Analytical thinking and analytical reasoning are two concepts that are often misunderstood. Many people think that they are the same thing, but this is not the case. In fact, analytical thinking and analytical reasoning are two very different things, however, related. Analytical thinking is an important aspect of analytical skills. Most of us do not realize how to use analytical thinking and often end up solving the problem incorrectly or half-heartedly. As data analysts or data scientists, it would be of utmost importance to acquire this skill well. In this blog post, we will learn these concepts with the help of some real-life examples.
What’s Analytical Thinking?
Before we get into understanding what is analytical thinking, lets understand the word, analysis, which forms the word, analytical.
The word “analysis” comes from the Ancient Greek ἀνάλυσις (analysis, “a breaking-up” or “an untying;” from ana- “up, throughout” and lysis “a loosening“).
From above, it can be comprehended that loosening anything or any problem up can said to be analyzing the thing or the problem. And, loosening a problem or a thing can be represented as breaking down the problem or thing into further components (sub-problems or sub-things).
Given above, let’s understand what is analytical thinking?
Analytical thinking is the process of thinking about any topic or an issue / problem which involves breaking down a problem / issue or a topic into smaller parts in order to better understand it in a better manner. When working with a problem or an issue, analytical thinking helps find great solutions. When trying to understand a topic, analytical thinking helps understand topic and related concepts in a better manner.
Analytical thinking fits in very well with the first principles thinking in the sense that reasoning from first principles requires you to break down a thing, an idea or a problem into its most basic elements to know the problem / idea / thing well. Thus, one would need to do analytical thinking or be analytical to do first principles thinking. This can be done by asking questions such as some of the following:
- What is the problem which needed to be solved? For example, how do we run the schools in Covid times in a safe manner? How do we increase sales? How do we achieve cost savings target in procurement? How do we reduce loss due to credit card fraud transactions?
- Are there sub-problems which when solved can help solving the problem as a whole? For example, the covid problem can be broken down into sub-problems related to vaccination, masking, sanitization, social distancing, covid communication etc. Similarly, the challenge of achieving cost savings in procurement can be broken down into avoiding pricing deviations, increasing reuse, proper budget planning, proper inventory planning, etc. These can further be broken down into its most basic elements based on first principles thinking.
- Why are we trying to solve the problems and/or sub-problems? What is going to be the ultimate change which will happen as a result of solving the problem? For example, the schools opening up will prove to be good for mental and physical well-being of the students. The cost saving in procurement will result in positive impact on the bottom line of the company.
- What constitutes the core of the problem/sub-problems? Who can create impact and who will be impacted? For example, the school and its students, teachers, staff members, parents, etc. In first principles thinking, this is also called as efficient cause. It is important to understand this aspect as solution could also be related to changing the way the end user deal with the situations related to the problem space.
- What processes will be impacted? Or, what processes will need to be changed?
Here is the diagram representing the above:
You may want to check a related post which will help you break the problem into sub-problems using first principles thinking. Here is the post titled – First principles thinking explained with examples.
Asking questions is key to analytical thinking
Asking questions holds key to the analytical thinking or breaking down the problem or thing into smaller components (sub-problems). The questions can be of different forms such as the following:
- Assumption questions: Assumption questions are a type of question that seeks to determine what is presumed or understood about the topic being discussed. An assumption question is a way of getting to the heart of a situation, examining underlying assumptions and beliefs. They can be used to uncover difficult issues, understand opinions and perspectives, and work out solutions to problems. Cartesian method of doubt can be adopted for assumption questions. Questions can be asked with the phrase “for certain”. The following are some of the examples of assumption questions:
- What do we assume about this situation?
- What assumptions have been made here that may need to be challenged?
- What are the underlying assumptions behind this decision?
- Clarification questions: Clarification questions are inquiries that are used to gain further information about a topic or situation. They help with analytical thinking by providing a clear direction for the conversation and allowing for deeper insight into the subject at hand. Clarification questions can also help to ensure that all parties involved in an exchange have the same understanding of the topic, thus improving communication. Here are a few examples of clarification questions:
- Can you explain what you mean by ABC?
- What do you think the implications of XYZ could be?
- How does PQR compare to other similar topics?
- What other perspectives should be considered in relation to EFG?
- Consequence questions: Consequence Questions are a type of analytical thinking practice that helps to uncover deeper issues related to a problem or question. These are the questions which help understand the “so what” related to the problem. They ask what the potential consequences of an action might be, encouraging creative and outside-the-box thinking. This type of questioning encourages people to think beyond just the immediate effects but instead consider long-term, downstream effects – both positive and negative – that may be associated with a given idea or action. By doing this, it can help individuals identify opportunities for improvement or new innovation that they hadn’t thought of before. Here are a few examples of consequence questions:
- How could this decision impact our business in the long-term?
- What would be the consequences if we implemented this solution?
- How will our customers react to this change?
- What unintended outcomes may result from us taking this approach?
What’s Analytical Reasoning?
Analytical reasoning is the process of forming conclusions from given premises by applying valid inference rules while considering different interpretations or perspectives on them. It is the ability to use logic and reason to solve problems. It helps us use what we know to figure out what else might be true. This can be done by following the processes. The steps will be described using the example.
- Set hypothesis: First and foremost, one needs to set a hypothesis. For example, walking 5 KM on daily basis reduces risk of heart attack.
- Figure out reasons; The reason for reduction in risk of heart attack is the reduction in cholesterol to acceptable limits, or, keeping the cholesterol well within acceptable limits.
- State general proposition with examples: State general proposition which is universally accepted as true with the help of examples. The cholesterol level well within the universally accepted limit is found to be indicator of good health of heart. As a data scientist, one can perform hypothesis testing at this stage to sight multiple examples. You may want to check one of my related posts: Hypothesis testing explained with examples.
- Re-instate the general proposition to the current hypothesis; Walking 5 KM a day keeps the cholesterol level well below acceptable limits. This can be outcome of the hypothesis tests you performed in the previous step.
- Draw the conclusion from the previously stated general proposition. Walking 5 KM on daily basis reduces the risk of heart attack.
The above reasoning process is inspired by Nyaya Syllogism from Nyaya philosophy (One of the schools of Indian Philosophy).
Here is another simpler and common example of analytical reasoning:
- Set hypothesis: The house is burning with fire
- Figure out reasons: There is smoke coming out of house
- State the general proposition with examples: When there is a smoke, there is a fire. For example, kitchen
- Re-instate the general proposition to the current hypothesis: The house is having lot of smoke.
- Draw the conclusion: The house is burning with fire.
Analytical Thinking Sample Use Case – Problem of Traffic
In this section, we will take a look at the approach we take to understand / analyze the problem of traffic in the city during peak hours.
To break down the problem of traffic in the city during peak hours, we can start by asking a few key questions and identifying sub-problems. Here are some possible sub-problems to consider:
- What causes traffic during peak hours?
- Are there specific areas or routes where traffic is particularly bad?
- What is the impact of traffic on commuters, businesses, and the environment?
- Are there any existing solutions or initiatives in place to address traffic during peak hours?
- What are the trade-offs or unintended consequences of different solutions?
To further analyze these sub-problems, we can ask a range of questions. Here are some examples of assumption, clarification, and consequence questions that might be helpful:
Assumption questions:
- What assumptions are we making about the causes of traffic during peak hours?
- Are there any biases or preconceptions we need to be aware of when analyzing this problem?
- Are there any assumptions we’re making about the impact of traffic on different stakeholders?
- What assumptions are we making about the behavior of drivers and commuters during peak hours?
- Are there any assumptions we’re making about the types of vehicles or transportation modes that contribute to traffic during peak hours?
- What assumptions are we making about the impact of new technologies, such as autonomous vehicles or ride-sharing platforms, on traffic during peak hours?
Clarification questions:
- What data do we have about traffic patterns and congestion during peak hours?
- What are the specific pain points for commuters and businesses when it comes to traffic?
- Are there any regulations or policies that are contributing to or exacerbating traffic during peak hours?
- What are the key factors that contribute to congestion during peak hours, such as bottlenecks, intersections, or merging points?
- How does public transportation, such as buses or trains, impact traffic during peak hours?
- Are there any data or studies on the impact of remote work or flexible schedules on traffic during peak hours?
- What is the impact of school start and end times on traffic during peak hours?
Consequence questions:
- What is the ultimate goal of solving this problem?
- What are the benefits of reducing traffic during peak hours for different stakeholders?
- What are the costs of not addressing this problem?
- How might solving this problem impact the environment, public health, and safety?
- What are the long-term implications of reducing traffic during peak hours for the city’s infrastructure and economy?
- What are the potential unintended consequences of different solutions or initiatives?
- What are the trade-offs involved in implementing different solutions?
By breaking down the problem into sub-problems and asking these types of questions, we can gain a more comprehensive understanding of the issue and identify potential solutions to alleviate traffic during peak hours.
Presentation on Analytical Thinking
The following slides / presentation is an introduction to analytical thinking. Analytical thinking is characterized by the ability to break down complex information into smaller, more manageable parts to better understand it. The presentation covers the characteristics of analytical thinkers, the analytical thinking process / workflow, the benefits of analytical thinking and what should one do for developing and improving analytical thinking skills.
Analytical Thinking with ChatGPT / Generative AI Tools
Ever wondered how analytical thinking could take your interactions with ChatGPT and other generative AI tools to the next level? 🤔 Brace yourself for an exciting exploration of the dynamic relationship between analytical thinking and these cutting-edge AI technologies. In this section, we’ll look at some of the prompts which can help you do analytical thinking while leveraging generative AI tools such as ChatGPT.
The following are some of the example prompts to give you an idea on how could you start doing analytical thinking around a topic or a problem / issue.
The following prompt can be used with code interpreter in ChatGPT 4. You can upload a CSV file consisting of customer reviews or a feedback.
Act as a data analyst who is tasked with analyzing customer feedback and reviews to uncover insights about the reasons for the decline in sales.
Utilize analytical thinking to categorize and analyze customer sentiments, complaints, and preferences related to the company’s products or services. Look for recurring themes or issues that customers have been expressing to identify areas that require immediate attention. Leverage sentiment analysis to extract valuable insights from unstructured data sources. Based on your analysis, create a comprehensive report consisting of appropriate plots / charts highlighting the key customer pain points and propose data-backed strategies to address these concerns, ultimately revitalizing the company’s sales performance.
Conclusion
Analytical thinking and analytical reasoning are two very different things. In this blog post, we learned the concepts with the help of some real-life examples. Analytical thinking is about breaking down a problems or issues into smaller parts in order to better understand it and find solutions. The steps involved include asking questions such as: “What is the problem which needed to be solved?”; “Are there sub-problems that need to be solved as well?”; “Why are we trying to solve these problems? What will happen if we do successfully solve them?”, etc. Whereas, analytical reasoning includes forming conclusions from given premises by applying valid inference rules while considering various interpretations or perspectives on those premises. Please feel free to share your thoughts or suggestions or ask questions for clarifications.
Check out my latest book titled as First Principles Thinking: Building winning products using first principles thinking.
- Agentic Reasoning Design Patterns in AI: Examples - October 18, 2024
- LLMs for Adaptive Learning & Personalized Education - October 8, 2024
- Sparse Mixture of Experts (MoE) Models: Examples - October 6, 2024
I found it very helpful. However the differences are not too understandable for me