast of when every language community will have access to Scripture in a language they understand best (named the All Access Goals by ETEN). Rule three of was especially relevant to the forecasting we do.
Rule three tells us to avoid premature enumeration. Harford opens the chapter with an example about infant mortality being higher in the rural areas of England than in London. In reality, this was due to a difference in reporting choices between classifying deaths as a miscarriage or as an infant death.
In Bible translation, we have been tracking the forecast to reach the All Access Goals closely, and publish monthly updates based on the most recent data. At times, there can be major fluctuation in the data, and therefore the forecast. There are many reasons we can have a ‘good’ month in which it does not actually mean that the pace of Bible translation has sped up. It could be that we discovered Scripture that was translated some time ago, that an organization was behind on reporting and just caught up in a single month, or many other reasons. This premature enumeration can lead us to inaccurate conclusions, just as the case with infant mortality in England.
I have long felt that statistics can be manipulated or used strategically and the difference between the two may not always be clear. Some may consider either the visualization on Iraq’s Bloody Toll or the alternative view of the data Iraq: Death on the Decline  to be manipulative or pushing an agenda. In working with and presenting data, I have felt a burden to steward this responsibility with integrity, and yet there is a tension between showing compelling visualizations and presenting the data in a way that promotes an agenda.
May the Lord give us all wisdom in using data wisely!
 Schulz, Kathryn, Being Wrong: Adventures in the Margin of Error. New York, HarperCollins, 2011.
 Tim Harford, “How to Make the World Add Up: Ten Rules for Thinking Differently About Numbers”, The Bridge Street Press, 2020, 21-50.
 Ibid, 71-92.
 Ibid, 248.