Guide to Laboratory Writing
Table of contents
4 Report structure and content
Appendix A: Decimal places, units and unit symbols
1 Introduction
Engineering and Product Design students need to acquire a number of key skills in order to qualify as effective professionals in their chosen fields. One of the most important is the ability to communicate technical information, both verbally and in writing, to a wide variety of people.
This document is designed to introduce you to some of the basic requirements of laboratory report writing. It should be used alongside the guidance you will receive as part of the year 1, term 1 'Technology for 21st Century Society' module. You should refer to it during your laboratory sessions and be advised that the lab supervisor or demonstrators may also refer you to it.
2 Originality
Usually you will have done the laboratory assignment with another person, so the results may not be entirely your own; that is always understood and accepted. But the report that is derived from those results must be your own work.
There are two ways of breaking this rule: collusion and plagiarism; both are equally unacceptable. Collusion means working with another person to write a report and then presenting it as your own unaided work. Plagiarism is making use of other people's ideas, facts or words, without acknowledgement, and presenting them as your own. Examination Boards take a serious view of these offences, and can impose penalties ranging from loss of marks to expulsion from the University.
The offence of collusion does not prevent you from discussing the report with another student; that would be unreasonable, and contrary to the idea of a university. It does mean, however, that the writing of the report must be an individual effort and not a joint effort. In particular you must make your own analysis and interpretation of the experimental results, and plot your own graphs.
You need to be clear about plagiarism and the restrictions on copying. It is never permissible to copy from another student; if acknowledged, the work will receive no credit; if not acknowledged it becomes plagiarism. However, you may use material, including facts and ideas, from publications such as textbooks provided you cite (refer to) your sources in the report. Section 4.11 describes the correct methods for citing sources. If you reproduce any material word for word you must enclose it in quotation marks as well as citing the source. Note, however, that it is seldom desirable to quote word for word in an engineering report and you should do so as little as possible. Formulae or equations (with no text) do not require quotation marks.
This guidance applies equally to information obtained from the Internet. It is very easy for markers to identify words and images that have been copied directly from web sites. If you do this without acknowledging the source of your information, by citing the URL of the source site and by putting the copied words in quotation marks then, again, you are likely to be found guilty of plagiarism.
Material that is common knowledge does not need to be cited in this way. This includes all the mathematics that you encounter in the course, and the basic engineering principles that are found in numerous textbooks. You must always cite the source of material that is not common knowledge - this is usually from specialist books and technical journals. In conclusion, if in doubt, cite the source.
3 Writing in the laboratory
Of course, a laboratory report can only be as good as the experimental work on which it is based. What is sometimes overlooked, however, is the quality of the record written in the laboratory. For some experiments you will be given specially prepared report forms that specify exactly what is required. For others you are required to record the experimental results and notes in a laboratory logbook, where you will need to make your own decisions about what to write. No matter what method of recording is specified, the following procedures are strongly recommended. They apply to all experimental work, both in the degree course and in professional work afterwards.
- Always record your results and observations in a secure form; this may be a special report form or a logbook. You must not record results on loose sheets of writing paper, even if they are filed later, because it is all too easy to lose these or to get them out of sequence.
- Never record results 'in rough' for subsequent copying, because copying wastes time and frequently introduces errors.
- Write in ink; this is your primary record of the experiment, so it must be permanent. Do not use correcting fluid. If you make a mistake, simply cross out the entry and write it again. Your laboratory record is the core of an 'audit trail' from the original readings to the final report.
- Record measurements and instrument readings in their original form. This means the actual numbers on a scale or display. If the numbers must be multiplied by a scale factor to get the final result, record the scale factor and give the calculated result as a separate step. This reduces the chance of error in recording the results, and enables the calculation to be checked.
- Include explanatory headings for the results so that they will be intelligible later. Do not rely on memory.
- Where graphs are to be plotted from experimental results, record the data in a table before plotting the graph. Plotting points on the graph directly from measurements without a table of values is likely to introduce errors, and it may degrade the accuracy of the data.
- Record any other information that may be required for a report. There is no need to copy any material from an instruction sheet, but you should record additional information such as the make and model or other identifier of the equipment used.
The aim of these procedures is to leave you with a secure, permanent and accurate record of the experiment from which you could write a report at any time, even when you have forgotten the details of what happened in the laboratory.
Calculations
Frequently you will need to make some calculations from your measurements. Be aware of the precision of your data, so that you do not quote the results of calculations to more significant figures than necessary. Most experimental data will not be known to more than three significant figures, even if a digital display shows more digits than this. Many values will only be known to two significant figures. It follows that three significant figures will be sufficient for the results of most calculations. Exceptionally, results may be given to four significant figures where the precision of the data justifies it, or for intermediate calculations where values are going to be subtracted. Subtraction magnifies errors, so it is important to preserve the full accuracy of the data.
Graphs
Whenever you are taking a range of measurements for different values of some variable, it makes sense to plot a graph in the laboratory. The standard laboratory logbook has graph paper on alternate pages for this purpose. These graphs are usually plotted quickly from the 'raw data' without any calculation. Their purpose is to reveal any suspect data points that do not lie on a smooth curve, and to see whether any additional points are required. If these graphs are to be of any use, they must be plotted as the experiment proceeds. They should also have titles and labelled axes - see section 6.3.
4 Report structure and content
There is an important distinction between formal reports and the short logbook reports required for many experiments. A formal report is complete in itself, so it must contain an account of the procedure and theory. A logbook report is to be read in conjunction with the laboratory instruction sheet, so it should not contain details of procedure or theory. It should have the following components:
- Results and graphs
- Discussion
- Conclusions
In addition, it should include notes of any difficulties encountered, or changes from the procedure given in the instruction sheet.
A formal laboratory report is a particular case of a technical report, which has the following basic structure:
- Summary
- Introduction
- Main text - several sections, with headings
- Conclusions
- References
- Appendices
The form and content of the main text will vary with different kinds of report. For a formal laboratory report, the conventional pattern for the main text is:
- Theory
- Equipment
- Procedure
- Results and graphs
- Discussion
The summary is a report in miniature, normally of not more than 200 words. It will state the main objectives of the work, and the principal results and conclusions; it will omit all inessential detail. A summary should not merely describe what the report is about; it should also give some information about the results. Where appropriate, it may include quantitative results. The summary must stand alone; it must be intelligible without reading any other part of the report except the title.
The purpose of the introduction is to lead the reader into the topic of the report; its nature will depend on the gap between the topic and the reader's background knowledge. For this reason the introduction on a laboratory instruction sheet will seldom make a good introduction for a formal report (quite apart from any question of plagiarism). Like the summary, the introduction is probably best written after the rest of the report.
A laboratory report normally includes a list of the objectives of the experiment; this will usually form part of the introduction.
If an experiment depends on a special piece of theory, then a theory section is required in the main text before the procedure section. If a piece of theory is used only in the analysis of the results, then the theory section can be placed after the procedure or results or even omitted completely.
When the amount of theory is too small to justify a theory section in the main text, you need only refer to a textbook source if the theory is well known, or include the theory in an appendix if it is not.
The report as a whole must contain enough information about the equipment for a reader to be able to repeat the experiment, but not all of it needs to be listed under equipment (in science reports this is usually termed apparatus). This section is not meant to be a catalogue of every item used in the experiment; you should include only those major items that it is useful to describe before the procedure section.
If you are investigating a standard manufactured product such as a motor, then you should give full particulars: all the information on the manufacturer's rating plate, for example. Also give its laboratory identification number. If the equipment has been made specifically for the experiment, then you will need to give its identification number, a sketch or a circuit diagram, and a brief description.
Usually it is not helpful to list standard measuring instruments such as voltmeters or pressure gauges under equipment; they are best identified in diagrams or descriptions in the procedure section. But you should include any measuring equipment that plays a major role in the experiment; this could be a special instrument, such as a spectrum analyser, or standard instruments where the measurement accuracy is crucial. In both cases give full particulars, including the serial numbers, so that the experiment can be repeated and the instrument calibration checked if necessary.
The procedure section of the report describes what was done, but it must not be reproduced from the laboratory instruction sheet; its purpose is quite different. Laboratory sheets are necessarily detailed, and the procedure is usually specified in a series of instructions. When the procedure is reported, it must be written in the past tense, and it must not be very detailed. For example, an instruction sheet may say:
"Set the load bank switches to 10 kW, and adjust the variac to give a primary voltage V1 of 230 V. Take readings of the primary current I1, the primary power P1, the secondary voltage V2 and the secondary current I2. Repeat the experiment with the load bank switches set successively to 8, 6, 4 and 2 kW; adjust the variac to give a primary voltage of 230 V in each case."
This procedure could be reported as follows:
"Primary and secondary quantities were measured for loads in the range 2 - 10 kW, with the primary voltage held constant at 230 V."
In this example there is no need to specify which quantities were measured, or the individual load values, because all of this information should be contained in the results.
Where necessary, give circuit diagrams or sketches in addition to the written description of the procedure. Indicate the range of any measuring instrument, for example '20 A' beside an ammeter symbol in a circuit diagram. If important instruments have been listed in the equipment, these must be identified in the procedure so that the reader will know which measurements were made with them.
There is one important respect in which the procedure section of the report can go further than the laboratory instruction sheet. It should record any difficulties that were encountered, any steps taken to overcome them, and any special precautions that were taken.
The results section of the main text must contain the results in a form that will be most useful to the reader, which usually means the final results as graphs or tables. The original raw data and the tables from which any graphs were plotted are essential components of a logbook report. However, for a formal report, these should not go in the results section. Instead, all material of this kind should go into an appendix, where it is available for reference but will not interrupt the flow of the main text.
Where the final results are derived from the original measurements by calculations, the formula should either be stated in the results section or referenced by an equation number in the theory section. It is not necessary to show any details of calculation in the results section, but an example calculation should be included in an appendix with the raw data.
When experimental results are compared with theoretical results, it is important to identify their origins. Tables of results should show clearly the values that are theoretical and those that are derived from measurement. Graphs should show theoretical results as smooth curves, with no points visible; experimental points must be clearly marked with distinctive symbols. See section 6.3 for further information about graphs.
The discussion must be related to the objectives of the experiment. It should be short and to the point, but it must be the outcome of careful thought and analysis. If one of the objectives is to compare theoretical and experimental results, then the discussion must include an analysis of errors; a statement like "results agreed within the limits of experimental error" is meaningless without this analysis. See section 8 for the treatment of errors. Often there will be a discrepancy that cannot be explained by experimental error; the discussion should then consider possible causes.
There are two pitfalls to avoid when comparing theory with experiment. One is to force the experimental results to fit the theory, for example by drawing a straight line through a set of points that actually lie on a curve. Textbook theories are often over-simplified, and the experimental results may be nearer the truth. The second pitfall is to advance impossible explanations for the difference between theory and experiment. If you think that a particular effect may be responsible for the difference, try to quantify the effect and see whether it is at least possible.
The conclusions section should be a short summing-up of the outcome of the experiment, drawing together the main findings from the results and the discussion. It should show how far the object of the experiment has been achieved. The conclusions should not introduce any new material; in particular, this section should not include any discussion of the results.
It is important to appreciate the difference between discussion and conclusions. These should always be distinct sections in the report, and never lumped together to form a single 'discussion and conclusions' section.
Section 2 stressed the importance of citing the source of material that is not common knowledge.
Whenever you refer to published sources of this kind, for example when quoting technical specifications or specialist theory, you must include full particulars of the source in a numbered list of references. As an example, section 9 shows the correct format for referencing books. Internet sources should be referenced by quoting the site url. Note that every item in the reference list must be referred to by inserting its number in the appropriate section of text. This is done using either a superscript1 or square brackets [1], as in the examples in section 5.
Appendices are used for any material that must be included in the report for the sake of completeness, but would disturb the flow of the report if placed in the main text. Examples are the laboratory instruction sheet, the original experimental results, sample calculations, and pieces of theory that do not warrant a theory section in the main text. It must be possible to understand the main part of the report without reading the appendices.
The report must begin with a title page, set out as specified below.
- Author's name and laboratory group in the top left of the page
- Date of experiment in the top right of the page
- Experiment title in the centre of the page
- Summary in the bottom half of the page
If preferred, the summary may be placed on a separate page immediately after the title page.
A list of contents should follow the title page (or summary page if this is used), listing all the sections of the report with their page numbers. Do not include the title page, summary page or contents page in the list of contents, and do not number these pages. All the pages that follow the contents page should be numbered consecutively starting at 1.
If required, a list of symbols can be placed after the contents page. Include this list if the report uses many symbols, or if the symbols are different from those normally used.
Print the report on A4 paper, using one side of the paper only. Leave ample margins for the marker's comments. Binding is at your discretion; a secure staple in the top left-hand corner is the minimum requirement. Note that the report, including graphs, must be easy to read without removing the binding.
5 Writing style
The style of writing which you use to convey technical information is, in general, different from that used for other purposes. There are a number of excellent textbooks on the subject such as references [1], [2] and [3], which cover this topic. Make sure that you refer to this guidance whenever you write laboratory reports. All of your report markers will expect you to be familiar with it and use the correct style.
6 Tables diagrams and graphs
Consider whether a table is the best way of presenting the data. Alternatives are graphs, bar charts or pie charts. If you choose a table, observe the convention that related items are grouped in adjacent rows, with the details listed in columns which progress logically from left to right.
Each table should have a table number and a caption and should be referred to in the text. Choose the table ruling to display the data clearly; see Turk and Kirkman [3] for an excellent treatment of table layout.
Always use the simplest diagrams that will serve the purpose. Often this will mean drawing them specifically for the report, rather than reproducing existing diagrams. Each diagram should have a figure number and a caption and should be referred to in the text.
Small diagrams should be placed in the text, as close as possible after the text reference. Do not put them before the text reference; unexplained diagrams distract the reader. Full-page diagrams and graphs may be placed before or after the text reference, whichever is more convenient for the reader. It can be helpful to group full-page diagrams together at the end of the report, particularly when they are referred to in several places.
The first consideration is whether the axis scales should start at 0. False origins, where the scales do not start at 0, should be used with caution; they can be very misleading. As a rule, only use a false origin if the values occupy less than 50% of the range starting at 0. In particular, do not use a false origin simply to fit the graph onto a large area of graph paper. These considerations do not apply to logarithmic scales, which must have false origins because log(0) = -∞.
Choose scales that have a simple relationship to the main 5 mm and 10 mm rulings on the graph paper, if necessary at the expense of graph area. If you use a scale such as 35 mm to one unit, it is difficult to read intermediate values from the graph.
Choose the orientation of the graph paper - portrait or landscape format - to make the best use of the available area. Leave enough room for margins and captions.
Labelling of axes
Each axis must be labelled with the quantity and the units (if any). Scale factors should be applied to units, not to numbers or quantities, using the preferred SI prefixes for multiples and sub-multiples (see Appendix A). For example, an axis showing lengths ranging from 0 to 120 mm should be labelled:
length, mm
and not length, m × 103 or length × 103 m.
When non-dimensional quantities are plotted, it is ambiguous to use a scale factor in the axis label. For example, if an axis is labelled p × 103, it is not clear whether p has already been multiplied by 103, or whether a number read from the graph must be multiplied by 10-3. With quantities of this kind, put the scale factor at the end of the axis.
Layout and presentation
Where you have a choice, draw the graph so that it can be read with the page in its normal position. This will always be possible with portrait-style graphs where the y-axis is longer than the x-axis.
The correct way of labelling the y axis (following the technical drawing convention that the text is viewed from the right-hand side) is to have the text running vertically from the bottom of the axis to the top. The axis labels should be positioned centrally, with the text running parallel to the axis.
Mark experimental points clearly with distinctive symbols such as '×' or '+'. For curves drawn through theoretical points, do not show the points; the curve alone is sufficient. Where several curves are drawn on the same axes, they must be clearly identified on the graph. Either label the curves, or use different symbols for the points and a key to the symbols.
Every graph must have a figure number and a title. If possible the title should be placed underneath the graph as with other diagrams, but computer packages for plotting graphs usually put the title at the top.
Make the title informative. Use words and not symbols, and do not reproduce the axis labels. For example, the title "Graph of force against distance" is worthless if the axes are labelled "force, N" and "distance, m". An informative title would be "Graph of lift force against distance from magnet poles". Units are not required in a graph title; the axis labels give this information.
Wherever possible, draw curves smoothly. If you are drawing by hand, use a Flexicurve or similar aid. If you are using a computer, most application programs give the option of drawing smooth curves instead of joining the points with straight lines. A particular problem with computer-drawn graphs is in fitting a smooth curve to experimental data, where the curve does not necessarily pass through all the points. If the equation of the curve is known, then a least-squares curve fit is possible with packages such as MATLAB.
7 Numbers and units
Beware of the calculator accuracy syndrome: writing numbers to the full display accuracy of the calculator. Use no more significant figures than the data warrant. In most laboratory work three significant figures will be sufficient - see section 3.2.
When giving numerical values of quantities in SI units, keep the number in the range 1 to 1000 by selecting the correct decimal prefix. For example, a length of 0.102 m should be written as 102 mm. Avoid writing these numbers in scientific notation; do not use 1.02 × 10-1 m for 102 mm. These rules do not apply if the units are not SI, or the number is dimensionless. When writing such numbers in scientific notation, use '×', not a dot, for the multiplying symbol.
Numbers with more than four digits before or after the decimal point should be grouped in threes, starting at the decimal point. Use a space (strictly, a half space) to separate the groups, not a comma, because the comma is used as a decimal point in some countries. Examples: 3.141 592 65; 40 000 ft; £5600 (four digits - no separator)
As far as possible, use SI units in the report. The names of all SI units begin with a lower-case letter, even when a unit is derived from a person's name, for example the newton. If a plural is required, it is formed by adding an 's'; thus the correct plural of henry is henrys, not henries.
Approved abbreviations for SI units are known as unit symbols. They begin with a capital letter when the unit is derived from a person's name, but they never end with a full stop. Unit symbols never take a plural form. Avoid non-standard abbreviations for units; for example, s is the unit symbol for second; sec is incorrect. There is a particular problem with this unit symbol, however, because s is the symbol for the Laplace transform variable (which has units of 1/s!). To avoid possible confusion, use the abbreviation sec in this context.
In a word-processed report, use normal upright type for units and unit symbols. By convention, italic (sloping) type is used for algebraic symbols, which helps to avoid confusion between quantities and units.
Decimal prefixes are always written next to the unit symbol, without a space or a full stop, for example kW. In compound units, use a slash (/) rather than a negative power to denote division; write m/s, not ms-1. Multiplication needs a little care, particularly when m is one of the unit symbols. Thus Nm is a newton-metre, but mN is a millinewton. If a metre-newton is intended, it should be written m N or m.N. Appendix A lists the common units, unit symbols and decimal prefixes.
8 Experimental errors
There are three main kinds of error in experimental work: errors of observation, systematic errors, and instrument calibration errors. Errors of observation are essentially random variations that affect most physical measurements. They can be treated by statistical methods [4], and they are easily identified by repeating the same measurement several times. In principle they can be made small by repeating the measurement many times, but there will be a limiting value set by the instrument scale or digital display. These are often the least significant errors in an experiment.
Systematic errors represent defects in the measuring equipment or the experimental method that cause the measured value to differ from the true value. By definition they cannot be reduced by repeating the measurement, and they can be very difficult to eliminate.
Instrument calibration errors are systematic errors of a particular kind. They represent imperfections in the measuring instrument as a difference between the true value and the indicated value; they have nothing to do with the way the instrument is used. For example, any voltmeter draws a current that will affect the circuit under test. This can introduce a systematic error, because the voltage at the meter terminals will not be the same as the original circuit voltage. The voltmeter calibration error is additional to this; it is the difference between the actual terminal voltage and the value indicated by the meter.
Instrument calibration errors are often the dominant errors in an experiment. For analogue instruments, these errors are expressed as a fraction of the full-scale reading (FSR) of the instrument, and they can introduce large fractional errors when the reading is low. For example, if a voltmeter has a full-scale reading of 300 V and the accuracy is specified as 1% of FSR, then the reading can be in error by +/- 3 V at any point on the scale. If a particular reading is 30 V, then the possible error is +/- 10% of the reading, quite apart from any errors of observation.
With digital instruments, the calibration errors are usually expressed as a fraction of the actual reading together with a number of digits, for example +/- 0.5% of the reading +/- 2 digits.
The error in a single measurement will be a combination of the error of observation and the instrument calibration error. There is no way of knowing whether they have the same sign or opposite signs, so the sum of the two errors must be taken as the possible error in the measurement.
With analogue instruments, errors of observation can be estimated from the instrument scale markings. It is usually safe to take the error to be half of the smallest interval between scale marks; the error is not likely to be greater, and can be considerably smaller. With a digital instrument, take the error to be +/- 1 in the last displayed digit.
Instrument calibration accuracy is often marked on the instrument or stated in the instruction book. This should always be treated as an optimistic estimate unless the instrument has been calibrated recently by a standards laboratory. Few analogue instruments will be better than 1% of FSR, and many will be worse than this. In the absence of other information, assume a calibration error of 2% of FSR for analogue instruments and 0.5% of the reading for digital instruments.
Often a quantity is derived from several different measurements. It is necessary to calculate the possible error in the derived quantity, given the errors in the individual measurements. Topping [4] describes how this is done and derives approximate expressions for the errors in combinations of quantities.
9 References
[1] - Davies, J W: Communication Skills - A Guide for Engineering and Applied Science Students (2nd ed., Prentice Hall, 2001)
[2] – Houp, K.W., Pearsall, T.E., Tebeaux, E. and Dragga, S. Reporting Technical Information (OUP, 2006)
[3] - Turk, C and Kirkman, J: Effective Writing (2nd ed., Spon, 1988)
[4] - Topping, J: Errors of Observation and their Treatment (4th ed., Chapman & Hall, 1972)
Appendices
Appendix A: Decimal places, units and unit symbols
Decimal prefixes:
| Prefix | Symbol | Multiplying factor | |
| tera | T | 1012 | |
| giga | G | 109 | |
| mega | M | 106 | |
| kilo | k | 103 | |
| hecto | h | 102 | Avoid using |
| deca | da | 101 | Avoid using |
| deci | d | 10-1 | Avoid using |
| centi | c | 10-2 | |
| milli | m | 10-3 | |
| micro | μ | 10-6 | |
| nano | n | 10-9 | |
| pico | p | 10-12 | |
| femto | f | 10-15 | |
| atto | a | 10-18 |
SI units:
| Quantity | Unit | Symbol | Comment |
| length | meter | m | |
| volume | litre | l | 10-3 m3 |
| mass | kilogram | kg | |
| mass | tonne | t | 1000 kg |
| time | seconds | s | |
| absolute temperature | kelvin | K | |
| luminous intensity | candela | ca | |
| luminous flux | lumen | lm | |
| illumination | lux | lx | cd sr |
| plane angle | radian | rad | |
| solid angle | steradian | sr | |
| force | Newton | N | |
| pressure | pascal | Pa | N/m2 |
| work, energy | joule | J | N m |
| power | watt | W | J/s |
| frequency | hertz | Hz | |
| electrical current | ampere | A | |
| electrical charge | coulomb | C | A s |
| electrical potential | volt | V | J/C |
| electrical capacitance | farad | F | |
| electrical resistance | ohm | Ω | |
| electrical conductance | siemens | S | 1/Ω |
| magentic flux | weber | Wb | |
| magentic flux density | tesla | T | W/m2 |
| inductance | henry | H |
Other units:
img/
| Quantity | Unit | Symbol | Comment |
| pressure | bar | bar | 105 N/m2 |
| volume | gallon | gal | 4.546 l |
| temperature | degree celsius | oC | interval 1K |
| temperature | degree farenheit | oF | interval 5/9K |
| length | inch | in | 25.4mm |
| length | foot | ft | 304.8mm |
| length | mile | mile | 1.609km |
| mass | pound | lb | 0.4536kg |
| force | kilogram force | kgf | 9.807N |
| force | pound force | lbf | 4.448N |
| power | horsepower | hp | 746W |
| rotational speed | revolution per second | rev/s | 2π rad/s |
| rotational speed | revolution per minute | rev/min | 2π/60 rad/s |
Updated and revised by the Department of Engineering & Design, November 2022