Data Analysis is just an interview

Data Analysis is just an interview

Knowledgeable? or just gathering information?

Nowadays, it’s very easy to be overwhelmed by the huge amount of information thrown at us through the internet, but surprisingly, even though the information is widely accessible to almost everyone, very few people are knowledgeable.

Being knowledgeable about a subject is understanding it at a deeper level and being able to connect it with other concepts and explaining it in different formats. It’s like being able to answer consecutive Whys until you reach a solid common knowledge.

I thought a lot about how to learn and transfer the skill of data analysis ( DA ), and by DA I don't just mean simply visualizing data and statistics, I mean extracting actionable insights from the data. Today’s topic is Data Analysis and how we can leverage it to build knowledge.


What makes DA a frustrating process ?

The most common way we follow in data analysis is to start by visualizing the data and state the obvious. We also zoom into the data using filters and use different tools to reveal interesting findings. Even if we do so, it’s still a very chaotic and non justified method that can easily be biased.

The dangerous part is that it’s very frustrating and stressful to do it that way, it feels like trying to google something even though you don’t know if it exists.

A change in perspective

An interesting perspective to look at DA from is to imagine yourself interviewing a your data as if it was a guest in your podcast, or maybe a criminal. You need to know as much information about it as possible, you’ll do this by asking questions, rephrasing answers and asking follow up questions to deepen your understanding.

Example

Imagine you have a dataset containing the grades of students, and you start talking to your data :

  1. What’s the student rankings in Maths ?
  2. What are the common factors of having good grades in Maths ?
  3. Are there students whose parents are engineers but have bad grades ?
  4. What are the weaknesses of students who are good in Maths ?
  5. ....

Every question is either generating or verifying a hypotheses ? and once we can extract an interesting insight we can easily backtrack to the series of questions which led to the answer in order to present the finding with a story.

Mistakes to avoid

Assumptions

It’s very common for data analysts to stop questioning the data by relying on personal assumptions. Assumptions are the enemy of truth. Sometimes, it’s difficult to dig deeper due to the lack of data. In this case, we continue the questioning by seeking support from experts or by referencing a relevant finding.

Bias

It’s very common for data analysts to stop questioning the data by relying on personal assumptions. Assumptions are the enemy of truth. Sometimes, it’s difficult to dig deeper due to the lack of data. In this case, we continue the questioning by seeking support from experts or by referencing a relevant finding.


It’s all about perspective, viewing data analysis as a simple interview brings more joy and peace to the process. Also, it makes it easier for people to learn this skill and develop a passion for it.

I hope this article added value to you, do not hesitate to read more on the blog and share with your friends. Thank you for your time.