Operationalising Data Ethics: 2 of n
By NATHAN KINCH
In part 1 of this series I introduced you to the defintion of a Data Ethics Framework. I highlighted the importance of social preferability experiments and how to conduct them.
This post will focus on another area of your framework, the Key Decision Log.
But before that, let’s quickly revisit the definiton.
It’s important that I highlight this again. The definition sits at the heart of our approach to operationalising data ethics. It helps us move beyond feel good statements and start changing the way we behave. It keep us accountable – to ourselves and our stakeholders – by design.
Let’s get to it.
Just because we can, should we?
Ethics is a decision-making process.
An operational approach to data ethics challenges you to design consistent methods for decision-making that are inclusive, participatory and empirically supported.
Over time we’ve learned that a central, easily accessible and highly useable/useful decision log helps.
The basic idea is that decisions are proposed. A decision making process is then executed. This likely includes a Social Preferability Experiment. If the result is a pass, the decision is logged, the action added to a backlog, eventually prioritised and then implemented and refined over time. If it’s a fail, you go back to the drawing board to evolve the approach or stop completely.
This features in multiple workflows. The decision-making context can be proactive, retroactive or retrospective. It becomes a key feature of how the organisation as a whole operates.
Here’s a diagrammatic representation of a Data Ethics Framework. It’s effectively a ‘system’, with inputs, throughputs and outputs.
This matters for a bunch of different reasons. But two stand out most:
- Internal clarity: Everyone in the organisation has access to this resource. It helps establish precedent. It guides and accelerates decisions and actions. It helps to progressively demystify complex, nuanced and ambiguous issues. And
- Meaningful accountability: The organisation can no longer hide behind a veil of secrecy. In fact, if this is done the way we propose it is, it’s quite the opposite. The organisation becomes open by design. Key stakeholders – customers, regulators, staff, independent advocacy groups, shareholders, academics etc. – can become part of the decision-making processes that directly and indirectly impact them.
This is powerful for a few reasons. The most important is that it enhances the trustworthiness of the organisation. It helps you do ethics better. And ethics is 3x more important than competence when it comes to trust. Turst disproportionally impacts bottom line business outcomes. So, in short, this really matters.
Where to start?
Select one key decision with ethical implications. Execute a basic social preferability experiment with a diverse group of stakeholders. 10 participants in total is enough. Publish the process and the results for everyone within the organisation to see.
This will be interesting. People internally will talk about this. It will help you begin designing for better outcomes together with the people that will be directly impacted.
If you wanna move beyond this and make meaningful progress faster, talk to us. We’re here to help.