Having satisfied yourself that a topic for research is feasible, you then need to persuade others (supervisor, manager or lecturer) that the research is not only feasible, but valuable too. This applies in cases where you generate your own topic and also to situations where the broad area of concern is dictated to you.
Usually, it is possible to discuss potential research informally, but eventually an area of interest must be chosen, and the topic analysis is often the first attempt a researcher makes to formally justify the intended research.
A topic analysis of two or three pages is a way of summarising various aspects of what may be several potentially acceptable topics. In most cases, it is the first step a researcher makes toward developing a proper research proposal and ultimately completing a study. The topic analysis and the research proposal are actually of similar structure, although the latter provides new (and usually more technical) information, as well as elaborating many areas of the topic analysis.
Some degree of planning is necessary at all levels of research. At this stage, your answers to most questions may be vague and uncoordinated – probably still being held within your own mind. The aim now is to develop a realistic plan of action with clear objectives, which takes account of resource constraints, and which has a likelihood of being achieved.
A further advantage of submitting views in writing for consideration by others is that redirection of research can be accommodated more readily now than at later stages.
A topic analysis usually covers the following areas:
2. Problem Statement
3. Research Objectives
4. Prior Work
5. Hypothesis [in the case of Quantitative Research]
6. Probable Approach
These are outlined below, together with some examples. The depth required in most sections is not great as there are benefits in being cautious during early stages of research. However, some sections in the topic analysis do require you to be very clear about what you want or what you expect to achieve – notably the problem statement, the research objectives, and research value.
The project should be given a title that indicates its scope and content e.g. “Market Opportunities for Australian Beef in Japan”, or “Job Satisfaction among Telstra Call Centre Staff in Melbourne”. The title should be concise however, so expressions like “A study of”, or “Comments on”, which are often redundant, can be left out in many cases.
2.2 Problem Statement
The “problem” in the study may be an unsatisfactory situation that has been encountered. Alternatively, it may be an existing state of affairs that may not seem problematic but merits investigation.
a. The Trobe Company is a major manufacturing company in Campbeltown, Melbourne. Management is faced with high absenteeism in the packing department, which is leading to difficulties in allocating work and delays in meeting production targets.
b. Over the last decade expenditure on racing in Victoria (as a proportion of total gambling expenditure) has declined dramatically. Expenditure on lottery products (lotteries, lotto, pools, and instant scratch-its) has declined in a similar fashion over this period. At the same time, expenditure on casino gaming (wagers on table games, gaming machines and casino keno systems) and gaming machines (machines in clubs and hotels) has increased equally dramatically.
Try to avoid problem statements that are concerned only about dealing with a situation. Focusing on the means to the exclusion of ends, especially at an early stage of the research, can be a limitation – the potential “solutions” to a problem will always depend on the particular problem set.
By changing the way you look at a situation, the problem also changes – opening up a whole different array of possible solutions. To illustrate, the problem statements for the Trobe and Victoria gambling examples may have been written thus –
a. The management of the Trobe Company must determine how to reduce absence rates in order to achieve planned work allocations and production targets.
b. Racing Victoria Limited and Tattersall’s must examine potential means by which horse racing and lottery products can be marketed to ensure their survival against the impact of casino gaming and gaming machines.
Both examples suffer from similar faults:
i. They do not take a flexible, open perspective of what are both very broad and complex situations. Instead, example a. focuses immediately on absence rates and example b. focuses on marketing – probably reflecting the researcher’s premature belief about where the solution lies.
ii. Both imply that “problems” spoil an otherwise good situation, so therefore must be eliminated. This is casting the “problem” in a negative light.
iii. Likewise, they take for granted that returning to a former problem-free state is best – not allowing for any criticism or evaluation of that state. This is casting the old state in a positive light.
iv. By doing this the researcher is not recognising that change can have beneficial aspects.
With problem statements like those above, a researcher would be unlikely to contemplate things like –
i. changing out of date work allocations and production targets.
ii. identifying all of the factors that contribute to absence.
For Victoria Gambling:
i. changing the product mixes for racing and lottery products
ii. withdrawing obsolete products, e.g. the pools and reinvesting in others.
Although the problem statement need not include all sub-topics, it should provide a brief and accurate synopsis of the area for study, and indicate clearly to the reader where the point of interest lies. The following statement is not very helpful –
“The Workplace Relations Act 1996 makes Equal Employment Opportunities programmes mandatory in the Public Sector. The government is also exploring ways of implementing EEO in the Private Sector.
It is considered that awareness and activity in this area is low in the Private Sector. To overcome this problem the Government is contemplating legislation aimed at addressing this situation in the Private Sector.”
i. Who considers awareness and activity to be low?
ii. What is the link between awareness and activity?
iii. What is the problem?
2.3 Research Objectives/Purpose of Research
Often, the problem is likely to be too broad to be expressed as a single issue, and the process of defining it will raise several questions. The overall purpose (objective) of the study is to answer these questions, so they must be explicitly addressed at an early stage in the study.
The specific objectives (questions to be answered) for a study of the Trobe Company could be:
a. To estimate the:
i. actual level of absence and
ii. timing and duration of the period(s) of absence.
b. To determine:
i. what proportion of absence from work is “approved” and
ii. what proportion is “unapproved”.
c. To compare absence in the Packing Department with:
i. other departments in the company and
ii. overall absentee figures in the industry
d. To ascertain the causes of “unapproved” absence from work.
For the Victoria Gambling problem the research objectives may be:
a. To quantify the effect of casino gaming and gaming machines on racing and Lottery products.
b. To account for these effects, as far as possible, by differences in people’s perceptions of the four forms of gambling.
It should be made clear that objectives need not be exclusively quantitative. Most research studies have purposes other than assigning values to things. Objective d. for Trobe aims only to discover causes of absenteeism (a qualitative activity) – measuring their importance may be beyond the scope of the study.
Qualitative does not mean “vague” however. Unless your objectives are clear ones, they will probably not be conclusively met because you never really said what you wanted to achieve.
“Our first objective is to determine public feeling on the issue of changes in charging methods within the health service.”
i. Which public are they talking about?
ii. What is “feeling” – is it knowledge, attitudes?
iii. Specifically, which changes in what part of the service?
iv. For what purpose?
Examination of the problem area can yield a seemingly infinite set of questions to be answered. However, because of constraints in resources (e.g. time, money) and requirements of possible future users of the study, it becomes essential to identify those questions that can be addressed:
i. within your constraints and
ii. which match the needs of any person or organisation using the study.
If your research has a “practical” application, then you should try to agree with management on questions of problem definition, information requirements and uses of data.
2.4 Prior Work in the area
Previous work is either that of others (literature review) or your own (pilot study).
The process of searching through the literature will help you to identify what others have found to be important variables in the area. It will also provide a foundation for your own understanding of the problem.
At this early stage you are unlikely to have conducted the full literature review, but some indication should be given of your awareness of the area, and which further sources you intend to investigate.
Two main sources of information in this area have been located so far –
i. Review of Credit Legislation.
Victorian Ministry for Fair Trading, May 1995.
ii. Fair Trading in Victoria.
Victorian Law Today, Fair Trading Act, 1999.
No surveys on consumer attitudes in this area have been conducted in Victoria, although we believe some useful British studies exist. These are currently being obtained.
If you have done any prior work in the area, it should be referred to briefly. Usually, because pilot research is conducted under less strict conditions, the actual results of a pilot study are not worth much – it is more useful as an indication of how practical the methods for the current study are. Details of other less formal exploratory research can also be included.
To begin my study of car owners‟ attitudes toward petrol brands and service stations, I interviewed a sample of 10 car drivers at two different petrol stations and worked through my original questionnaire with them. They suggested a number of changes in wording which were incorporated into the revised questionnaire design.
For example, in one question, directed at customers‟ attitudes toward different brands of petrol, several respondents misinterpreted the service station for the brand of petrol it sold.
In another question concerning the „number of children‟ the respondent had been living at home, an insufficient choice range was offered. A greater range has therefore been included in the final questionnaire (see revised questionnaire attached as Appendix 1)…etc.
2.5 General Hypothesis to be Examined (for a Quantitative Study)
Although you may have identified some important variables in a situation and predicated some relationship between them (as shown in the rationale above), you are less likely to have completed other groundwork (such as the literature review).
Therefore you may not be in a position to make specific, detailed statements about all the relationships you believe exist. In the topic analysis it is often better to identify one or a few hypotheses with considerable potential for testing.
For that reason, the criteria for hypotheses are relaxed but nevertheless should-
a. Be free from ambiguity.
b. Express a relationship between variables or groups of variables.
a. The success of casino gaming has occurred, to some extent, because it has presented a significantly different and more attractive image to the gambling market.
b. Lower than expected second quarter sales (units) were due to:
i. Recent re-direction of corporate advertising strategy in Victoria
ii. Some activities of the parent company overseas reflecting badly on the image of the Australian operation.
2.6 Probable Approach
Some areas of the research design will already be worked out, while the clarification of others may depend on you having settled on specific questions like hypotheses. For the topic analysis, it should be possible to offer a brief description of the following areas:
i. What kind of approach (Quantitative or qualitative or mixed methods)?
ii. What kind of design?(Exploratory, descriptive, analytical, predictive etc).
iii. What kind of study?(Field study, experimental study etc).
iv. What kind of subject/unit will be studied?(Individuals, groups, households, organisations etc).
v. From what population will the sample be drawn?(Directory lists, census data, pedestrians etc).
vi. What method will be used to sample the population?(Simple random sample, stratified, convenience, quota sample etc).
vi. By what methods will they be studied?(Interview, observation, mail questionnaire etc).
vii. Where will the study take place?(Home, work, street, cafe etc).
viii. What will be the time horizon?(Immediate, continuous over time, test-retest etc).
ix. What is the value of the study?
a. We will conduct a formalised analytical study into current aspects of individual buyer behaviour. This will involve personally interviewing a sample of male and female shoppers over 15 years of age in all central Melbourne retail footwear outlets. Initially, respondents will be selected by convenience sample, as only people entering shops are questioned. However, once in the outlet, people will be randomly selected. We will question them on their shoe size and style preferences, and their attitudes and opinions about what they like and dislike about shoes that are currently available. The same survey will be conducted again during the media campaign to assess its impact on shoe buyers.
b. Can a causal relationship be established between a car‟s age and its propensity to be involved in an injury accident (per kilometre travelled)? To answer this question, an analytical study of all Traffic Accident Reports collected by the year 1999 will be undertaken.
It is important at this stage to indicate the purpose, size, and expected duration of your research project. For example, a PhD will require a larger sample size as well as a more detailed literature review. It may also require a more robust form of data analysis which will extend well beyond simple descriptive statistics. The latter quantitative analysis method may be perfectly satisfactory for an MBA research project but would probably result in a „fail‟ for a PhD.
2.7 Value of the study
Informing a manager or lecturer about the value of your research is more important at the topic analysis stage than at any other. This is because acceptance or rejection of the entire project is most likely to occur here – later, alterations are usually made to aspects of a previously approved study.
Also, because your analysis may be insubstantial in parts, it is even more important to accentuate the merits of the proposed study and to “sell” it on those merits.
The potential utility of the study will probably lie in one or both of two areas –
i. What it contributes to the existing knowledge of an area, or
ii. What practical applications arise from it.
Where the study is to be used by management, suggestions for its possible use should be made.
Where knowledge of the area is fairly good, some general prediction of the likely outcomes could be made and the importance of each explained.
a. The results of the Reliant Company car viewing survey of 200 respondents will assist management to determine:
i. The likely reception of the new car in the six-cylinder sector of the market
ii. The most likely target group for the car in terms of their:
– demographics, and car buying requirements
– attitudes towards cars and driving
– image perceptions and perceived advantages of the new Reliant car
– level of interest in purchasing the new Reliant car
b. Following the completion of this research project, we will be able to:
i. Describe the demographics and geographics of hirers in general
ii. Ascertain customers’ perceptions of Victor DVD equipment rental
iii. State the factors which encourage or discourage potential hirers
iv. Indicate which customer segments should be targeted as potentially
v. Draw inferences as to the market for DVD rentals amongst non-Victor customers
vi. Determine the marketing success or failure to date prices likely DVD hirers