Empirical Software Engineering

Hello, my name is Oliver and this is my blog for the ISAD 504 Module of my MSc.

The first seminar session that I attended was on Empirical Software Engineering where I learnt the importance of applying scientific techniques to the practice of software engineering. This is a systematic and logical approach that is based on the principles applied to scientific experimentation whereby we can take measurements and devise hypotheses based on our results.

This approach is somewhat contrary to what us coders tend to do in our work where people such as myself tend to jump in without too much planning or thought about why we're doing what we're doing. Also, I myself have seen this approach in industry where the boss decides that they need to use the latest technology because everybody else is or because somebody recommended it. In particular I found this with social media where people decide that they need to get into the social world without really thinking through their aims, objectives and the method that they will follow to achieve these.

As recommended in Shirley's blog I read this article : http://www.netmagazine.com/features/15-top-web-design-and-development-trends-2012 which I found quite interesting.

The section on the demise of flash and it's predicted resurgence was quite interesting as most people seem to be writing it off these days , also it was a surprise to see Ian Lobb quoted! I found this article has quite an interesting point and also shows that developers tend to be somewhat elitist and determined to use the latest technology rather than the best tool for the job. It also raises the question as to whether the business case is driving the use of the latest technology.

Section 12 of the article regarding mobile workforces was also interesting as we're entering a time when many people are choosing to shun the 9 to 5 lifestyle in favour of freelancing, whether this is because it is difficult to get a job at the moment or some other factor I believe that this can only be a good thing. I expect that we will begin to see many more small (no more than five person) online startups.

In terms of this article and empirical software engineering it would be interesting to measure the engagement time of users on a flash based website against a static website and one written using html 5 and to gather the user's feedback, this could then be used to see whether the technology is improving the user experience. It would also be interesting to look at how much software is produced by small (less than 5 person) teams and their annual turnover to see whether the team size affects the level of success achieved. In this case it would be interesting to see if an end user prefers the software developed by a small company over that of a larger one. This would be of particular use with regards to apps where anyone can create them without a large financial backing.


I stumbled across this article during my reading which actually should have some relevance to my final project. The article titled Semantic web for e-learning bottlenecks: disorientation and cognitive overload quite interestingly studies the limits of what a learner can take in at once and applies it in a web context. This is quite interesting as we can often be bombarded with information on a web page and this can disorient us. In the context of empirical software engineering we could experiment with the amount of different content on a page and user's engagement, from this we could deduce the optimal amount of content that should be shown in our software.

Finally, it seems a bit strange that people are expected to pay for scientific journals these days when information is shared so freely across the web. In fact in this video http://www.ted.com/talks/lang/en/michael_nielsen_open_science_now.html Michael Nielsen provides an interesting view on the currently closed position of academics and makes the case for open research whereby anybody can contribute their ideas and thoughts to a project and he also highlights how freely sharing information has created some powerful open source software.