Do not blindly follow the data you have collected; make sure your original research objectives inform which data does and does not make it into your analysis. All data presented should be relevant and appropriate to your aims. Irrelevant data will indicate a lack of focus and incoherence of thought. In other words, it is important that you show the same level of scrutiny when it comes to the data you include as you did in the literature review. By telling the reader the academic reasoning behind your data selection and analysis, you show that you are able to think critically and get to the core of an issue. This lies at the very heart of higher academia.
It is important that you use methods appropriate both to the type of data collected and the aims of your research. You should explain and justify these methods with the same rigour with which your collection methods were justified. Remember that you always have to show the reader that you didn’t choose your method haphazardly, rather arrived at it as the best choice based on prolonged research and critical reasoning. The overarching aim is to identify significant patterns and trends in the data and display these findings meaningfully.
3. Quantitative work
Quantitative data, which is typical of scientific and technical research, and to some extent sociological and other disciplines, requires rigorous statistical analysis. By collecting and analysing quantitative data, you will be able to draw conclusions that can be generalised beyond the sample (assuming that it is representative – which is one of the basic checks to carry out in your analysis) to a wider population. In social sciences, this approach is sometimes referred to as the “scientific method,” as it has its roots in the natural sciences.
4. Qualitative work
Qualitative data is generally, but not always, non-numerical and sometimes referred to as ‘soft’. However, that doesn’t mean that it requires less analytical acuity – you still need to carry out thorough analysis of the data collected (e.g. through thematic coding or discourse analysis). This can be a time consuming endeavour, as analysing qualitative data is an iterative process, sometimes even requiring the application hermeneutics. It is important to note that the aim of research utilising a qualitative approach is not to generate statistically representative or valid findings, but to uncover deeper, transferable knowledge.
The data never just ‘speaks for itself’. Believing it does is a particularly common mistake in qualitative studies, where students often present a selection of quotes and believe this to be sufficient – it is not. Rather, you should thoroughly analyse all data which you intend to use to support or refute academic positions, demonstrating in all areas a complete engagement and critical perspective, especially with regard to potential biases and sources of error. It is important that you acknowledge the limitations as well as the strengths of your data, as this shows academic credibility.
6. Presentational devices
It can be difficult to represent large volumes of data in intelligible ways. In order to address this problem, consider all possible means of presenting what you have collected. Charts, graphs, diagrams, quotes and formulae all provide unique advantages in certain situations. Tables are another excellent way of presenting data, whether qualitative or quantitative, in a succinct manner. The key thing to keep in mind is that you should always keep your reader in mind when you present your data – not yourself. While a particular layout may be clear to you, ask yourself whether it will be equally clear to someone who is less familiar with your research. Quite often the answer will be “no,” at least for your first draft, and you may need to rethink your presentation.
You may find your data analysis chapter becoming cluttered, yet feel yourself unwilling to cut down too heavily the data which you have spent such a long time collecting. If data is relevant but hard to organise within the text, you might want to move it to an appendix. Data sheets, sample questionnaires and transcripts of interviews and focus groups should be placed in the appendix. Only the most relevant snippets of information, whether that be statistical analyses or quotes from an interviewee, should be used in the dissertation itself.
In discussing your data, you will need to demonstrate a capacity to identify trends, patterns and themes within the data. Consider various theoretical interpretations and balance the pros and cons of these different perspectives. Discuss anomalies as well consistencies, assessing the significance and impact of each. If you are using interviews, make sure to include representative quotes to in your discussion.
What are the essential points that emerge after the analysis of your data? These findings should be clearly stated, their assertions supported with tightly argued reasoning and empirical backing.
10. Relation with literature
Towards the end of your data analysis, it is advisable to begin comparing your data with that published by other academics, considering points of agreement and difference. Are your findings consistent with expectations, or do they make up a controversial or marginal position? Discuss reasons as well as implications. At this stage it is important to remember what, exactly, you said in your literature review. What were the key themes you identified? What were the gaps? How does this relate to your own findings? If you aren’t able to link your findings to your literature review, something is wrong – your data should always fit with your research question(s), and your question(s) should stem from the literature. It is very important that you show this link clearly and explicitly.
Need help with dissertation writing?
Our dissertation writing service, offered by our network of over 3,000 world-class academic writers, can provide you with a model dissertation you can use as a customised map to the results you need.
Find out more
This is the final post in my P&D series, where I will provide an overview of the Research Analysis and Writing the Dissertation stages.
The approach used for the data analysis differs depending on the research strategy and data collection methods used, so here’s a quick summary of the method I followed for Grounded Theory Research using Interviews.
As I alluded to in my previous post, one of the challenges I encountered was what to do with 15 interview recordings, each of around 30-45 minutes length – transcribing these was likely to take between 4 and 8 working days before I even started the analysis! However, reading the literature about Grounded Theory introduced me to the concept of coding: this involves taking the salient points from the interviews and assigning them to a ‘code’ which represents the essence of the statement. For example, the comment “a strategic alliance is where two organisations combine to produce a new product offering that does not exist today and brings value to both organisations” could be assigned the code “joint value-generating product offering”.
There are many different ways of undertaking qualitative data coding, included dedicated software packages such as NVivo. However, I decided to use Excel given that I didn’t want to spend time learning a new software application, and am very comfortable manipulating data in Excel. I started by listening to the recording of each interview and summarising the key statements made (as opposed to transcribing every word), as well as using a macro to record the timestamp for each comment:
Then I reviewed all the statements across all interviews, assigned each of them to one of my research objectives (eg. long-term strategy, success factors, organisational readiness), and gave them an initial code to represent the statement:
The advantage of this two-pass approach was that the initial recording of comments helped me become familiar with the data, and allowed me to get a view of the trends appearing between interviews, so I could choose codes that were consistent across all the comments.
The final stage of the analysis involved focusing on each of the individual research objectives separately, where I used an Excel PivotTable to summarise the codes assigned to each one. This was generally quite a long list (eg. 30-40 different codes), but reviewing all the codes together helped me to see trends or identify gaps in the level of detail. I was then able to return to the original data and re-code some of the statements (focused codes); this resulted in a more manageable list (eg. 10 different codes), which I could then compare across the different types of individuals interviewed.
An additional benefit to this approach (in addition to consolidating 816 statements down to about 60 codes), was that it provided a quantitive data set to work with alongside the original qualitative comments. Although the sample size was not big enough to draw statistically valid conclusions, I was still able to draw numerical comparisons of the data. More importantly, this exercise took around 27 hours rather than the expected 50-60!
Writing the Dissertation
Undoubtably the most daunting aspect of undertaking the P&D is writing the 15,000 word dissertation (or possibly longer with supervisor approval!). Although there is a typical structure for a dissertation, there is flexibility in how it is written. My advice would be to work closely with your supervisor to ensure it is structured in a way that suits them – not only because they have lots of experience in reviewing them, but also because they will be first marking your work. I had a really helpful supervisor who provided me with a pro-forma beforehand explaining what he typically likes to see in each section – by aligning my dissertation to this meant it would be presented in a way he expected, and ensured I did not miss any fundamental points. I was also in regular contact with him whilst writing my dissertation, not only to gain initial feedback on each chapter, but also to seek advice about topics such as the level of depth required for the literature review, and how to divide content between results, analysis and recommendations.
Here’s a brief summary of how I approached my dissertation
Introduction – I wrote this chapter prior to starting my research, and although I found it necessary to change it over time, I would definitely recommend writing this chapter as you start your P&D; it forces you to think about why your research is important, ensures you capture sufficient background information on the subject of your research, and encourages you to think about how you will go about your research – all of which help bring the P&D to life very quickly.
Literature Review – I’ve discussed the literature review in a previous post, but one of the biggest challenges I faced with this was how to avoid writing too much. Having a specific set of research objectives helped with this, as it provided focus to the review, and avoided me discussing the many interesting (but not relevant) findings on the topic in general.
Research Methodology – This chapter was slightly easier to write than others, and like the introduction I would suggest writing this before undertaking the research, as what you learn whilst writing this section may change how you approach the research itself – as I mentioned previously, I originally planned to pursue a case study approach, but writing this section resulted in me changing to the grounded theory approach, which was far more suitable to my P&D objectives.
Results – I encountered some challenges writing my Results chapter – the initial feedback by my supervisor highlighted that I had used this section to begin my analysis, with the consequence that I did not display my results effectively. This highlights the importance of getting your initial draft to your supervisor early, as I was able to work closely with him to understand how to split content between the Results and Analysis chapters. Following a number of discussions, I focused on using the Results chapter to compare the various statements and codes identified for each of the two sample groups I interviewed. I used a combination of text, tables and charts to make it easy for the reader to interpret (and also to keep my word count under control!), which also helped highlight trends and patterns for discussion in the analysis chapter. Another minor challenge I encountered when writing the results was how to use references for comments made by interviewees, as all interviews were anonymous. I overcame this by using a code to classify different types of interviewees (eg. DH1..5 were department heads and SH1…5 were individuals working with strategic alliances).
Analysis – The analysis section focused on evaluating the research findings against the literature review. This was a much more descriptive section than the Results chapter, and the section that required the most thought. It was really interesting to write this section, as bringing the literature review into context both helped explain some of the results I had seen, and also allowed me to identify some new theories: both specifically to the subject of my research, and also more generically to the field of strategic alliances itself.
Recommendations – This section is essentially the outcome from your P&D; ie. after undertaking the research and analysis, this is what my recommendation is moving forward. Having spent four months looking into the subject matter and speaking with many colleagues about the topic, I started writing this section with a view on what my recommendations would be. However, by taking the time to justify these recommendations using my literature review and analysis, I felt that I could really stand by these recommendations as they were proven by my research, as well as identifying some new recommendations that I had not previously considered.
Conclusion – There were three parts to my conclusion; firstly I stated a specific answer to the original research question. Although this may seem obvious, this is something that I didn’t naturally bring out in my first draft, and yet including it provided the dissertation with a strong point of closure. I also used this chapter to highlight some of the limitations of my research, suggest how it could be generalised to other subjects, and also propose potential ways of taking the research further.
And that was it … 6 months later, 110 pages (including appendices and references) were submitted to Warwick Business School. Not only did this represent the completion of my P&D, but it also marked the completion of my MBA – pressing the ‘Submit’ button for the final time was a wonderful feeling! Since then I have had my dissertation marked, and my MBA has been officially confirmed by the University. Next step … graduation!