Investigation is a central part of MYP Science. Through investigations, you learn to frame a question, design a method, collect and analyse data, and draw a conclusion that the evidence supports. This guide explains how to approach each stage of an MYP Science investigation and how to produce work that meets the criteria that assess it. It applies to the current IB MYP Sciences framework at all year levels, with the understanding that the demands rise as you move through the programme.

A note on scope. This guide uses the current IB MYP Sciences assessment criteria. Investigations draw most directly on Criterion B (Inquiring and designing) and Criterion C (Processing and evaluating). Investigation tasks vary by school and year level, so confirm the specific requirements of your investigation against your school's instructions and the current IB MYP Sciences subject guide. A tutor can help you develop the skills investigations require; a tutor cannot complete assessed investigations for you.

Developing a Research Question

A strong investigation starts with a strong question. A good research question is focused, clear and investigable: it states what you are investigating and how. "How does X affect Y?" is a workable frame; "the science of water" is not. To develop a question:

Start from something you genuinely want to know. Investigations are easier to do well when the question interests you.
Narrow it. Broad questions cannot be investigated in the time and with the equipment available; a specific question can.
Make it testable. You should be able to imagine what data would answer it.
State it clearly. A reader should understand your question from a single sentence.

A weak question produces a weak investigation, no matter how carefully the rest is done. Spend the time to refine the question before you commit to a method.

Formulating a Hypothesis

Where the task asks for one, a hypothesis is a reasoned prediction, not a guess. A strong hypothesis:

Predicts a relationship between the independent and dependent variable.
Is grounded in the science, with a brief explanation of why you expect the relationship.
Is testable, in the sense that data could support or contradict it.

A hypothesis is not a statement of what you hope to find; it is a prediction you will test against evidence. Being wrong is not a failure of the investigation; it is information the investigation produces.

Identifying Variables

A clean experimental design rests on clearly identified variables:

Independent variable. The variable you deliberately change. It should be measurable and span a sensible range.
Dependent variable. The variable you measure in response. It should be measurable with adequate precision for the relationship you hope to detect.
Controlled variables. The variables you keep constant so that any change in the dependent variable can be attributed to the independent one. Weak control of variables is one of the most common reasons a conclusion becomes unreliable.

Identify your variables before you write the method. The method is the plan for changing the independent variable, measuring the dependent variable, and holding the controlled variables constant.

Planning the Method

A good method is one that produces data precise enough to answer the question. When planning:

Choose appropriate measuring instruments with enough precision for the relationship you hope to detect.
Decide the range and number of values of the independent variable. Too few values leaves the relationship unsupported; too many can be impractical.
Plan repeats. Repeating measurements and reporting a mean improves the reliability of your data and gives you something to discuss in the evaluation.
State how variables are controlled. For each controlled variable, say how you keep it constant.
Plan for safety and ethics. Consider any risks and how you will manage them.
Do a pilot if you can. A short trial run often reveals practical problems that are easy to fix early and expensive to fix late.

A method that cannot be carried out with the time and equipment available produces no data. Feasibility matters as much as ambition.

Observations and Measurements

During the investigation, record observations and measurements systematically. Useful habits:

Record raw data in a clear table with headers that include units.
Record what you observe, not just what you expected. Unexpected observations are often the most informative.
Note conditions that might affect the data, such as temperature or other environmental factors.
Do not discard outliers without reason. If a reading looks wrong, investigate why; do not silently drop it.

Careful recording at the time is cheaper and more reliable than trying to reconstruct observations later.

Working With Data

Once you have raw data, the next step is to process it into a form that reveals the relationship. Processing might include calculating means, converting units, computing derived quantities or linearising data. At each step:

Explain what you are computing and why. A column of processed numbers with no explanation is hard to follow and easy to mark down.
Track units through calculations. Unit errors are a common and avoidable source of mistakes.
Keep appropriate precision. Report values to a number of significant figures consistent with the precision of the measurement.
Round only at the final step unless instructed otherwise, to avoid premature rounding changing later results.

Using Graphs

A well-chosen graph is often the clearest way to reveal a relationship. When graphing:

Plot the dependent variable on the vertical axis and the independent variable on the horizontal axis.
Label both axes with the quantity and its unit. A graph without units is incomplete.
Choose a scale that uses most of the available space and is easy to read.
Draw the best-fit line carefully, and do not force it through the origin unless the science demands it.
Use the graph to identify the relationship, linear, proportional, inverse, or otherwise.

Where you want to test a relationship, a linearised graph, plotting a derived quantity so that a straight line is expected, is often the clearest way to confirm it and extract a meaningful gradient or intercept. Explain what the gradient and intercept represent in scientific terms.

Interpreting Results

Interpretation is the move from data to scientific meaning. It is not the same as describing the data. A strong interpretation goes: data, relationship, scientific meaning.

Start by identifying the relationship in your processed data, what the graph shows. Then interpret that relationship in scientific terms: does it match what the underlying science predicts? If your data suggests a proportional relationship, say so and connect it to the theory. If the relationship deviates, explain why that might be. Avoid generic lab-report phrasing such as "the graph goes up"; say what the shape means for the science you are investigating.

Drawing Conclusions

A conclusion answers the research question, supported by the evidence. A strong conclusion:

Answers the question that was asked.
Is supported by the data, including its uncertainty, not stronger than the evidence allows.
Connects to the science, explaining what the result means.
Acknowledges limits where the method or data has them.

A conclusion stronger than the evidence is a common and costly mistake. If the data does not support a strong claim, make a weaker claim that the data does support.

Evaluation

Evaluation is where you judge the quality of your evidence and your conclusion. A strong evaluation discusses:

Limitations. What constrained your investigation: equipment, precision, time, uncontrolled variables.
Sources of uncertainty. Where the main uncertainties came from, and how large they are relative to the effect you measured.
Impact on results. How those limitations and uncertainties actually affected your data and conclusion, not just that they existed.
Realistic improvements. Specific, practical changes that would reduce the uncertainty or address the limitation. "Use better equipment" is weak; "use a motion sensor instead of a stopwatch to reduce reaction-time error" is specific and useful.
Whether the conclusion is supported. An honest statement of how far your evidence supports the conclusion you drew.

Genuine evaluation is reflective and specific. A short paragraph listing generic "human error" earns little because it shows no understanding of what actually limited the investigation.

Suggestions for Improvement

The improvements you suggest should be specific and practical, not generic. For each limitation you identify, ask: what specific change would reduce this limitation, and how much would it improve the data? An improvement that is specific, justified and feasible earns more than a list of "use better equipment."

How Investigations Vary by School and Year Level

Investigation tasks are not the same everywhere. Schools design their own investigations within the MYP framework, so the topic, the scope, the time allowed and the emphasis vary. The demands also rise across year levels: an investigation in MYP 1 is a simpler task than an investigation in MYP 5, and the level descriptors expect more of older students.

Before you start, confirm with your school:

What the investigation is and what it is assessing.
How much time you have and what equipment is available.
Which criteria the investigation will be assessed against.
What your school's policy is on collaboration, drafts and feedback.

Common Investigation Mistakes

Research question too broad. A question that cannot be investigated in the time and with the equipment available.
Weak control of variables. Uncontrolled variables that make it impossible to attribute changes in the dependent variable.
Insufficient data. Too few values of the independent variable, or no repeats, leaving the relationship unsupported.
Unexplained processing. Derived quantities or calculations with no stated method or justification.
Graphs without interpretation. A plotted graph that is never connected to the science or the conclusion.
Generic evaluation. "Human error" and "use better equipment" with no specifics.
Conclusions stronger than the evidence. Claiming a relationship is confirmed when the data and uncertainty do not support it.

How MYP Science Tutoring Can Help

Legitimate tutoring support for investigations focuses on helping you produce better work, not on producing it for you. A tutor can help with refining an idea into a focused research question, understanding the underlying science, thinking through variables and experimental design, planning data analysis, interpreting results, and giving feedback on clarity and structure. The student remains responsible for producing and submitting their own assessed work. If you would like support, see our MYP Science tutoring in Hong Kong.

Practice Between Lessons With HK Study Hub AI

We're building an AI-powered practice platform to help students continue learning between tutoring sessions with targeted questions, feedback and personalized practice. Explore AI Tutor (currently under development).

MYP Science Investigation FAQs

How do I choose a topic? Start from a question you genuinely want to answer, check it has measurable variables and feasible data collection, and make sure the science is appropriate to your course. Avoid topics chosen only because they look impressive.

What makes a good research question? A single, focused, clearly stated question with an independent and dependent variable you can measure. "How does X affect Y?" is a workable frame.

Do I have to include uncertainty? Estimating and reporting uncertainty is part of good scientific practice and is expected where relevant. Treat it as part of the evidence, not an add-on.

How should I analyse my data? Move from data to relationship to scientific interpretation. Identify the pattern, connect it to the science, and explain what the graph means.

What goes in the evaluation? Specific limitations, the main sources of uncertainty, their impact on your results, realistic improvements, and an honest judgement of whether your evidence supports your conclusion.

Can a tutor review my investigation? A tutor can give feedback on clarity, structure, experimental reasoning and interpretation. They cannot write or rewrite assessed work for you.

Can a tutor write any part of it? No. The investigation must be your own work. Tutoring that crosses into writing the investigation for you is academic misconduct and risks your result.

For more, see our assessment criteria guide. Browse all MYP Science resources, MYP resources or MYP tutoring in Hong Kong.

MYP Science Investigation Guide