Further lessons on improving your Data Interpretation skills
This is part two of the two part series on data interpretation skills that we’ve been sharing. Helping to address the imbalance that there its too much focus on just data or technical skills. Not enough advice is shared on how to interpret statistical results into a story that is relevant for your organisation. A communication that drives action as a result.
As I shared at the start of this series, these posts were written by Selby Cary & Martina Pugliese. Selby is serial entrepreneur, innovator & engineer, now Head of Technical Ops at Testcard. Martina is a senior data scientist at Shopify. So, together they bring plenty of real world pragmatism to knowing what is needed to communicate actionable interpretations.
This post builds on the first one in which Selby & Martina shared the first two lessons. Why there is so much more than just statistical results to share from your data. How important context and other aspects of interpretation are to delivering helpful output. Now in this post they move on to share the last two lessons. Empowering people through data literacy and exercising critical thinking skills…
Lesson #3 — Empower People through Data Literacy
We desperately need to improve society’s data literacy. Being able to interpret data, see its pitfalls, understand its strengths and debate its value is an essential skill in our modern world. Data helps us understand the present and better forecast the future. For many people around the world, it’s more than a tool, it’s a path to freedom.
The same piece of data can imply different things depending on your perspective, background and set of assumptions. To illustrate this point, let’s consider the phrase “we only use 10% of our brains”. There are several problems with this statement, and a little critical sense can easily spot them. Before you accept this statement as fact, you may wonder:
- What does the 10% actually refer to? Is it a fraction of the brain’s mass? Is it a proportion of the total active neurons in a given moment? If there are an estimated ~86 billion neurons in our brains, that would mean we use 8.6 billion of them.
- How has the 10% been measured, and when? Do we know more about the experiments that were carried out to generate this figure (if any)? What were their assumptions, inputs and error rates?
- Which area of the brain does this refer to? Do the same regions or neural pathways activate, or does it change? If so, how? A brain is a complex machine made of different interacting components, so it would be important to understand which areas we are talking about.
- Who is this figure referring to? Is this an average statement or is it based on a particular group of individuals? Is this figure meant to represent an average across a population or lifetime? If so, what demographics or age ranges were used?
The reality is, the 10% statistic is a long-standing myth whose origin is unclear but its popularity persists today. The habit of proclaiming statistics to promote a certain idea is everywhere — media, politics, advertisement and activism (to name a few). How many marketing campaigns (or bus slogans), especially with compelling facts, have impacted the decisions of buyers, community projects or even elections? The COVID-19 pandemic aggressively surfaced this issue, proving to be fertile ground for the spread of misinformation and a painful wake-up call for the world.
How do we combat misinformation?
Primarily, decision-makers and educators need to promote programs that enhance data literacy in schools for all age groups. In a professional sense, data science focussed regulations and organisations (similar to the ICO for GDPR) will be needed to ensure good practice for all AI systems. The latter is no easy task, especially in a rapidly evolving world, but there is progress on the horizon.
What can you personally do?
You can exercise critical thinking. Each one of us needs to be more sceptical of the data we see, share and interact with. You could start by asking “where did this figure come from?” or wondering “how did they conduct this experiment?”. It’s certainly no easy task, but acknowledging that things are more complex than they appear is a good start.
How can you empower data literacy in your business/organisation?
- Consider Hiring a Data Scientist— Sometimes, knowing the fundamentals is not enough and we need specialised assistance. To make sense of your data and inform future product development, you may need a data professional.
- Build Cross-functional Channels — It’s easy to fall into the trap of specialisation silos and forget about the importance of cross-functional teams. Providing opportunities for data scientists and engineers to mingle with other business functions is the best thing you can do. Like a jigsaw, each team is an important piece of the data puzzle.
- Make Data Accessible across your organisation — All stakeholders need to have access to accurate, consistent and reliable data to make informed decisions. Providing a centralised data source for all stakeholders will save time and money whilst enhancing the value of your product.
- Contextualise your Data — Data professionals need context around the goals they are tasked to achieve and the reasons behind them. Clear, consistent and specific goals are required to maximise your team’s efforts. As discussed in a previous blog, your data needs a purpose.
Lesson #4 — Exercise your Critical Thinking
Understanding the intricacies of a piece of data and thinking critically is challenging, even for the best of us. Debunking myths or rebutting information as fake news is a mammoth undertaking and can require specialist skills, such as neuroscience or even virology. However, we can all apply the scientific method in our questioning.
Most people today get their daily dose of news and information from social media, which favours short and snappy posts that grab your attention. Communicating a complex topic, such as cold fusion, in one sentence is no walk in the park. Especially, when your audience is the general public.
The accessibility of information has created a situation in which reading, researching, exploring and investigating is perceived as a costly endeavour. Even respected news outlets are not immune from “accidentally” publishing unchecked “news” for the sake of playing the social game and attracting an audience. This is bad for everyone.
How can you avoid falling for ‘fake news’?
- Ask Questions — Research sources, cross-reference claims, and do some due diligence before you accept something as fact. In short, spending a little time learning about the topic will go a long way! Asking questions to deepen understanding is an ancient philosophical practice, dating back to Socrates. Like many old innovations, it survived the test of time and is still the best starting point for a fruitful conversation.
- Source Quality — Don’t trust what you read or hear unless it comes from a qualified, respectable source. This could be an accredited individual or institution, ideally with peer-reviewed competencies. Obviously, this is much easier said than done, but a little digging can reduce the enormous amount of noise we swim through on a daily basis.
Being able to quantitatively interpret the world around us has always been an important survival skill. In the technology-driven world of today, data is behind every decision we make and it defines our choices. Whether we’re choosing which products to buy or whom to vote for in our local elections, we can all be swayed by the information we see. So, next time you see a fact or figure — be curious and ask questions!
More further reading for those who found this post helpful
Once again thanks to Selby & Martina for this helpful series. Once again they have shared some further reading for those who want to think further about this challenge. Here are their recommended books. I’m delighted that two are ones that I have previously recommended on this blog.
If you would like to learn more about the misuse of data and what you can do about it — we highly recommend exploring the online resources and books mentioned below (some have been referenced in the blog already):
- Calling Bullshit, data reasoning in the digital world (course, books and material), C Bergstrom, J West, University of Washington
- Factfulness, the reasons we’re wrong about the world (book), H Rosling and the Gapminder team, 2018
- How to lie with statistics (book), D Huff, 1954
- The surprising power of questions (article), HBR, 2018
(Originally published on Selby’s Scaleup Lessons blog on Medium.)