Alhamdulillah, completing the Storytelling with Gender Data Fellowship gave me reason to be grateful to the mentors and fellow participants who shared the journey. It also gave me a stronger interest in the relationship between evidence and a story. I do not want figures merely to decorate what I already believe. If the evidence raises an uncomfortable question, that question belongs in the work too.
Define the question before choosing the chart
A question about change over time requires comparable observations from different periods. A question about distribution requires information about how something is shared. A question about individual experience may need interviews as well as numbers. The method should follow the enquiry rather than the available visual format.
The scope also matters. A national dataset may not support a claim about one local community. A small local study may not establish a national pattern. A story can still be valuable at its actual scale. Enlarging the claim beyond the evidence usually makes the piece less reliable rather than more important.
Let evidence complicate the opening idea
Good research may reveal that the first interpretation was too simple. A trend may have exceptions. Two sources may measure different things. An apparent improvement may reflect a change in reporting. The writer should allow these findings to alter the story, even when the original title was more dramatic.
Consider a hypothetical article about access to a training opportunity. Registration figures might show interest, while completion records reveal a different pattern. Interviews could help identify questions about scheduling or access. None of these pieces of evidence should be asked to do all the explanatory work alone. Their value grows through careful comparison.
Human examples need the same discipline. A real person's account can make a pattern understandable, but the writer should establish what it represents and obtain appropriate permission. An invented quotation presented as testimony would undermine the entire piece. Where an example is hypothetical, the reader should be told directly.
The working title should remain provisional until the evidence has been examined. This reduces pressure to defend an early framing. A less dramatic title that accurately describes the finding serves readers better than a memorable one requiring the article to overstate its case.
Make the conclusion no larger than the finding
The article should explain what the evidence supports and what remains uncertain. A useful ending may identify a practical question for institutions or a need for better information. It does not always need a sweeping prediction or an urgent demand. Sometimes the strongest contribution is making a neglected issue easier to examine.
Sources should be close enough to the relevant claims for readers to follow them. A long list of links at the end does not help if it is unclear which source supports which statement. Clear attribution is part of the explanation, not an administrative attachment to it.
A strong data story leaves readers with a more accurate understanding than they had before. The chart, narrative, and examples should serve that purpose together. Before asking whether a number is striking, ask whether it helps answer the question. That is where the story's responsibility begins.
0 Comments