This blog is for you if your work involves being able to diagnose a business challenge using data.

My recent work with market research specialists and data analysts has led to a conclusion, which is, there is a particular skill that contributes to you being able to analyse data really well.

What is it ?

Your ability to challenge or decode the findings.

Here is a wonderful story of what that looks like?

1n 1991, a first-year PhD student named Amy Edmondson now Novartis Professor of Leadership & Management at Harvard Business School began visiting hospital wards, intending to show that good teamwork and good medicine went hand in hand.

*Amy started visiting recovery rooms, talking to nurses, and paging through error reports. She noticed a lot of reported errors and she explains that the errors were not happening because there was incompetence, but because hospitals are really complicated places and there's usually a large team involved in patient care. Its easy to make mistakes due to the complexity involved.

However, some hospitals seemed more prone to errors then others. She made an observation that one department that was chatty, informal, had pictures of their kids on the wall and the other department was sedate, the nurse managers wore business suits and kept the public area free of personal items and clutter.

Amy found this an opportunity to investigate if good team work and good medicine go hand in hand

She designed a survey with suitable questions expecting that this will prove good team and good medicine went hand in hand but the survey findings shocked her, the wards with the strongest team cohesion had far more errors. She checked the data again. It did not make any sense.

She then went back to the questions in the survey and looked at each one of them. One question in the survey was " If you make a mistake in this unit, it is held against you." This response to this question revealed what was going on ?

It wasn't that that wards with strong teams were making more mistakes. Rather, it was that nurses who belonged to strong teams felt more comfortable reporting their mistakes. The data indicated that one particular norm- whether people were punished for missteps - influenced if they were honest after they screwed up.

This a wonderful example of how a data analyst should not data puke. There are 3 things that are important to tell the data story

Why are you findings the numbers peak or drop?

So what if its peaking and dropping ?

If the data analysis reveals results that are against your instincts...Challenge it.

In my recent work in Data Storytelling I have come across several situations where we have challenged the data being presented and upon investigating, found what was really going on.

Since when did data become smarter than a human?

Pic Credit: Inside Big Data

*Story Source: Smarter Faster Better by Charles Duhigg