Repsonsible storytelling

Without some element of storytelling, data analysts wouldn’t be able to effectively communicate the findings of their analysis to broader teams, leadership, and stakeholders. Communication to industry, community, and government are all integral parts of conducting research.

Data analysts are on the front line of providing metrics to executives and leaders. They have a responsibility to themselves and to their reports to portray the figures accurately. To achieve this, analysts must adopt responsible storytelling practices, void of bias.

Results that are directly or indirectly manipulated due to personal bias, or to fit desired organisational outcomes, ideologies, or agendas, devalue the overall practice of data analytics and compromise its future. Keith McNulty, analytics leader at McKinsey, outlines the responsibility of data analysts and the role of data analysis itself.

‘Associating a piece of analysis with an objective within the organisation, implying that the key reason for undertaking the analytics is to develop a narrative that will entertain that objective. Data analytics should only ever be associated with a question, never an objective.

Determining a “desired narrative” before any analysis is conducted, automatically installing a bias on decision making and a pressure on analytics professionals to entertain the narrative. Narratives should only be built when results are completed, validated and where any potential weaknesses are highlighted.’ – McNulty [1]

McNulty’s framework guides a responsible storytelling practice that focuses on maintaining a research-based approach at all times.

1. Context: Outline the reasons for undertaking the analysis. Make sure the business or organisational question is clear. Outline earlier, related work. Ensure that researchers have the freedom to make conclusions based on the data.

2. Methodology: Record the methodology. If a methodology option has been chosen, explain why. Be transparent about gaps or weaknesses in the methodology and the effect these might have on the accuracy and reliability of the results.

3. Results: Perform the analysis in a way that is repeatable. Ensure that appropriate statistical standards are adhered to. Record all instances of when results do and do not meet those standards.

4. Discussion: Ensure there is a thorough critique and peer review where possible. If not possible, declare so. Compare with any other results from earlier work. Be fully transparent if conclusions can’t be directly drawn from the results. Clearly highlight where causality cannot be assumed.

5. Conclusion: If conclusions are solid and pass scrutiny, consider the most compelling way to communicate them to stakeholders. If there are uncertainties, present possible avenues for further research and refrain from over-promoting results in text or graphics.

Over to you

How do you, as a data analyst, practice responsible storytelling?

Share your answers with your fellow learners in the comments.

References

1. McNulty K. Beware of ‘storytelling’ in data and analytics. Towards Data Science; 2018 Jul 22. Available from: https://towardsdatascience.com/beware-of-storytelling-with-data-1710fea554b0