How can you encourage your team to think critically about data?

The biggest mistake analysts make, is to take the first question asked and dive into figuring out the "how" (e.g. writing the SQL statement). Seldom is the first question asked the one worth analyzing. I usually do one of the following things: a. I apply the 5x Why Framework (in a playful way, bluntly asking why won't do you any good.) b. I ask "If I get you this piece of data, what action will you take upon it"? On average, I would say 80% of questions asked are not worth spending time on. If you learn to filter out that 80% you become a more impactful analyst straight away.

You don’t know what you don’t know. Make sure you are a subject matter expert as your knowledge will generate the right questions and constructive curiosity. Challenge assumptions, compare sources, and connect the dots.

Encourage your team to think critically about data by fostering a culture of open dialogue where questions are welcomed. Provide training on analytical techniques and regularly challenge assumptions to promote a thorough examination of data sources. Celebrate and reward instances where critical thinking leads to valuable insights, fostering a mindset that prioritizes thoughtful analysis.

Your team needs to be able to examine the data with questions and an understanding of the limitations rather than accepting it at face value. This involves engaging your team in the data collection and analysis process and explaining the methodologies used so they can approach the data with scrutiny. I personally teach my team of medical students about the basics of statistics so they also know the limitations of various tests.

Data analysis can be approached in two ways: observational and experimental. In my experience I have found clarity regarding the context to be crucial, as posing inaccurate questions can steer the analysis in an incorrect direction.

Assessing the statistical significance of results is the optimal approach for evaluating various types of data analyses. While a significance value of 0.05 or lower is commonly considered a standard, it may vary depending on the analyst's industry-specific understanding.

There are several important factors when it comes to evaluating evidence: 1. Understanding how to do a thorough literature search and find a comprehensive set of high quality evidence on a certain topic 2. Understanding what constitutes high quality evidence 3. Understanding biases and fallacies which may influence your literature search and the strength of the evidence you’ve found such as confirmation bias

I agree that data presentation is only as good as the presenter. I’ve attended many medical conferences where innovative technology which improved patient outcomes was hardly noticed by anyone because the presentation was long, confusing, and boring. When presenting data in a relatable way it is so important to practice your presentation and also to give your presentation to people outside your workplace or your professional community to get meaningful feedback. Those who know your work will know what you’re trying to say. Those who are totally unfamiliar rely on your communication skills to understand the value of your data analysis and can help you bridge gaps between what you want to communicate and what you’re actually communicating.

I personally find the DIKW (Data, Information, Knowledge, Wisdom) model quite useful to explain the process of extracting insights and value from data. As marketers we collect so much data that is easy to fall victim of data and analysis paralysis. Making your data useful requires putting that data into context and explaining what's happening. What patterns and correlations can you see? This step is very important before being able to interpret the data and turn it into insights (Wisdom). Using this data to tell a story is key. This means explaining the why behind results and weaving in context to the information you're sharing. Only then you'll be able to provide meaningful and actionable insights.