Meaning of "Problem Framing" in Data Science: High School Explanation

🎯 Learning Objectives

• 💡 Define "Problem Framing" in the context of data science.
• 🧠 Understand why problem framing is a critical first step in any data project.
• 🔍 Identify the key components involved in effectively framing a data science problem.
• 🤝 Relate problem framing concepts to relatable high school scenarios.

🛠️ Materials Needed

• 📝 Pen and paper or digital note-taking device.
• whiteboard or projector for group discussions.
• 🌐 Internet access for optional research or examples.

⏰ Warm-up (5 minutes)

Think about a time you tried to solve a problem – maybe planning a school event, choosing a topic for a science fair project, or even figuring out why your favorite sports team isn't winning. Before you could *solve* it, what was the very first thing you had to do? Share your thoughts with a partner or the class!

📚 Main Instruction: Unpacking Problem Framing

🧐 What is "Problem Framing"?

• 🕵️‍♀️ Imagine you're a detective. Before you can find the culprit, you first need to clearly understand: What crime was committed? Who was affected? What evidence might exist? This is like problem framing for a detective!
• 💡 In data science, "Problem Framing" is the crucial first step where you clearly define the specific question or challenge you're trying to solve using data. It's about figuring out the *right* problem before you start looking for answers.
• 🚧 Without proper framing, you might collect the wrong data, build a useless model, or even solve a problem that nobody cares about! It's the foundation of any successful data project.
• 🗺️ Think of it as setting your GPS destination accurately before you start driving. A wrong destination means you'll end up in the wrong place, no matter how good your car is.

🔑 Key Components of Effective Problem Framing

• 🎯 Understand the Goal: What is the ultimate business or real-world objective? What change do you want to see? (e.g., "Reduce student truancy," "Improve school lunch satisfaction.")
• 👥 Identify Stakeholders: Who are the people or groups affected by this problem and its potential solution? Who will benefit? Who needs to be involved? (e.g., students, teachers, parents, school administration.)
• 📈 Define Success Metrics: How will you know if your solution is working? What measurable outcome indicates success? (e.g., "Decrease truancy by 10%," "Increase positive feedback on lunches by 20%.") This often involves a target value, like achieving an accuracy of $X\%$ or reducing a cost by $Y$.
• ⚖️ Consider Constraints & Resources: What limitations do you have? What data is available? What's the budget, timeline, or ethical considerations? (e.g., limited access to student data, no budget for new software.)
• ❓ Formulate the Data Science Question: Translate the broad goal into a specific, testable question that can be answered with data. (e.g., Instead of "Improve grades," ask "Can we predict which students are at risk of failing an exam based on their last three quiz scores and attendance records?")

🏫 High School Examples of Problem Framing

• 📚 Academic Support:
  • 📝 Initial Goal: "Help students do better in math."
  • 📊 Framed Problem: "Can we identify students who are likely to score below a B on their next math test based on their homework completion rates, previous test scores, and participation in after-school tutoring?"
  • ✅ Success Metric: "Reduce the number of students scoring below a B by 15% in the next semester."
• 🚌 School Transportation:
  • ⏱️ Initial Goal: "Make school bus routes more efficient."
  • ⛽ Framed Problem: "How can we optimize bus routes to minimize total travel time and fuel consumption while ensuring all students are picked up within a 10-minute window of their scheduled time, using current student addresses and bus capacities?"
  • 💰 Success Metric: "Decrease total bus route duration by 10% and fuel costs by 8% per month."
• 🍎 Cafeteria Management:
  • 🗑️ Initial Goal: "Reduce food waste in the cafeteria."
  • 🍽️ Framed Problem: "Can we predict the daily quantity of specific food items that will be wasted based on historical consumption data, daily menu, and student attendance, to adjust ordering and preparation quantities?"
  • 📉 Success Metric: "Reduce average daily food waste by 20% in the next quarter."

✅ Assessment: Practice Quiz

Choose the best answer for each question:

1. Which of the following best describes "Problem Framing" in data science?
   a) Collecting as much data as possible.
   b) Clearly defining the specific question or challenge to be solved with data.
   c) Building a complex machine learning model.
   d) Presenting the final results to stakeholders.
2. Why is problem framing considered a crucial first step?
   a) It saves time in the modeling phase.
   b) It ensures you are solving the right problem and not wasting resources.
   c) It is required by most data science software.
   d) It helps you choose the programming language.
3. Which of these is NOT typically a key component of effective problem framing?
   a) Identifying stakeholders.
   b) Defining success metrics.
   c) Deciding on the specific algorithm to use.
   d) Formulating a clear data science question.
4. A school wants to "improve student engagement." Which of these is a better *framed* data science question?
   a) "How can we make students like school more?"
   b) "Can we identify factors (like club participation, attendance, and survey responses) that predict student engagement levels, and suggest interventions for low-engagement students?"
   c) "What are the most popular video games among students?"
   d) "Should we buy new computers for the library?"
5. A local charity wants to "raise more money." Which success metric is most appropriate for a data science problem aimed at this goal?
   a) "Increase total donations by 15% in the next fiscal year."
   b) "Have more people attend our fundraising events."
   c) "Get more likes on our social media posts."
   d) "Send out more email newsletters."
6. You're trying to frame a problem about "reducing tardiness for first period." Which of the following is a relevant constraint to consider?
   a) The color of the school lockers.
   b) The budget available for new attendance tracking technology.
   c) The history teacher's favorite subject.
   d) The number of trees on school grounds.
7. Transform the broad goal "Make the school website more useful" into a well-framed data science question.
   a) "What kind of pictures should we put on the website?"
   b) "Can we analyze user click-through rates, time spent on pages, and search queries to identify underperforming sections and suggest improvements for navigation and content relevance?"
   c) "How many pages should the website have?"
   d) "Which web designer should we hire?"