Research
Although research is a term in common use, many people use it incorrectly or carelessly. In everyday life, this is rarely a problem, but when you are doing an academic project, knowing what research is and to what extent you are doing it is important.
Background research
Before undertaking any kind of project, it is a good idea to find out what you can about the problem, the technology you will be using,the client if there is one and what work has already been done in the area. If you are doing a very practical project, then background research is a good term for this activity. In the context of a major academic project (Honours or Masters level) this type of activity is covered by the literature review and that is the preferred term. Please see the literature review section of the notes for more detail.
Research questions
In general, research is about using data to draw conclusions about some question that is identified in advance. There are exceptions to this - some types of research such as grounded theory seek to develop theories directly from observations - but they are unusual.
A research question can be understood as a further development of the aim of the project. Having one or more clear research questions can help to structure the project well. It allows you to set clear and specific criteria for the type of data you need, the appropriate research methods to use, the data analysis methods needed and the evaluation process. Research questions are discussed in more detail in the Methods section.
Primary and secondary research
If you are collecting your own data in order to answer your research questions, then you are doing primary research. The data can take many forms such as simulation results, physical measurements or survey responses. With primary research, a lot of effort needs to be devoted to the data collection method to ensure that the data is valid. The advantage is that the data collection can be specifically tailored to the aim of the project.
If you are using data that already exists in some form - published datasets, previous publications or other documentary sources, you are doing secondary research (sometimes called desk research). The data was originally collected by someone else for some other purpose. That is more convenient in the sense that it already exists and so you do not need to define the data collection method. On the other hand, it is important to define your selection criteria and any other methods used to ensure that the data is suitable for the purposes of your project.
Research methods
In all types of research, a systematic approach is essential. There are many research methods to choose from, and the first challenge is to select the one that you think most appropriate. After that, it is a question of following the rules for that type of research. At undergraduate and Masters levels, you need to demonstrate that you are aware of research methods and can follow established procedures. At PhD level, by setting out a systematic method, you are establishing the grounds on which your results may be judged valid and reliable.
Your choice of method depends on the type of project you are doing. Here are some common cases:
-
Are you comparing two technologies or methods for solving the same problem?
Consider an experimental approach. This involves eliminating any unwanted influences, setting up a test scenario, collecting data on the two alternatives and comparing the results. Ideally, you would define hypotheses in advance and use statistical methods to process data from several randomised trials.
-
Are you evaluating the application of a theoretical approach in a particular context?
Consider a case study. This involves the systematic collection and analysis of data from multiple sources related to the case in question. Ideally, several data collection methods should be used so that their results can be cross-referenced. This is known as triangulation.
A multiple case study collects data related to the same phenomenon across several different contexts and compares the results. This approach is particularly useful for developing theoretical models on the basis of observation.
-
Are you interested in the effect that you own actions have on an existing context?
Consider action research. This requires you to document the situation before and after the intervention as well as the nature of the intervention itself. A major consideration in this type of research is guarding against confirmation bias. It is essential to define a clear objective procedure and to show evidence that it was adhered to.
Whatever method you use, the more carefully you apply its techniques, the better for your project. If you are tempted to take a particular approach because it looks like an easy option, you will be doing yourself a disservice by avoiding the detail. This will have a detrimental effect on your marks. The links at the foot of the page provide more information on the range of research methods available.
The non-research option
At undergraduate level, and to some extent at Masters level too, it is possible to base a project on a practical topic rather than a research-oriented one. The typical example is the design and delivery of a software application. Because this type of project does not have an explicit research dimension, you need to demonstrate the same graduate-level qualities in a different way. For example, you should select an appropriate theoretical approach to the design and construction of your application, and you should apply appropriate techniques rigorously as demonstrated by supporting evidence. You should also include a well-designed objective evaluation of the final result.
Writing Good Software Engineering Research Papers