Paper structuresUniversity & graduate
Lab Report
The write-up of an experiment you ran yourself: what you expected, what you did, what happened, and what it means. It's the training version of the scientific journal article, built on the same IMRaD shape.
What it is
A report of an experiment you carried out, usually for a science or engineering course in high school or university. You state a question and a hypothesis, describe exactly what you did, present the data, and say whether the data support the hypothesis.
It sits in the middle of the ladder for a reason. The papers below it gather and explain what other people found. A lab report is the first paper built on evidence you produced yourself, so it has to show how you got that evidence and how far it can be trusted.
Most lab reports follow the same shape as a scientific journal article: Introduction, Methods, Results and Discussion, known as IMRaD. Courses often add a title page and an abstract at the front and references and appendices at the back. Writing lab reports is practice for writing research papers: the sections, the tense and the habits of honest reporting carry straight over.
Who reads it
Your instructor or teaching assistant, who usually knows what the experiment should show and is checking whether you understood it. It helps to write for a second reader too: someone who knows the science but wasn’t in the lab. That reader should be able to follow what you did, judge your results, and repeat the experiment from your Methods alone.
The parts
1Title and abstract
The title says what was tested or found, in a phrase rather than a sentence. Course guides often suggest ten words or so. Many courses also want a title page with your name, your lab partners, the section and the date.
The abstract is one paragraph that sums up the whole report: the purpose, the method in brief, the key results with numbers, and the main conclusion. Write it last. Length limits vary; one university guide asks for 100 to 200 words, an MIT course caps it at 250.
2Introduction and hypothesis
Say what the experiment was for, give the background the reader needs (the theory, and earlier findings where relevant), and end with the hypothesis. A good hypothesis names the independent variable (what you changed) and the dependent variable (what you measured), and predicts how they relate. It should be written down before you run the experiment, not fitted to the results afterward.
3Materials and methods
What you used and what you did, in enough detail that someone else could repeat it. Give quantities and settings, say which groups were controls and what they controlled for, and explain why a step was done if it isn’t obvious. Write it as an account in the past tense (“the solution was heated to 60 °C”), not as instructions copied from the lab manual (“heat the solution to 60 °C”). If you changed the procedure, say so: three repetitions instead of four, a substitute reagent, a step you skipped.
4Results
What you found, without interpreting it. Short text points the reader to tables and figures and names the main trends. Tables suit exact values; graphs suit trends. Number and title each one, label axes and columns with units, and don’t repeat the same data in a table and a graph. Report every result that bears on the hypothesis, including the ones that look wrong.
If the raw data are long, put them in an appendix and give a summary here.
5Discussion
What the results mean. Say whether the data support the hypothesis, and use “supported” or “suggested”, not “proved”. Explain results that differ from what you expected. Account for error with specific causes (an instrument’s limits, a contaminated sample, a calibration problem) rather than a vague “human error”. Compare your results with published values or earlier work where you can, and say what you would change if you ran it again.
6References and appendices
References list the sources you cited, in the style the course asks for. Appendices hold what supports the report but would clutter it: raw data, a sample of each type of calculation, extra graphs. Give each kind of material its own labeled appendix, and refer to each one somewhere in the report.
What this level expects
A clear, testable hypothesis, stated before the results, with the variables named.
Methods someone could repeat. Quantities, conditions, controls and any changes to the procedure.
What actually happened. Report the run that went wrong, the outlier and the step you changed. Hiding them is worse than explaining them, and the explanation is often where the marks are.
Numbers with their uncertainty. A measurement is reported with its uncertainty and units, and the value is given only to the precision the uncertainty allows: 3.45 ± 0.05, not 3.452 ± 0.05.
Results kept apart from interpretation. Data in Results, meaning in Discussion.
Past tense for what you did, present tense for things that are still true: the theory, the report itself, what a figure shows.
Common mistakes
Copying the lab manual. Methods written as numbered instructions, or pasted from the handout, show the instructor nothing about what you actually did.
“Human error.” It tells the reader nothing and suggests carelessness. Name the source of error and how it could have shifted the result.
“The results prove the hypothesis.” One experiment supports or fails to support a hypothesis. It doesn’t prove it.
Leaving out data that don’t fit. If a result looks wrong, report it and discuss it. In published research, omitting data so the work is misrepresented counts as falsification, a form of research misconduct.
Too many or too few digits. A value given to more decimal places than the uncertainty justifies claims precision you don’t have.
Unlabeled graphs. No title, no units, no axis labels: the reader can’t tell what was measured.
What varies
Almost every course sets its own format, so read the lab manual and the marking guide first. Some want an abstract and some don’t; some fold the hypothesis into the Introduction and some give it a heading; some call the Discussion “Conclusions”. Longer engineering reports may add sections for theory and calculations.
Voice is the most argued-over convention. The traditional rule is the passive voice (“the samples were weighed”), to keep attention on the work rather than the worker, and many instructors still ask for it and for no “I” or “we”. But the active voice (“we weighed the samples”) is increasingly accepted, and an editor at the journal Nature wrote in 2007 that it had long been the journal’s policy. If the assignment doesn’t say, ask, then use one choice consistently.
The word “error” needs care. In a lab report it usually means uncertainty in a measurement, random (scatter that careful work can’t remove) or systematic (a bias in one direction, such as a badly calibrated instrument), not a blunder. How much formal error analysis a course expects varies a lot.
Sources & further reading
- Scientific Reports — The Writing Center, University of North Carolina at Chapel Hill
- The Lab Report — Writing Advice, University of Toronto
- Tutoring Lab Reports — Purdue Online Writing Lab
- Guidelines for Writing Up Your Research, 20.109 Laboratory Fundamentals in Biological Engineering — MIT OpenCourseWare
- How to Write a Lab Report — UNSW Sydney
- Appendix A: Uncertainty and error analysis — introductory physics lab manual, hosted by WebAssign
- Definition of Research Misconduct — US Office of Research Integrity
- Maxine Clarke, “Active is better than passive or neutral,” Nature blogs, 2007
End of page · Back to top