Writing about a MATLAB experiment can be harder than running one. You may have working code, a graph that looks right, and a set of results, but that does not automatically make a strong academic assignment.

The important part is explaining what you actually did and why you did it. Your reader should be able to understand the purpose of the experiment, the method you followed, the results MATLAB produced, and what those results tell you.

I find it useful to think of MATLAB as the tool used to investigate a question. The assignment should focus on the investigation itself, rather than becoming a description of every command in the program.

What Should You Include in a MATLAB Experiment?

Before writing, ask yourself four simple questions:

1. What was I trying to find out?

2. How did I carry out the experiment in MATLAB?

3. What results did I obtain?

4. What do those results mean?

These questions give you a straightforward structure for the assignment.

A typical experimental report may use sections such as Introduction, Methodology, Results, Discussion and Conclusion. The exact structure depends on your course and your lecturer's instructions, so those requirements should always take priority.

The main thing is to keep the purpose of each section clear.

For example, instead of writing:

MATLAB was used to calculate the values and produce a graph.

you could write:

MATLAB was used to investigate how the calculated option price changed as volatility increased, while the remaining model parameters were kept constant.

The second sentence tells the reader what the experiment was actually investigating.

Start With a Clear Aim

Your introduction does not need to be complicated. It should give the reader enough background to understand the problem and explain what the experiment was designed to examine.

A useful aim normally identifies the main variable being investigated and the outcome being measured.

For example:

The aim of the experiment was to investigate the effect of volatility on the calculated value of a European call option. MATLAB was used to evaluate the pricing model for a range of volatility values while the other parameters were held constant.

That is much more useful than simply saying that MATLAB was used to calculate an option price.

If you are working on an engineering, mathematical or scientific problem, you can also briefly introduce the theory behind the experiment. You do not need to reproduce an entire textbook explanation. Include only the theory needed to understand what you tested.

Explain Your MATLAB Method

The methodology section is where you explain how you carried out the experiment.

A reader should have enough information to understand your approach without having to study your entire MATLAB script.

Describe the software and approach

If relevant, mention the MATLAB version and any specialist toolbox you used.

For example:

The experiment was implemented using MATLAB R2026b. A MATLAB script was developed to calculate the required output for each set of input parameters.

You do not always need to state the software version. It becomes more useful when the version, toolbox or computational environment could affect reproducibility.

State your input parameters

This is one area where students often provide too little information.

Don't simply say that you tested “different values”. Tell the reader what those values were.

For example:

The underlying asset price was set to £100, the strike price was £100, the risk-free interest rate was 5%, and the time to maturity was one year. Volatility was varied from 10% to 50% in increments of 5%.

Now the experimental conditions are clear.

Explain what MATLAB did

You do not need to describe every line of code.

Instead, explain the overall process:

For each volatility value, the MATLAB program evaluated the pricing equation and stored the resulting option price. The calculated values were then plotted against volatility to examine the relationship between the two variables.

That gives the reader a clear picture of the procedure without turning your methodology into a programming tutorial.

Mention important assumptions

Computational experiments are based on assumptions, just like physical experiments.

For example:

During the sensitivity analysis, all parameters except volatility were held constant. This allowed the effect of volatility on the calculated option price to be examined independently.

This kind of statement shows that you understand the design of your experiment rather than simply entering numbers into MATLAB.

Don't Use Your Code as a Substitute for Explanation

Including MATLAB code can be useful, particularly when the assignment specifically asks for it. However, code should support your explanation rather than replace it.

Suppose your program contains a loop that tests several values of volatility. You could explain it like this:

Volatility was varied between 10% and 50% in 5-percentage-point increments. For each value, the pricing model was evaluated while the other parameters remained unchanged. The resulting prices were stored and used to generate the sensitivity plot.

The reader now understands the experimental design without needing to work through the code themselves.

You can then provide the relevant code underneath if required.

This approach is particularly important in academic work because your lecturer is generally interested in whether you understand the method, not simply whether you can produce a working MATLAB script.

How to Describe MATLAB Graphs

A graph should never be left to speak entirely for itself.

If you include a MATLAB figure, give it a meaningful caption and make sure the axes, units and legend are clear where necessary.

A weak caption would be:

Figure 1: MATLAB graph.

A better caption would be:

Figure 1. Effect of volatility on the calculated European call option price while the underlying asset price, strike price, interest rate and maturity remain constant.

The surrounding text should then explain the important observation.

For example:

Figure 1 shows a positive relationship between volatility and the calculated call option price. As volatility increased across the tested range, the calculated option value also increased.

This is much more informative than:

The graph shows the results.

MATLAB's Live Editor can be useful here because it allows code, output and explanatory text to be kept together in a single live script. MathWorks also provides tools for exporting live scripts into formats commonly used for reports.

Separate Results From Discussion

A common problem in student assignments is mixing the results with their interpretation.

These are related, but they are not exactly the same thing.

Results: What did you find?

The Results section should report the evidence produced by the experiment.

For example:

Increasing volatility from 10% to 50% produced a corresponding increase in the calculated call option price. The calculated price was approximately £8 at 10% volatility and increased to approximately £21 at 50% volatility.

Use the actual values from your own MATLAB experiment rather than invented or rounded figures.

Discussion: What does it mean?

The Discussion section is where you explain the significance of the result.

For example:

The increase in option value is consistent with the theoretical relationship between volatility and a European call option. Higher volatility increases the range of possible future underlying-asset prices, which can increase the value of the option's asymmetric payoff.

The Results section tells the reader what happened. The Discussion explains why it matters.

That distinction makes the assignment much easier to follow.

Use Numbers Instead of Vague Statements

Academic writing becomes stronger when you support observations with actual evidence.

Try to avoid phrases such as:

• “The result increased significantly.”

• “There was a large difference.”

• “The simulation worked well.”

• “The results were close to the expected answer.”

Unless you define what “significant”, “large” or “close” means, these statements do not tell the reader much.

Instead, give the relevant measurement.

For example:

Reducing the numerical time step from 0.10 to 0.025 reduced the absolute error from 0.084 to 0.031.

You can then explain the importance of that change:

This represents a reduction in absolute error of approximately 63%, indicating that the numerical solution became more accurate as the time step was reduced.

Specific evidence makes your discussion much more convincing.

Explain Unexpected Results

Not every MATLAB experiment will produce exactly the result you expected.

That is not necessarily a problem.

If your result differs from the theoretical value, investigate the possible reasons instead of hiding the discrepancy.

Depending on the experiment, the difference could result from:

• numerical approximation;

• rounding;

• an unsuitable step size;

• assumptions in the model;

• implementation errors;

• limited input data;

• random-number generation;

• differences between theoretical and practical conditions.

For example:

The numerical result remained slightly different from the analytical solution, although the difference decreased as the time step became smaller. This suggests that numerical approximation contributed to the observed error.

That gives the reader a reason for the discrepancy and shows that you have thought about the quality of the result.

If your experiment uses random numbers, reproducibility becomes particularly important. MATLAB provides functionality for controlling and restoring random-number-generator states, which can help when you need to reproduce computational results.

Make Your Experiment Reproducible

A good computational experiment should be reasonably easy for another person to understand and, where appropriate, reproduce.

You should normally provide the important details, including:

• the input data;

• parameter values;

• parameter ranges;

• equations or algorithms;

• assumptions;

• number of simulations or trials;

• relevant MATLAB functions or toolboxes;

• important computational settings.

You don't necessarily have to include every technical detail in the main body. Some code, additional tables or supplementary results can go into an appendix if your assignment format allows it.

For larger experiments, MATLAB's Experiment Manager can also help organise parameter combinations and compare results systematically.

A Simple Structure for Your Assignment

If you are not sure how to organise the report, the following structure is a good starting point.

Introduction

Explain the problem, relevant background theory and the purpose of the experiment.

Methodology

Describe:

• the MATLAB environment;

• input data;

• parameters;

• equations or models;

• experimental procedure;

• assumptions;

• parameter ranges.

Results

Present:

• important numerical results;

• tables;

• graphs;

• comparisons;

• error measurements.

Keep this section focused on the evidence produced by the experiment.

Discussion

Explain:

• the main trends;

• whether the results agreed with theory;

• unexpected findings;

• possible sources of error;

• limitations;

• what the findings mean.

Conclusion

Return to the original objective and explain what the experiment showed.

Avoid introducing a completely new argument in the conclusion. Keep it focused on the findings you have already discussed.

Example of a Weak and Strong Description

Consider this example:

MATLAB was used to test different interest rates. The graph shows that the option price changed. The results were good.

There is not enough information here for an academic report.

A stronger version would be:

MATLAB was used to investigate the sensitivity of the calculated option price to changes in the risk-free interest rate. The interest rate was varied from 2% to 8%, while the underlying asset price, strike price, volatility and time to maturity were held constant. For each interest-rate value, the pricing model was evaluated and the resulting option price was recorded. The results showed a gradual change in the calculated call option price as the interest rate increased. This behaviour was consistent with the theoretical relationship predicted by the model for the tested parameter range.

The second version tells the reader what was changed, what was controlled, what MATLAB calculated and what the experiment demonstrated.

That is the level of detail you should aim for.

If your assignment involves financial modelling, for example, you may also need to distinguish between implementing a mathematical model and discussing its practical assumptions. Resources such as derivatives pricing options help in UK can be useful as supplementary material when you need additional context around MATLAB-based derivatives-pricing work, but your academic argument should still be based on your own calculations and authoritative sources.

Common Mistakes to Avoid

Explaining every line of code

Your assignment usually does not need a sentence explaining every for loop, variable assignment or plotting command.

Explain the purpose of the code instead.

Including screenshots without discussion

A screenshot of a MATLAB command window is not, by itself, evidence of good analysis.

If you include an output screenshot, explain what the important output means.

Using graphs without captions

A reader should be able to understand what a figure represents without guessing what the axes or lines mean.

Calling a result “successful”

Say what the result actually demonstrated.

For example, instead of:

The experiment was successful.

write:

The numerical solution converged toward the analytical solution as the step size was reduced.

Ignoring limitations

Every model has limitations. Acknowledging them does not weaken your assignment. In many cases, it shows that you understand the difference between a simplified computational model and a real-world system.

Describing MATLAB instead of the experiment

This is perhaps the biggest mistake.

Your assignment is not really about the plot() function or the MATLAB interface. It is about the problem you investigated using MATLAB.

How to Make the Writing Sound More Academic Without Making It Stiff

You do not need complicated vocabulary to make your writing sound academic.

Clear sentences are usually better.

Instead of:

The aforementioned computational implementation was subsequently utilised for the purpose of obtaining the aforementioned numerical outcomes...

write:

The MATLAB model was then used to calculate the numerical results.

The second sentence is easier to read and communicates exactly the same idea.

I would also avoid repeating phrases such as “it can be seen that” throughout the report. Replace them with direct statements:

The results show that...

The calculated error decreased as...

Increasing the parameter resulted in...

The simulation produced...

This keeps the writing natural and makes your conclusions more direct.

A Final Checklist

Before submitting your assignment, read through it and ask:

• Have I clearly explained the aim?

• Have I identified the variables I changed and measured?

• Have I provided the important input values?

• Have I explained the computational method?

• Have I stated important assumptions?

• Have I used actual numerical evidence?

• Are my figures labelled and captioned?

• Have I explained what the figures show?

• Have I separated results from interpretation?

• Have I compared the findings with relevant theory?

• Have I discussed errors and limitations?

• Could another reader understand how the experiment was performed?

• Does my conclusion answer the original objective?

• Have I referenced external information correctly?

The main principle is straightforward: describe the experiment, not just the MATLAB code.

When you connect the research question, experimental method, MATLAB output and interpretation, your assignment becomes much stronger. The reader can see not only that you obtained a result, but that you understand how the result was produced and what it actually means.