Back to Home Page
You will be writing reports constantly for the rest of your professional career, you might as well write good ones!
This document does not seek to replace the standard report-writing documentation that you received in first year, it just seems that I am constantly harping on a few points that may need clarification :-)
Writing a report takes planning. Many people do not plan a report-you have spent weeks designing a widget, now design the report-don't ``just write'' it! Set up your section headings, your subsections and subsubsections, and write a 2-3 line description of what you are going to say in each of these divisions. Stew on that design, and optimise the flow of thought. Ask yourself whether it makes sense to introduce the concepts at that point in the report etc. Don't forget to introduce the background to the subject gently-don't barge in with fancy maths!
By the way, the above paragraph is why I say that a hand-written report must be written at least four times!! It is simply not possible to sit down and write the entire report in the correct logical order without messing it up!
The next common mistake is to address the report to an idiot. ie ``My class has been given a project at the Department of Electrical Engineering at the University of the Witwatersrand, Johannesburg, Gauteng, South Africa, and we need to construct a student death warning device.'' I couldn't care HOW you got the project, what I want is a succint and informative PROBLEM STATEMENT. This does NOT mean copying the intentionally vague problem statement that was handed out to you!
You need to define your target audience as a relatively intelligent colleague, but perhaps not quite in your field, and you need to lead her into the topic.
The next common mistake is to bumble along. A report is essentially an argument. It has a premise (the problem statement) and a conclusion. You need to argue in a direct line without meandering about the thousands of design options you considered and rejected. Your report is about your final design, and if you need to refer to earlier designs as they influenced the final design, fine, but keep it to the point.
Quantify. Quantify. Quantify. Always quantify. Quantify the error in your reference voltage as a function of the supply voltage, and hence get it down to how much of a temperature difference its going to make at the end of the chain. ie Find and Define your errors. We do NOT live in an ideal world.
Explain the design at a high level first, ie break it up into logical functional blocks, then explain why a resistor is there, and how you got to its value.
When we say that the report must follow a format of Abstract, Introduction, Body, Conclusion, References; for heaven's sake don't call the ``Body'' section BODY. By the ``Body'', we mean the bulk of the report. Split that up into several sections eg Temperature Sensing; Setting the Reference Voltage; Comparing the Reference to the Temperature; Output Interface...
Any material which detracts from the line of argument (like long maths) needs to be placed in an appendix... BUT!!! the ``bottom line'' needs to be referred to in the text, else its a pretty appendix. Note that in general, appendices are NOT read.
Reference any statement, or derive it from first principles, unless it is basic knowledge. Use a PROPER referencing system-either a numerical system similar to the IEEE system, or a name-based Harvard system. Make sure that you use them properly. Pay particular attention to verbal communication and Web-based refs. Remember that the purpose of a reference is to enable the reader to get more information on the topic.
A conclusion is not:``I really learn't a lot in this project, and the tutor was very nice, and the department is very nice''.
Point Form is often useful in a conclusion:
The online version is "http://ytdp.ee.wits.ac.za/ReportWriting.html"
Back to Home Page