Very often, when referring to ethics in research, we are concerned about the people involved from whom we have obtained data that will subsequently be used to analyse our research problem. We need to protect the integrity, welfare and rights of the individuals who assisted us along the way. There is, of course, more to the ethical principle in research, as it covers the complete journey from formulating research questions to publication as we seek to advance knowledge through our work.

Some of the most relevant ethical principles all scholars should be aware of are:

1. The importance that our respondents give free and informed consent,

2. We guarantee their privacy and confidentiality,

3. We minimise harm and maximise benefits.

When researching a case-study environment and more specifically on an individual organisation, we need to ensure that the company has given its consent for the research to be undertaken, be named, and determine the level of exposure the final document should receive. Also, individuals must provide their support to be mentioned by name in the paper.

When transcribing interviews, always send the transcript back to the respondent to confirm that it is an accurate representation of what was discussed in the interview. I suggest that everything related to ethics is done in writing.

I do not doubt that your University has its code of ethics that you need to follow in conducting your research. Make sure to be up to date with this policy document.

by Prof. D. Dingli

When you are thinking through your thesis topic – long before you start writing up – you need to make sure you have a very clear and significant business-related problem validated by subject matter experts (SMEs). An SME is an independent, experienced and interested person, willing to help you, someone who knows your topic well and can confirm (or deny) the significance of the problem and the extent to which it is real and important. One SME is not enough. Three or four are better, and can help you with the next step – deciding on what you want to study and what is beyond the scope of your topic. This will help you focus. You mustn’t be too broad in your approach, or too narrow.

Your research question must not be too obvious or too easy to answer, or too difficult and impossible. You will then need to read enough literature (and possibly conduct even more interviews with your SMEs) to establish that this really is a new and interesting area of study. Your next job will be to explore the relevant theories and models which relate to your topic. These will help give you a theoretical framework for you to apply to your own context. Your context could be geographical or industry or sector-based, and might involve some comparisons. Then, you will be in a position to start thinking about your specific research questions. Again, make sure that your research problem and therefore your research questions are not too easy to answer, and it’s not too easy to collect data. Your examiners will expect you to study something challenging to show off your skills!

by Prof. K. MW.

You should only start reading when you have completed thinking through your research objectives and have “crash-tested” your research topic ideas as much as you can. You need to know the terminology of your subject, so you can start trawling through the literature. For example, if you are looking at employee retention – you might also look at “commitment”, “turnover” and other related topics. You should read widely about relevant theories and models before you make your choice of which you will use – and justify your choice, even on the grounds of convenience or parsimony – if it was because it was easily and economical. One of the most important things about your Literature Review is that you should try to create a dialogue with the literature items by comparing and contrasting the different studies and theories. Your job is to act as a moderator between them – questioning everything you read critically, and summarising the latest “state of the art” on your topic. You should return to this analysis when you have completed your data collection and written up your findings – have you found something new? Or confirmed existing reports? Ideally your research questions should be “symmetrical” – it shouldn’t matter if your research results support your hypotheses or propositions or those of other authors – or they don’t. Research is about searching for truth – whatever it is. And that truth might be already known – or not.

What is literature? Be careful with news, reports or any type of media information (such as social media posts) about the subject you have chosen. This may not be seen as “literature” by your thesis examiners – so you need to justify this carefully and consider using it in a section looking at background material only. You can’t go wrong if you focus as much as possible on fairly up-to-date peer- reviewed sources that are published in reputable academic platforms. If the authors are mostly university professors and the publication is called “The Journal of…” then this is promising… And focus only on relevant literature – discard irrelevant and outdated material – be selective – go for quality not quantity!

Use the APA manual to understand the difference between quoting, para-phrasing and using ideas. Always apply the APA manual accurately (therefore correctly) and thoroughly edit your work before handing in. It doesn’t matter how often you do an edit – you will always find one or two more gaps, omissions or duplications!

Keep reminding the reader of your research topic and research questions throughout your dialogue with the literature. Avoid disintegrated and disconnected islands of knowledge and quotes. Try to create a “map” of your research topic, where the links between each topic area are clear and it’s all inter-connected. And when you have done as much as you can, try developing a literature review table in which you summarize each article or study using columns and rows focusing on the main themes. Most examiners love this kind of stuff! And it shows a well-organised mind at work.

by Prof. K. MW.

This is not a common type of research design to be found in business related thesis papers, although it could provide some interesting insights n this setting too.

You have to think of experimental research as always having a minimum of two groups. One that receives some kind of treatment and one that does not, provided both groups are similar on all other relevant factors.

Random sampling is the only way to ensure that each group’s characteristics are similar or that anything else that could influence the result is randomly distributed between both groups.

Maybe one group could be exposed to a 35% price reduction and the other (control) group is not to observe what happens to demand when prices are lowered. In this way it would be possible to conclude that a decrease in price does cause an increase in demand if such behaviour is observed.

So the question is: does variable “A” cause variation in variable “B”? and how large is that variation?

In education, we might decide to expose one group to a particular teaching methodology to determine if it enhances learning or memory. This is what is known as testing for causal hypothesis. Example. Hypothesis: A 35% reduction in price for item X will induce a 10% increase in demand.

It could be possible that you create 3 groups where one group is now exposed to a price reduction of 15% to observe what influence that could have on demand.

Experimental design is a predictor of phenomenon. Design features could be conducting a pretest and a posttest on the subjects before conducting the experiment. Conducting a pilot study would be a wise decision too. It will ensure that the experiment is designed to measure what it should be measuring.

The researcher must always be aware of the dangers and threats of randomized experimental research design. There are three main causes that could negatively influence causal validity.

1. Contamination – this is where there is a possibility that subjects in different groups meet and interact to ensure the control group do not receive any treatment due to that interaction

2. Crossovers – This is where nonparticipants receive the treatment

3. Attrition – losing subjects that took part or should be taking part in the experiment

by Prof. D. Dingli

One of the most challenging tasks in research is designing a questionnaire through which data needs to be collected. All too often, scholars fall into the trap of using poorly designed questionnaires, collect tons of data and then apply every statistical tool in the book to that data. Unfortunately, if the data collection instrument is faulty, then most likely, the data analysis, results and interpretation that follow are flawed.

The first problem begins with the construct being measured and therefore, the questions being used to measure that construct. Respondents with similar score should be equal to each other with respect to the construct being measured and likewise different to each other if their scores are different. This concept determines the level of “discrimination” within the questionnaire.

Another critical concept is the validity of the measurement. Are you really measuring what you should be measuring? If a questionnaire is designed to measure “assertiveness” then it should not include questions related to eating habits.

The third important issue to ensure good questionnaire design is Reliability. The questionnaire should produce the same results under the same conditions, however, validity is an underlying condition for reliability.

Steps that you need to take in questionnaire design are:

1. Determine the construct you intend measuring – read about it to understand the underlying factors that constitute this construct

2. Select your measurement scale – Are you planning to use a YES/NO response or a Likert 1-5 scale?

3. Meet with Subject Matter Experts to discuss the items to include to measure your construct. Theory helps, however, expert advice could assist you to fine-tune those items. Avoid repetition of items or writing out statements that measure relationships rather than a single construct. Avoid bias or leading statements

4. Administer the questionnaire as a pilot study. Try it out on a select number of typical respondents to determine (a) if the statements are understood the way you wanted them to be. (2) To examine the distribution of responses (3) Test reliability (4) Fine-tune and make adjustments to questionnaire based on feedback

5. Administer questionnaire on sample population and collect the necessary data. Even this stage requires some thought. Are you going to collect data face to face being present as your respondent completes the questionnaire? Are you planning a telephone survey or will this a self administered questionnaire sent by mail or through some form of electronic means?

6. Carry out the relevant statistical analysis on the data collected.

by Prof. D. Dingli

1. In deductive research we tend to start from a strong theory. It is referred to as the top down approach. Based on the theory of interest a set of hypotheses may be stipulated which need to be tested. Data is then collected and processed in a way to be able to test the hypothesis. This is normally achieved through quantitative statistical data analysis. The result will either prove or disprove the hypothesis that was formulated. This approach treats society from a positivist stance. Results must be unbiased by the observer or the participants. This approach normally deals with numerical data as it reduces the possibility of any conclusions being influenced by external factors.

2. In inductive research, we could say, works the other way round. The starting point is the observation of reality and hence more appropriate for qualitative data. The data is manipulated in a way to determine patterns and themes, which then give the researcher scope to establish a hypothesis of interest worth exploring. The end results could be the emergence of a new or modified theory based on the derived conclusions. This kind of research takes on a very open-ended approach, while deductive research initiates with a very limited and well defined scope. It leans on the interpretivist methodology which is concerned with the collection of qualitative data, often through unstructured interviews or participant observations to collect this kind of data.

3. It is not unusual either to see both approaches combined in the same research as this produces a circular design where theory leads to observations which could lead back to new theories.

4. Essentially, the difference between the two approaches is centred around the researcher views the nature of reality. Quantitative approaches believe in the single version of reality that is objective and measurable and purely based on scientific principles. The qualitative researcher believes that reality is centred on the individual and that many natures of reality could exist generating different meanings for different people.

5. Scholars need to look back at their research questions and objectives to clearly understand if they are out to test theories through a search for evidence to support or refute their hypothesis or if various individuals need to be observed to derive themes which would lead to theories.

by Prof. D. Dingli

Before scholars embark on writing out a thesis, it is customary that they first submit a thesis proposal to their University. There are many reasons for requesting a proposal. Some of the most common would be determining if the student has an exciting topic to write about that will be acceptable for the course of studies, select an appropriate supervisor, and mainly challenge the student to think conceptually and holistically about their thesis.

The structure and length of proposals could vary from 10 pages up to 5,000 words. Some Universities have stringent guidelines, even where word count per section is concerned, while others are more flexible. What is important is that your proposal includes the most critical areas, these being:

1. A clear working title that could be modified at a later stage. However, a title clarifies the most essential reason behind the thesis.

2. Background – brief but essential. The topic is explained in a broad context introducing the reader to the subject, including a build-up to the next section.

3. Problem Statement – What do you want to solve? Is this a practical issue, a theoretical one, a company problem or an industry one? Clarity of a written problem statement is paramount.

4. Emerging out of the problem statement would be your Research Questions and research objectives. The reader needs clarity as to why this topic is worth investigating. What is going to be learnt if the research questions are answered? What will this study add to the body of knowledge on the topic?

5. Methodology – What kind of data needs to be collected, and how will the scholar collect it? This section also needs to explain what kind of data analysis would be necessary to be undertaken so that the processed data may be interpreted and conclusions derived in ways that the research questions are answered, and the objectives are met.

6. Some form of literature review must also accompany the proposal. Scholars need to show evidence that they have read the literature and can frame their problem within an academic dimension underpinned by a solid theoretical foundation.

The proposal might need to go through several drafts and revisions, especially once a supervisor gets involved in the process. It is an opportunity to identify fatal flaws in the paper's conceptual ideas that might not be resolvable at a later stage.

Scholars may be disheartened at this early stage when they see their supervisor sending the work back for updates, changes and improvements. The scholar worries as this is not even the thesis itself and assumes that the challenge ahead will be insurmountable. In fact, this process is explicitly designed to ensure that the thesis writing will develop without any unnecessary hitches along the way. Once the paper's major problems are resolved at the proposal stage, a thesis usually flows with ease.

by Prof. D. Dingli