Friday, July 23, 2021

Methods and Reasons in the Pillar Cell Phone Data Story

The Pillar Podcast just put up their latest episode, which provides answers to some of the questions that people have asked about the story which led to the resignation of the USCCB general secretary. Here are my notes from listening to it for those who don’t have the time.

They spend a bit of time explaining for listeners what the nature of the USCCB is and what the role of the General Secretary is in it.  The USCCB is mostly not responsible for setting policy and certainly not doctrine, but it does often issue guidelines and best practices. Recently many of those guidelines have related to how dioceses, parishes, schools, etc. should adopt practices to avoid sexual abuse. The General Secretary is described as being the day-to-day head of the organization, directed by the President, who is a bishop but spends most of his days running his diocese while the General Secretary is in the actual USCCB offices running things.  They compared the office of General Secretary to the CEO of a company, while the President is more like the chairman of the board.

They state that their data analysis was based on a broad dataset provided by a source which had legally purchased the data. The data did not provide names or phone numbers, it just gave locations, apps, and a unique id of sorts. The data included the use of a lot of different apps of different sorts, and whatever the motives of the person who gave them the data, they approached it looking for general trends not hunting for a specific person. 

As an aside (although I have published pieces at The Pillar, I had not involvement in this story and do not have any inside knowledge about its analysis) this is pretty much what I suspected.  Having worked with data vendors and marketing departments at large consumer product companies for much of the last ten years, I'm very familiar with this kind of data.  Most apps sell it (it's what in the terms of service you click to agree to without reading them) and it's used by marketers to do things like target ads towards people who live in Houston and have entered a Home Depot within the last three months. So marketers use it in the aggregate to identify groups of individuals they want to send ads to.  But there's no real reason the same data can't be used to identify a person based on where they've been as the NY Times did after Jan 6 in the story linked below.

After acquiring the data set, they say that their first move was to verify that it was legally acquired by the source and also to determine that the data was complete and what it purported to be. 

Ed pointed out that whenever a source provides information or data or a quote, there is always some reason for doing so. This is normal, and to a great extent not of interest unless it affects the veracity of the information. 

They pointed out this is not the first time journalists have analyzed data like this, a key example being the NYTimes story in which they traced some of the people who were at the storming of the Capitol to their homes.

JD stated that they had started very broadly looking to see what they could determine about the use of apps in the Catholic Church in relation to the Church’s efforts at reform 

As they were analyzing data in relation to Church related events, they noticed that a Grindr user seemed to keep showing up at events and places, the combination of which suggested this was some high USCCB official. 

They indicated that they had not come into this with a tip about the general secretary which they went to verify, but rather that they were looking to assess things in a wider way and saw this pattern jump out at them.

There was also significant discussion of why they believed this was news -- and why the particular nature of the general secretary’s role made this news in a way that the personal immoral behavior of some random priest or Church official would not. 

One thing they noted was that in addition to the fact that this was a very highly ranking cleric, with responsibility for (among many other things) helping to guide the formation of the Church's guidelines on questions such as how app usage by priests should be monitored or curtailed in a parish setting. Another is that Grindr itself has been a locus of a number of recent cases in the US and other countries of priests becoming sexually involved with minors.  In one recent case, the priest was not civilly charged, because the civil authorities determined that he had not known the person he made contact with via Grindr was in fact a minor. Thus, they said it was very much Grindr that was the key concern to them in terms of risk in relation to minors, not homosexuality in general.

There’s a lot of other material, dealing with questions and accusations made against them, but since I was primarily interested in the data questions I’ll direct you to the podcast to find out more about those. 

One other resource I'd direct you to is a very good piece which Brandon wrote over at Siris on the nature of detraction and its relevance to this particular news story.  A number of people have voiced concerns about whether running this story constituted detraction, and I think Brandon, as always, does a good job of looking at what precisely detraction is and how it would and would not apply.


No comments:

Post a Comment