Review - Part One - Telling Stories With Data


This is one of a series of reviews that I will be writing for an ethics and data science reading course in which I am currently participating. This post specifically is on part one of ‘Telling Stories With Data’ by Rohan Alexander.

Data contain a story. The numbers are not just purely objective measures of reality. They are a reflection of the way in which we understand and interact with the world. It is important to understand when communicating using data that we are presenting the story within that data. Or, at least a story. There are important things then to take into consideration when telling a story of data to ensure that an accurate story is being told. Not just the story we wish to be told.

Part one of ‘Telling Stories With Data’ presents a well-rounded introduction to the concepts to be considered and how to consider them. Beginning with an apt comparison to writing fiction, Rohan presents that when we write quantitative analysis, we consider the elements of character, plot, setting, theme, and style. That in the case of quantitative analysis these concepts are analogous to what is the data, who generated it and how (who is/are the characters involved)? What is the data trying to say, how can we let it say this (what is the plot of the story)? What is the broader context of the data, where and when was it generated, could other data have been generated/included (where and when is the setting of the story, how does that setting influence the story being told, could it be told elsewhere)? What are we hoping others will see from this data (what is/are the theme/s of the story, what do we hope people take away)? And, how can you convince them of this (how is the story told, through whose voice and perspective)?

I would add onto this to say study your favourite books, films, artwork, and designs. Study your favourite creative pieces. Critically examine how they communicate meaning through story and how convincing they are of the points that are being made. Critically examine why you were drawn into some creative productions more than others. What kept you from being immersed in the story and that world? Was it the lack of coherent plots, was it the lack of developed characters, or was it an undeveloped setting? Study logos and children’s media. These are exemplars of communicating in a simple and direct manner. Logos need to be able to communicate what the company or organization represents in a single illustration. Children’s media needs to present the often complicated concepts in an uncomplicated way so as to be easily understood. Look at Pixar’s work or the film ‘The Song of the Sea’ for excellent examples in doing this. Study graphic novels and comics. Look at Gary Larson’s ‘The Farside’ for communicating a story in simple, one-panel comics. Look at Jeff Lemire’s ‘Essex County’ for meaningful stories told through illustration. And, look at ‘The Arrival’ by Shaun Tan for an example of communicating without words. This last being particularly useful for understanding data visualization.

It is crucial to understand that with data we must tell the story the data is telling, not the story we wish to tell with the data. Do not fit the evidence to your narrative, let the evidence guide the narrative. The legendary baseball player Henry “Hammerin' Hank” Aaaron passed away this January 22nd. In many memorial messages towards Mr. Aaron he was remembered for his grace in the face of racist adversity in his chasing and surpassing Babe Ruth’s home run record. That he ignored the racism levelled against him with “stoic dignity”. However, in his autobiography, Mr. Aaron writes of how the experience in surpassing Ruth left a foul taste in his mouth and tarnished baseball. While what is presented is the mythologized image of a Black man confronting racism, much in the way the image of Jesse Owens has been used against Nazi superiority, with stoicism, what in actuality is happening is the experiences of a Black man being erased for a narrative. The data are being used to tell a story, while the story of the data is not being told.

What I’ve written here is more an addition to what was included in the introductory chapter, or my thoughts as I read through the introduction. So, less a review and more a response I suppose. As a review I will say this. I know what to expect in the notes from the introductory chapter. I know how I should be thinking of data and data science. I know what software will be used and where to get it. Other than a few areas of grammar, and who am I to really consider that a short-coming, this is a strong introduction to the topic of telling stories with data. As a final note, the header colours are awesome.

I’ve linked to ‘Telling Stories With Data’ here as well as at the beginning of this post. Please, whoever reads these other than Rohan, go check it out. It’s an important topic to explore and understand. Even if you don’t use R, this is a great way of learning about the software, and the concepts you’ll learn are universally applicable.