Dissertation writing is a great challenge for most students. Having successfully helped more than fifty PhD candidates of business management during past five years, Dr Mike Fung would like to share his experience about how to write a dissertation. His method is also applicable to other Master thesis or bachelor papers.
How to choose a topic
Many students chose their favorite topics, but they often neglect the following points:
-
Are the collected data abundant and complete
First of all, please make sure you gathered abundant and complete materials. Otherwise, it is hard to write a good dissertation. In this case, you should consider another topic.
Sometimes, on the basis of an English literature paper, certain paper could be written via changing the place of data collection. For example, the paper 'Snack Bars in Hong Kong' could be derived from 'Snack Bars in UK'.
-
Is it a substantial topic? Who could benefit from it
The topic must be something that matters, such as strategy, market, corporate culture and so on, which is important for many relevant persons. And above all, it should look like a doctoral thesis topic. The following two examples were rejected topics. 'Communication Skills in Hair Salon' is gossip news tricks which can hardly appeal to refined taste. Another one, 'The Market Mesearch of Marriage Planers, is too simple to fit for a Ph.D. paper.
-
Is there appropriate methodology
The methodology of the thesis should be demonstrated in one chapter. The methods need to adapt to the requirement of the dissertation, as follows:
-
The research methods employed in the thesis are
- Action Research
- Quality Research
- Quantity Research
- Secondary data research
- Primary data research
- Building a model
- Improving an existing model
- Testing or Implementing an existing model
-
Data Sources
Ensure that data and information are accurate and reliable.
-
Study Groups and Data Collection
The definition of study groups should aim at answering the research questions. The data collection should be accurate and reliable.
-
The sample design
Narrow down the research scope so that specific study could be carried on. For example, if the research is about 'Hong Kong People's Viewpoint of Healthy Food', it's better to gather questionnaire survey sample from a representative small chamber of commerce rather than by cold call.
-
Research Tools, Legitimacy and Reliability
Research tools could be commonly used theoretical analysis, statistical analysis, model analysis, first-hand information and secondary resources. The legitimacy and reliability are provided by appropriate usage of the tools.
-
Data processing and statistical treatment
Excel or SPSS are usually employed as data processing software. Take heed that student version of SPSS might be inadequate due to its limited case processing ability.
-
Assumption or Hypothesis
Normally every dissertation has an assumption or hypothesis, which needs to be confirmed or disproved by evidence. Relatively lower level of data support is enough to prove an assumption, such as mean, median or mode. By contrast, proving an hypothesis demands higher criteria, like more than 95% or 99% of statistical significance.
Generally an assumption or hypothesis could be fractionized into several smaller assumptions or hypotheses. For instance, the assumption 'Hong Kong Civilians like McDonald's Food' could be fractionized into the following assumptions:
'Hong Kong civilians like McDonald's hamburgers';
'Hong Kong civilians like McDonald's fish fillet burgers';
'Hong Kong civilians like McDonald's French fries';
'Hong Kong civilians like McDonald's Red Bean Pie', etc. -
Research Design
There are a series of issues relative to the thesis topic, involving assumption, hypothesis, theoretical reasoning or data compiling. Readers should consult those articles.
-
Questionnaire survey design
It's the outline of the questionnaire.
-
Question setup
Specific questions listed in the outline.
-
-
Is it convenient to collect and analyze data
Prior to the research, the author need to know if the data collection is feasible and easy to obtain. It's hard to obtain information and data for certain topics, such as 'The tax avoidance in China', 'The state of business integrity' or 'The operation of Hong Kong ICAC (Independent Commission Against Corruption)'.
-
Is the thesis unique and original
The thesis must be a unique paper instead of a replica of others' work. The more innovative value it has, the better your thesis is, such as 'The influence of China High Speed Railway', 'The development of democracy under socialism with Chinese characteristics', 'The Japanese nuclear crisis', etc.
Changing research title provides an easy access to compose new topics. Firstly, data could be collected from different regions/countries/industries/markets. Secondly, narrowing down the investigate scope might do the trick, for example, replacing a country with a province, a town or an enterprise. Thirdly, you can always make a different match, such as turning 'the loss of restaurant employees' into 'the loss of trading corporation's employees', 'quality management' into 'strategic quality management', 'project management' into 'strategic project management', 'cost effectiveness' into 'strategic cost effectiveness', 'enterprise culture of factories' into 'corporate culture of snack bars', and so on.
Thesis structure
Commonly there are five chapters in a dissertation, including:
-
The problem and its background
In this chapter, one should clearly state what problems of the research and their background are. Ordinarily many papers are learning experiences of authors. Their background could be the authors' working context.
-
Theoretical framework (literature research)
This part is a review of literatures relative to the thesis. If the cited literatures have direct connection to following research analysis, the thesis could be much more convincing.
-
Methodology
It contains the content of (1)C, which should be covered in university courses.
Usually a good dissertation creates a new model or an improved one, and then proves its validity with data. Poor quality do the research by applying other's model on data gathered from a new domain.
The data source could be second-hand（i.e. it is collected by others）. In the case of the thesis 'Medical Tourism', the author found a comprehensive report on internet, which is mostly news on international newspapers. He wrote a great paper by analyzing and commenting the news, along with his firsthand investigation in Hong Kong.
The number of surveys should be more than 32. This is the bottom line of acquiring reasonable analysis in a small group questionnaire. Normally a survey involving 100 to 200 participants is pretty well.
-
Research result and analysis
In this step, the author compiles research results and performs an analysis. Analysis could be theoretical or quantitative. Data analysis is statistical analysis which might be simple or complicated depending on circumstances. For example, the data analysis of finance papers is very demanding. If the paper implemented a model, its data analysis also requires fairly high standard.
-
Summary and conclusions
Summary describes research results. Conclusion is the final settlement of the analysis and research, which should be both instructing and enlightening.
The method of writing
It would be better to make up an outline and key points first, followed by writing down the description of each point. In the end, words and sentences are linked into an intact dissertation. Never write them all at once, otherwise one might find it's impossible to put those arguments together later.
The Japanese research method is to design the framework of summary and conclusion first, and then do the research. In such a way, the research won't go in the wrong direction.
About the Author
Dr Fung is a Professional Engineer of Canada with more than 20 years experiences in project management of application software development. In addition, he had supervised numerous industrial projects as well. He is an international WHO'S WHO of Professional of USA (1988).
For reader enquiries, please contact: Prof. Dr. Mike Fung
Telephone: (852)536-28-567
Email: dr.mikefung@yahoo.com.hk