Effective Questioning — Data Reflections

Designed in Canva by me

The great art of learning is to understand but little at a time. — John Locke

I’m working through the course, Ask Questions to Make Data-Driven Decisions. The first week focused on Problem-solving and Effective Questioning. The first prompt for reflection was, “What does data mean to you?” So, I begin this post with an introspective definition of what data means to me.

Data is a collection of the footprints of decisions and activities. These prints have the capacity to convey information about the past, present, and future. This definition, in my subconscious, relates to a person whose actions and inactions are being generated and collated. Consider going to the bank (depositing cheques, saving, and withdrawing for Christmas goodies), going to the hospital for regular checkups, or paying a visit to a pregnant spouse. All these simple actions generate data that, given the right tools, could pre-empt future actions or rationalize past decisions.

As a data analyst, when a dataset is handed to you, you must draw out meaningful insight from it. The first step is to structure the analytics process into smaller manageable chunks. This structuring could be achieved through these four activities:

• Recognize or identify the current problem or challenge
• Organize the available information.
• Reveal gaps and opportunities
• Identify your options.

Recognizing the Problem

In our bid to recognize the problem, experience has revealed that problems have similar traits and can be categorized. The common issues that analysts deal with usually require that they:

1. Make predictions, thus using data to forecast how things may be in the future.
2. Categorize things: assign information into clusters based on similar features.
3. Spot something unusual by identifying data that is different from the norm.
4. Identify themes by grouping the categorized data into broader themes.
5. Discover connections by finding similar challenges faced by different entities and combining the data and insights to address them.
6. Find patterns: use historical data to understand what happened in the past and what could happen again.

NB: Identifying the problem is the most important step in the data analysis process.

After determining the specific problem, the next step is to form questions that would direct the workflow. We learned the value of asking questions throughout the analytical process in the first article of this series; the more relevant the questions, the higher your chances of having a good influence on the business. As a result, asking effective questions would yield valuable information. The questions should be SMART. i.e.;

1. Specific: Does the question address the problem contextually? Could the questions uncover some meaningful information?
2. Measurable: Will the answers realized be quantifiable?
3. Action-oriented: Will the information lead to change?
4. Relevant: Is the question related to the problem of interest?
5. Time-bound: Given timelines, will the answers be specific to the times?

Designed in Canva by me.

While at it, these are the red flags that cloud the smartness:

• Leading questions: questions that have specific answers and have the answer suggested in the question.

Eg. You think this place is worse, don’t you?

• Closed-ended questions: Questions that require one-word answers.

Eg. Have you been living here for a very long time?

• Vague questions: questions without context or specificity.

Eg. Do you drink?

My favorite reflection for Week 1 (paraphrased)

As a junior analyst in a new role, the company requires that you do a ‘deep dive’ into the weekend sales dataset. To get started, you must ask questions to get some information. How can the questions help you to examine the time, audience, security, resources, and objectives?

My first inclination was to write out these questions:

1. What is the age range for prolific buyers?
2. What is the rush hour(s) for sales? What are the peak times when everyone is buying?
3. What are the top-selling products?
4. Are there missing items?

All of these questions would have been relevant if I had set goals that I was working towards. The idea was to get questions to form the basis for the main work goals. Hence the better approach should have been:

1. Objectives: What are the goals for this deep dive? Are there challenges I have to keep in mind?
2. Timelines: What are the timelines for this project?
3. Audience: Who are the stakeholders in this project? Who will I present the findings to?
4. Resources: What are the resources available? Do the tools and domain experts I could consult exist? Is there an extra dataset that I would need to complete this project?
5. Security: who would have access to the collected data? How long will the data be archived before ultimately being destroyed?

Find me working through the remaining weeks of this course.

Designed by me.

Resources

1. Google Data Analytics Professional Certification — Ask Questions to Make Data-Driven Decisions (Week I)
2. Canva for all the designs