Formulating Experimental Questions

TL;DR

Experimental questions guide your entire scientific investigation by clearly defining what you want to find out. A good experimental question focuses on a relationship between variables, is specific, measurable, and answerable through an experiment. Think of it as a clear roadmap for your scientific journey.

1. The Mental Model

Imagine you're trying to solve a puzzle. Your experimental question is like knowing exactly what picture you're trying to build. Without it, you're just randomly fitting pieces together; with it, you have a goal and a direction.

2. The Core Material

Formulating a good experimental question is crucial because it sets the foundation for your experiment. It helps you design your procedure, identify your variables, and analyze your results effectively.

What makes a good experimental question?

Photo by Ann Hannes on Pexels

A good experimental question often follows the "PICO" framework (sometimes adapted as "PICOT" or "FINER" in different contexts, but PICO gives a great start for science experiments):

• Population/Problem: Who or what are you studying?
• Intervention/Independent Variable: What are you changing or testing?
• Comparison: What are you comparing it to (if anything)?
• Outcome/Dependent Variable: What are you measuring as a result?

Also, think of it as being SMART:
* Specific: Is it clear and focused?
* Measurable: Can you actually measure the results?
* Achievable: Can you do the experiment with the resources you have?
* Relevant: Is it important or interesting?
* Time-bound: Can you do it within a reasonable timeframe? (Less critical for basic questions, but good for planning).

Independent and Dependent Variables

At the heart of an experimental question are your variables:

• Independent Variable (IV): This is the one thing you change or manipulate in your experiment. It's what you are testing.
• Dependent Variable (DV): This is what you measure or observe. It's the outcome that changes because of what you did to the independent variable.

A good question usually explores the relationship between the IV and the DV. Often, it can be framed as "How does [Independent Variable] affect [Dependent Variable]?" or "What is the effect of [Independent Variable] on [Dependent Variable]?"

Examples of good vs. bad questions:

Photo by Ann Hannes on Pexels

• Bad: "Do plants grow?" (Too general, not an experiment, just an observation.)

• Good: "How does the amount of sunlight affect the height of bean plants after three weeks?" (Specific IV: amount of sunlight; Specific DV: height; Population: bean plants; Time: three weeks.)

• Bad: "Is sugar good for you?" (Too subjective, broad, and difficult to measure experimentally in a simple way.)

• Good: "Does adding 5g of sugar daily to a rat's diet affect its body mass over four weeks compared to a control group receiving no added sugar?" (Specific IV: 5g sugar daily; Specific DV: body mass; Population: rats; Comparison: control group; Time: four weeks.)

The Question Formulation Process

Photo by Leeloo The First on Pexels

Here's a visual flow of how to think about formulating your question:

graph TD
    A["Have a general idea or observation"] --> B["Identify potential variables (things that can change)"]
    B --> C1["What can I change? (Independent Variable)"]
    B --> C2["What can I measure as a result? (Dependent Variable)"]
    C1 --> D{"Are these variables specific and measurable?"}
    C2 --> D
    D -- Yes --> E["Draft the question: How does IV affect DV?"]
    D -- No --> B
    E --> F{"Is the question SMART (Specific, Measurable, Achievable, Relevant, Time-bound)?"}
    F -- Yes --> G["Refine question for clarity and conciseness"]
    F -- No --> E
    G --> H["Final Experimental Question"]

3. Worked Example

Let's say you're interested in how different types of fertilizer impact plant growth.

1. General idea: Fertilizer makes plants grow.
2. Potential variables: Type of fertilizer, amount of fertilizer, type of plant, amount of water, sunlight, temperature, plant height, number of leaves, plant mass.
3. Identify IV & DV:
   • Independent Variable: Type of fertilizer (e.g., nitrogen-rich vs. phosphorus-rich vs. no fertilizer).
   • Dependent Variable: Plant growth (e.g., plant height, number of leaves).
4. Draft a question: "How does the type of fertilizer affect plant height?"

5. SMART check and refinement:

   • Specific: "Type of fertilizer" is good. "Plant height" is good. But "plant" is vague. How long will the experiment run?
   • Measurable: Yes, you can measure height.
   • Achievable: Yes, if you have plants and fertilizers.
   • Relevant: Yes, for gardening or agriculture.
   • Time-bound: Needs to be specified.

   Refined Question: "What is the effect of using nitrogen-rich fertilizer versus phosphorus-rich fertilizer on the average height of tomato plants after four weeks, compared to plants receiving no fertilizer?"

   • IV: Type of fertilizer (nitrogen-rich, phosphorus-rich, none)
   • DV: Average height of tomato plants
   • Population: Tomato plants
   • Comparison: Plants receiving no fertilizer
   • Time: Four weeks

This question is clear, focused, and directly guides how you'd set up your experiment.

4. Key Takeaways

• An experimental question defines the goal of your entire investigation.
• It focuses on the relationship between an Independent Variable (what you change) and a Dependent Variable (what you measure).
• Use the "How does [IV] affect [DV]?" format as a starting point.
• A good question is SMART: Specific, Measurable, Achievable, Relevant, and helps set a Time-frame.
• It should clearly state who or what you are studying and any comparison groups.
• Your question must be something you can actually investigate through an experiment.
• Refining your question before starting saves a lot of time and effort later.

Common Mistakes to Avoid

• Too broad or vague: Asking "How do chemicals affect people?" is unanswerable.
• Not testable: Asking "Is the sky pretty?" is a subjective opinion, not an experiment.
• Only one variable: A good experiment needs both an independent and a dependent variable to explore a relationship.
• Confusing IV and DV: Accidentally trying to change what you should be measuring, or vice-versa.
• No clear population or time frame: Not specifying what you're studying or when makes the question ambiguous.

5. Now Try It

Think about a common observation you've made, like "listening to music helps me focus." For 15 minutes, try to formulate a testable experimental question around this idea.

1. Identify the potential Independent Variable (what you would change).
2. Identify the potential Dependent Variable (what you would measure).
3. Draft a question using the "How does [IV] affect [DV]?" structure.
4. Refine your question to be SMART: specific, measurable, achievable, relevant, and with a time frame. Consider who (or what) you'd test and what you'd compare it to.

Success looks like a single, clear question that you could realistically start designing an experiment for. For example: "What is the effect of listening to instrumental music on the number of correct answers on a 15-minute math test for high school students, compared to students who study in silence?"