Route B: Generalisation

All of the approaches that can be adopted within Route B: Generalisation could require you to make changes to the research design used in the main journal article when setting your own research design (i.e., changes that reflect your choice of experimental, quasi-experimental or relationship-based research design). However, the goals that have been set within the main journal article (i.e., your goal of describing, relating or comparing) are likely to remain the same. Nonetheless, the nature of these changes and the extent to which changes should be made will depend on the approach you have adopted.

Most population-based approaches involve relatively small changes in the characteristics of the population between the main journal article and your dissertation, if any changes are required at all. For example, you may simply be examining whether a new teaching method was as effective at increasing exam performance amongst postgraduate students (i.e., the new population in your dissertation) as it was amongst undergraduate students (i.e., the population studied in the main journal article). In such a case, you can probably just use the research design applied in the main journal article (e.g., an experimental, repeated measures research design, assuming that you were also able to randomly assignment participants to groups, and apply the same treatments used in the main journal article). However, sometimes the characteristics of the population in the main journal article and your dissertation will vary to such an extent that some changes to the research design used will have to be made. For example, imagine that you wanted to examine whether the same teaching method was as effective at increasing exam performance amongst school children. In such a case, there are a number of reasons why applying the same experimental research design is difficult if not impossible. School teachers may have less autonomy to change their teaching methods than lecturers at university, making it difficult to apply a new teaching method (i.e., the local education authority may have strict rules on the curriculum taught to students and the way that such a curriculum must be taught). Also, at the school level, it is often viewed as unethical to provide students with different teaching methods (i.e., one group of students follows the teaching method they always did, whilst a second group is given the new teaching method). This is more likely to be the case if previous research has shown that one teaching method (i.e., the new teaching method) may be more effective, giving those students that are given this new teaching method (i.e., the treatment in the experimental research design) an advantage over the other group. As a result, you may either be unable to use such a research design or have to make changes to such a design to suit the new population being studied (i.e., in this case, school children). If the ethical issue does not prove to be a problem, it may still be possible to carry out an experiment with a treatment group and control group, but it may not be possible to randomly assign participants (i.e., students may be already assigned to different classes). In such a case, you may have to carry out a quasi-experimental research design, rather than an experimental one. However, if you cannot separate the groups at all, you may be forced to use a relationship-based research design (i.e., a non-experimental research design), which is also likely to lead to a change in research methods.

If you are taking on a context/setting-based approach or time-based approach within Route B: Generalisation, the same rules apply as above. You have to examine the changes in the characteristics of the context/setting or time before the main journal article and your dissertation, and assess whether you will need to adjust the type of research design used in the main journal article when setting your own research design.

However, if you are taking on a treatment-based approach within Route B: Generalisation, you are likely to have to make quite a few changes to the research design used in the main journal article when setting your own research design. After all, the focus on a treatment-based generalisation is to examine whether the treatment characteristics have to be the same when applied to different populations or settings/contexts to arrive at the same conclusions. For example, imagine that the main journal article showed that a new teaching method was more effective at increasing exam performance amongst undergraduate students than the traditional teaching method. In the main journal article, the students in both the control group (i.e., given the traditional teaching method) and the treatment group (i.e., given the new teaching method) receive two 1-hour lectures each week for a 12 week period. However, your treatment-based approach aims to examine whether different treatment characteristics would lead to the same result; characteristics such as the number of lectures given, the interval between each of these lectures, the length of the lectures, and the time period of the experiment. Perhaps you want to know if the new teaching method is as effective at increasing exam performance if the length of the lectures was different (e.g., 1 hour, 1.5 hour, and 2 hour lectures). In any of these cases, a treatment-based approach within Route B: Generalisation will require you to make a number of changes to the research design used in the main journal article when setting your own research design. However, it is likely that you will want to change the treatment characteristics within the existing research design (or a similar design) rather than changing the research design fundamentally (i.e., you would change the length of the lectures, but still maintain an experimental or quasi-experimental research design, as was used in the main journal article, rather than switching from an experimental/quasi-experimental research design to a non-experimental one, which would also lead to a change in the research methods used, and make comparisons between the results from the main journal article and your dissertation more difficult).

Route C: Extension

Like in Route B: Generalisation, all of the approaches that can be adopted within Route C: Extension could require you to make changes to the research design used in the main journal article when setting your own research design (i.e., changes that reflect your choice of descriptive, experimental, quasi-experimental or relationship-based research design). Also like in Route B: Generalisation, the goals that have been set within the main journal article (i.e., your goal of describing, relating or comparing) are likely to remain the same. However, whilst most aspects of the research design applied in the main journal article will remain the same in your dissertation, there can be significant changes to the research design used in your dissertation in the case of design-based extensions (one of the approaches within Route C: Extension). Before we discuss the implications of this approach, we touch on population and context-driven extensions.

If you are carrying out a population or context-driven extension, the implications for your research design will be similar to those of population or context-based generalisations, discussed in the previous section. Whilst it may be assumed that since extension goes further than generalisation, you would have to make more changes to the research design during a population or context-driven extension compared to a population or context-based generalisation, this is not necessarily the case. After all, such extensions typically differ from generalisations because constructs/variables have to be omitted, modified or added in population or context-driven extensions. However, this does not mean that such omissions, modifications or additions in constructs lead to a fundamental change in research design. Rather, like in population and context-based generalisations, most population and context-driven extensions will only require small changes in research designs, if any changes are needed at all.

Final thoughts

Before moving on to STEP THREE: Research methods, make sure that you are in no doubt what the goals of your research are (i.e., your goal of describing, relating or comparing) and the research design that you plan to follow (i.e., your choice of descriptive, experimental, quasi-experimental or relationship-based research design). This is important because your choice of research design and your goals will influence the choice of research methods you select, together with the route you have selected, and the approach within that route, which we discuss next.