Week 8: Stats

Last week’s question set me on a wild goose chase looking for an infographic about gun ownership and gun control that made the rounds on Tumblr soon after the shootings at Sandy Hook Elementary School. I vaguely remember this infographic as being exceptionally balanced and informative, and (most rare) politically nonpartisan.

I have yet to find this now semi-mythical infographic. With any luck I will return with a supplementary post, but in the meantime, having stubbed my toes on them whilst chasing my proverbial goose, I present the two infographics flanking this post: both on the topic of gun control, touting some similar statistics*, and telling two very different stories.

The obverse of the bad surveys Glen showed us last (last) week is bad data presentation. As prof. Galey said last class, researchers’ biases can make them blind to the proverbial gorilla strolling across their data if they aren’t careful enough in formulating their questions. It also stands that those biases can lead people to leap on the statistical patterns in ball-passing, and yet ignore the gigantic gorilla walking plain as day through data that was collected with great care. Obviously, this effect is heightened if personal or political values are at stake, or if the goal has morphed from one of discovery to one of persuasion.

Infographics are an engaging and effective way to present a research essay, but they also stand at a troubling confluence where both statistical literacy and visual literacy are core skills– neither of which are actively taught or fostered in mandatory schooling (unless there have been significant changes since I graduated high school). On top of this, the infographic is a form of expression which seems accessible to non-experts. Amateur efforts abound (and occasionally succeed) without the guidance of official standards. Rarely do I see an author’s name on an infographic, or for that matter any other reliable way to assess the document’s authority except painstakingly taking a micro-comb to the list of sources (if there is enough information to follow them).

I suspect that the selfsame factors which make infographics and similar data visualizations appealing and accessible also bear the potential to subtly obfuscate and obscure instances in which data is not being presented in good faith. Infographics often seem to trade detail & accountability for accessibility. This is where I have to respectfully disagree with Glen about the language of statistics. I am in favour of conventions which support greater specificity and clarity of communication within a specialized field. Statistical language, like the language of mathematics or physics, is intended to reduce the ambiguity of the English language so that ideally** person A hears and comprehends exactly what person B intended to say. It may look like English, but it’s an illusion. When a physicist says “power”, it is not power as in “great responsibility”. What they mean is “the amount of energy consumed per unit time” as measured in Joules per second. I am not personally fluent in statistics, and I can’t speak to the potential for abuse, but it looks like a similar situation to me, that is, a specialized mathematical language which has been designed to minimize miscommunication.

The downsides of a field-specific language are clear: the alienation of many people; significant barriers to the spread of statistical literacy; the inclination toward an insular and cliquey atmosphere within the discipline. I don’t know how to bridge this gap in accessibility, but I don’t think that the answer is to dispense with specialized terminology. I will say the burden to bridge the gap should not be on the reader – perhaps statisticians need to learn how to translate back into English (proper English) as part of their training. Should detailed statistics reports be bilingual, available in both English and Stats? And might it help to incorporate a mandate for accessibility into their professional code of ethics [http://www.amstat.org/about/ethicalguidelines.cfm]?


*Similar, but not the same. As far as I can tell (the sources listed in these two infographics make me want to hug a copy of the APA style guide and whisper Shhh, it’s okay, they can’t get us through the internet– but I digress) it seems the two sets of references have only one point of possible overlap (they both refer to the CDC).

** Caveat: we’re all human here.


Leave a comment

Filed under Blogging Question

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s