Monthly Archives: August 2010

How to Make your Project Reports Better….

As time goes by I get less and less patient with bad writing and sloppy presentation…  here are some thoughts about how to make that technical report a bit better:-

  1. Remember that you are telling a story.  It has to be backed up with evidence but there should be a beginning, a middle and an end.  At all times you should have a narrative.
  2. Your report is not a murder mystery.  Please don’t leave the reader waiting until thirty or even one hundred pages [1] in before you tell them what you are doing, you should explain your key results in the summary and abstract.  The introduction should say what you actually did.
  3. Your report sections are not a set of in-laws who don’t talk to each other.  If you are talking about your experimental set up mentioning how many tests you ran or that the results will be along soon in section five is something to be encouraged.  Remember point 1 – you need a narrative from beginning to end.
  4. Your literature review should do three things: 1) Explain why your topic is important 2) Explain what has been done before about it and 3) Describe the “gap in the market” that you work is going to address.
  5. The point of the report is to demonstrate that you have done X,Y and Z and are therefore worthy of a degree A,B or C.  Avoid long rambles about the state of the world or the urgent need to solve the problem of global warming, cheese shortages or  the need for a free market – stick it in a paragraph and move on to what you did.  It’s your achievements that will determine whether or not you get the qualification.  Engineers respect actions so show that you took some during your project.
  6. Read what you write.  Read every word of it out loud and slowly – you’ll be amazed at how many little errors you pick up doing this.
  7. Less is more.  A polished interpretation of your work with a few key graphs is much more valuable than a data dump of everything with no explanations.  Brevity is brilliant.

Now although I work at a higher educational establishment and must have examined nearly fifty MEng, BEng and MSc project reports in the last five years this post reflects my own personal views and not those of my employer.   It might also not apply to the particular institution where you work but is provided in the hope that it will be useful.

1. Yes I have read work that got up to one hundred pages before telling what they had done.  I was not happy.


To Desire or not to Desire

After an early difficult period me and my HTC Desire featuring the Android operating system started to get on.  The integration with Google calendar was great and a task manager called Astrid interfaced nicely with the Remember the Milk website.  The web browsing was very neat and the GPS and on-line maps were groovy.

Until on Sunday the phone decided to annihilate my text messages.  They all just vanished in front of my eyes – one moment the desktop widget was displaying them and the next “no messages” appears.  After two hours of trying I failed to get them back…..

So I no longer know what time my friend’s train is on Saturday, or what my sister wanted me to call about or whether everyone was there yet for curry at the Le Raaj last Thursday night.  O.k. the last one isn’t very important…  but to have a communications device that randomly drops data is really really annoying.

There and then I decided that a phone that lost my messages was worse than having no phone at all and reset it to factory defaults, cleared all my user data off it and put it back in the original packaging.  I’ve gone back to the trusty Nokia 6300 a pretty slick embodiment of the candybar phone form factor.  A phone which also so far hasn’t randomly deleted all my text messages.

Now I’m wondering whether I should go back to the shiny unreliable device which tells you the weather and has a five mega pixel camera or stick with a phone that actually works.  To Desire or not to Desire?  Whether to suffer the slings and arrows of outrageous programming or to take arms against a lack of regression testing?  Time will tell.