The Data Setting Duty of Care

by | Jul 28, 2020 | ICT4D |

Fadli stood staring at John. He blinked a few times. Then he said, “I have always been told I was born the day Temu struck gold.” John added this as a note. He knew a bit about running and Kenyan history, so assumed Fadli was referring to the Kenyan running legend winning Kenya’s the first gold medal in the 1968 Summer Olympics. Given it was a summer olympics, John entered Fadli’s birthday as July 1, 1968. And then moved on to the next question.

John didn’t realise the summer olympics that year happened between October 12-27. Not July. He also didn’t realise, while Temu was a Kenyan running legend, Temu was also a common local name. And there was a local legend of a Temu striking gold, in the early 50s.

Context matters. Data is not objective.  All data has a setting. And therefore, the data setting helps provide meaning to the data. Birthdays are a simple example. While power imbalances or incentives leading to silences in the dataset or selection bias is more complex.

The more the data setting can be passed along with the data itself, the better chance those receiving the data will understand it more fully. And those ‘receiving’ the data can be partners or can be future staff or staff in other parts of your organisation.

And yes, there is an elephant in the room. To do this well, takes time. It is ‘extra’ work. However, maybe that’s just our framing of it. Too often we view data as objective, neutral, as fact. However, it is not. Data is infused with a perspective, opinion, with worldview. Data tells a story.

Therefore when we present it as fact. When we separate it from the setting in which it came from, it loses some of its value.

And so yes, it might take more time at the point of capture, but it might save errors and cause less harm in the future. Therefore, perhaps we should frame it as being a professional or as our duty of care. Maybe we need to be able to make notes in our systems. Or maybe the ability to check a box indicating this birthday, this data field is a ‘guess’. And then in the explanations accompanying the dataset we explain what standard ‘treatment’ we gave to all ‘guesses’.

Duty of care is a choice. And the choice is up to us.

Photo by Getty Images


Submit a Comment

Your email address will not be published. Required fields are marked *