Dissertation Support Guide

The purpose of the research strategy is to guide the researcher through the steps needed to answer the research question.

The main research strategies identified by Saunders et al. (2012) are:

 Action Research

The findings of action research are used to actively instigate change within an organisation (e.g. The researcher investigates the effectiveness of a process used in their organisation to assess how it can be improved).

The researcher has to be involved in a particular role in the organisation (or be in charge of a project), and has to be able to directly implement changes. Both qualitative and quantitative methods of data collection can be used.

You should use this method if you are in the position to directly implement change in your organisation on the basis of the findings of your research.

Usually associated with: Both qualitative and quantitative research.

 Experimental Research

The researcher conducts a number of experiments where only one variable is manipulated at a time. (e.g. To test perception of time, two control groups may be placed in two identical rooms of the same amount of time. The first test group is subjected to painful stimuli while the second test group is subjected to pleasurable ones. At the end of the experiment members of the two groups will be asked to assess how much time they spent in the room).

This method is difficult to implement on a tight timeline as the experiment has to be repeated multiple times; it may also be difficult finding test subjects.

Usually associated with: Qualitative research.

 Case Study

Case study research is the assessment of an entity (usually an organisation, individual, event, or action), in order to establish its key features (Bryman, 2012).

You should conduct a case study if you have a personal interest or an in-depth knowledge of the case (for example, if you work in a company, you may decide to chose it as the subject of your case study).

Usually associated with: Both qualitative and quantitative research.

 Survey Research

Surveys are commonly used in quantitative research projects, and involve sampling a representative proportion of the population (Bryman & Bell, 2015). Surveys produce a large amount of data that is then analysed to identify causal relationships (cause and effect).

Survey research examines the impact of a particular independent variable (the cause) on the dependent variable of interest (the effect).

Example:

You wish to investigate the impact of education on earnings. In this instance, the level of education is considered the independent variable (cause) and the amount of income is the dependent variable (effect). Likewise, if you wish to investigate the impact of length of services on earnings, length of service is the independent variable and the amount of income is the dependent variable.

Usually associated with: Quantitative research. Uses the deductive approach (you have to develop hypotheses to test).

 Grounded Theory

Grounded theory involves the collection and analysis of qualitative data. Theories are developed only after you have collected and analysed the data.

When performing research based on grounded theories you should:

• Identify the main area of interest.
• Avoid preconceptions (no hypotheses) and focus on the data findings only.
• Interpret the subtle meaning of the data and analyse, not only the answers, but also how the questions have been answered. Were answers delivered hesitantly or with conviction? Were the participants of a focus group agitated or calm when expressing their point of view?

Usually associated with: Qualitative research. This method uses the inductive approach.

 Ethnography

Ethnography analyses groups in a specific setting to assess how individuals relate to each other and their environment. (Bryman, 2012).

A subset of ethnography, called digital ethnography, is becoming increasingly popular to conduct market research online, for example, through the analysis social media analytics, blogs and online forums.

Usually associated with: Qualitative research.

 Archival Research

This type of strategy involved a systematic review of archival material in order to identify patterns and establish the sum of knowledge on a specific topic.

Archival research is based on the analysis of raw data that is unprocessed, such as records, market reports and statistics.

The researcher collects and analyses the data to give it new meaning.

This type of research is considered quantitative if the raw data is numerical in nature, whereas it is considered qualitative if the data is textual.

Usually associated with: Both qualitative and quantitative research.

References

Bryman, A. (2012). Social research methods. 5th edn. Oxford: Oxford University Press.

Saunders, M., Lewis, P. and Thornhill, A. (2012) Research methods for business students. 6th edn. Harlow: Pearson Education.

When discussing your methodology, you have to clarify if you will approach the research problem inductively or deductively.

 Inductive approach

The inductive approach is used to generate new ideas, and is associated with qualitative research.

Researchers gather a better understanding of the problem as they progress through the research. Researchers will NOT develop hypotheses* at the initial stages of the research as they are not sure about the nature of the findings until the research is completed.

This approach can be divided in 3 stages:

1Data collection and analysis

2Identification of patterns

3Development of a theory

 Deductive approach

The deductive approach is used to test theories, and is associated with quantitative research. If you chose to use the deductive approach you should develop hypotheses* to test.

Researchers develop hypotheses based on existing theories and design the research strategy to test the validity of the hypotheses. This approach is usually associated with quantitative research.

4 Stages:

1.Developing a theory based on existing research

2.Formulating hypotheses to test

3.Data collection and analysis

4.Hypotheses are either confirmed or rejected.

*Hypotheses are statements, developed on the basis of the findings of existing research, that clarify what you expect to find in your research.

The research population consist of a large group of individuals who share similar characteristics that are relevant to your study. Due to the fact that many populations are too large to address directly, it is necessary to adopt sampling techniques to identify a smaller group to represent the larger population.

To begin, you should determine your sample size, or the number of individuals who will take part in the primary data collection process. When determining the size of a sample you need to keep in mind that:

• A larger sample size will be more representative of your population than a smaller one.
• Quantitative research requires a greater sample size than qualitative research.
• If you email a questionnaire, you need to start with a large sample size, to account for a low percentage of responses.
The sample should not be either too large or too small. If it is too large, it will take a long time to analyse the data; if it is too small, it will not be representative of the population at large.

Sampling methods are commonly divided two categories: probability and non- probability (Saunders et al., 2012, p.211).

When using probability sampling, each member of the population has equal chances to be chosen to take part in the study; this type of sampling is mostly used when dealing with very large populations.

Probability sampling techniques include: simple random sampling, systematic sampling, stratified random sampling, cluster sampling.

 Simple Random Sampling

This is a completely random method of selection.

This technique may involve assigning numbers to all the research subjects and then using a random number generator to chose random numbers until the sample size is reached.

 Systematic Sampling

In systematic sampling, the researcher selects samples at regular intervals.

The researcher starts by assigning a number to each member of the sampling frame (a sampling frame is a list used to define a researcher's population of interest) starting from 0.

The first case is selected using a random number; then, the researcher has to calculate the sampling fraction (proportion of the sampling population that the researcher needs to select).

Example

If the sample size is 1/3 of the population every third person after the original random number will be selected.

If the original random number is 5 and the interval is 3, the sequence will look as follows:

5;8;11;14;17;20; 23 etc…

 Stratified Random Sampling

This sampling technique involves dividing subjects into mutually exclusive groups (ex. males and females) and then using simple random sampling to choose members from these groups.

The advantage is that it is more likely to be representative of the population; the disadvantage is that it is only possible if you can easily identify the distinguishing parameters of each group.

For example, if one of the groups has to be comprised of part-time staff and the other of full-time staff, you need to ensure that you know the employment status of all your subjects, in order to assign them to the correct group.

When using no-probability sampling, the selection is not random, therefore, only some members of the population will have an opportunity to take part in the study.

Non probability sampling techniques include: quota, purposive, snowball, convenience, self-selected.

Quota Sampling

This is a non-random method used for surveying large populations.

The researcher divides the population into subgroups and tries to identify what portion of the population each subgroup represent. The same proportion should be then applied in the sampling process. The researcher then selects subjects from the various subgroups while taking into consideration the proportions noted in the previous step; this will ensure that the sample is representative of the entire population.

This is the ideal technique to use if the aim of your study is to investigate a particular characteristic of a certain subgroup.

Purposive Sampling

Purposive sampling allows the researcher to select which individuals are more fit to provide the information needed to answer the research question.

For example, if the purpose of the research is to investigate the effectiveness of staff retention techniques within a company, the research sample should be comprised of members of the company’s Human Resources department.

This sampling method is often used in case studies.

Snowball Sampling

Snowball sampling is normally used if the population size is very small. The researcher asks the initial subjects of the research to identify other potential subjects that meet the research criteria.

The main risk associated with using this technique is bias* because the initial respondents are likely to identify potential respondents who have similar points of view.

*Bias is a lack of objectivity; either favouring something or being prejudiced against it.

Convenience Sampling

Convenience sampling is one of the most commonly used sampling techniques.

The samples are selected on the basis of accessibility. The sample selection process is continued until the researcher reaches the required sample size. This technique is the least time consuming but it is not always representative of the entire research population.

For example, if the focus of your research is on social media behaviour in Ireland, and you survey mainly people belonging to your age group(friends, classmates, etc), your sample will not be representative of the research population which is comprised by individuals of different age groups.

Self-Selection

In self-selection sampling, individuals are not directly approached by the researcher and choose to take part in a study on their own accord.

For example, the researcher may share the link to their questionnaire on a social media platform or advertise the need for participants to a focus group.

This technique ensures that the sample is comprised by people who have an interest in the topic but the downside is that may be difficult to find volunteers.

How can you choose the appropriate research design to answer your research question?

The answer to this question is complex, but you can simplify it by asking yourself what kind of data you need to collect.

 Qualitative research design:

Qualitative data is text based, and it does not employ mathematical techniques. This type of design is associated with the analysis of individuals’ views of the world, which are non-quantifiable.

 Quantitative research design:

Quantitative data is numerical in nature, and it is used to test variables in your population responses.

 Mixed method design:

At times, you may need to collect both qualitative and quantitative data in order to answer your question,in which case you will have to use a mixed method design.

Qualitative research is used to gather an in-depth understanding of a problem and answer questions such as: “why?” and “how?”.

Commonly used qualitative data collection methods are interviews, focus groups (and mini focus groups), and open-ended questionnaires.

 Interviews

One-on-one in-depth discussions, they can be:

• Structured interviews: The researcher asks the same questions to all participants in a predefined order without deviation.
• Unstructured interviews: Flexible format where questions can change depending on the participant's responses. The interviewer can diverge from the interview schedule.
• Semi-Structured interviews: A set of core questions is used, but interviewers may explore responses in more detail. It strikes a balance between structure and flexibility.

Tips for Interviews

• Style: Ask open-ended questions to get descriptive answers.
• Bias: Avoid leading questions (e.g., "Do you agree that...").
• Language: Use simple terms. If using technical language, explain it clearly.
• Be concise: Keep questions short and specific.

 Focus Groups

A focus group is a small (usually 7+ people) but diverse group representing broader opinions. Groups of 4–6 are referred to as MINI FOCUS GROUPS.

Recommended for exploring views dependent on group interaction. They can be:

• Structured: The moderator follows a set agenda without deviations.
• Unstructured: Free-flowing discussion, often used in product or general opinion research.

Tips for Focus Groups

As a moderator, you must guide the group while maintaining a relaxed atmosphere. Keep the session under 2 hours.

Before the session:

• Design an interview schedule with time allocated for each question.
• Organise a suitable location.
• Ensure all participants can attend.
• Establish rules at the start.
• Obtain explicit consent for data use.
• Stick to your schedule to cover all key topics.
• Intervene if someone dominates the conversation.
• Request permission to record and take detailed notes.
• Consider tone and non-verbal cues as part of your analysis.

 Questionnaires with Open-Ended Questions

Unlike multiple-choice formats, open-ended questions require full responses, not just “yes” or “no.”

Examples:

• What steps is your company taking to improve performance management?
• What type of marketing strategy is your company currently employing?
• Do you feel comfortable in your work environment?
• Have you ever felt you were being discriminated against in your workplace?

These questions allow deep exploration but have drawbacks:

• Lower completion rates due to time demands.
• Risk of misinterpretation or vague answers.
• Time-consuming analysis, especially with large datasets that require manual coding.

Tip: Open-ended questions can also be added at the end of multiple-choice surveys to gather detailed insights.

Quantitative research is associated with statistics and calculations and is used to answer questions such as: “what?”, “which one?”, “how often?”, “how many?”.

Quantitative research uses variables (age, gender, culture, role in the workplace, etc.) and examines their impact on the responses collected. This type of research is appropriate when the researcher wishes to test assumptions deductively and it supports the development of hypotheses.

The most common quantitative data collection method is the multiple-choice questionnaire.

Questionnaire Design Tips

• The questions included in your questionnaires should always be based on your research objectives and hypotheses.
• When designing a questionnaire remember to include relevant variables to be tested. The variables have to be mutually exclusive, meaning that only one option can be selected (ex. male/female; full time/part-time; under 30/30-50/over 50). Using variables will allow you to divide the responses by group. For example, If the focus of your research is on gender-based marketing, you need to know if your respondents are men or women;
• whereas, if you are investigating generational marketing, you need to know which age groups your respondents belong to.
• You can use one or more types of questions to design your questionnaire:
• Multiple choice questions: Respondents are asked to choose from a distinct set of pre-defined responses.
• Rating scale: Respondents are asked to assign a numeric value to the rated object. Ex. Rate how much they enjoy a product on a scale 1 to 10.
• Rank order questions: Respondents are asked to compare a list of different objects to one another and rank them. If the list has 10 objects, the object rated 1 is the most important and the object rated 10 is the least important.
• Checkboxes: Respondents can select one or more of the pre-defined responses.
• Likert scale: Respondents are asked to specify their level of agreement or disagreement for a series of statements.
• Open-ended questions: Respondents are not given options to choose from and invited to express their opinion. Open-ended questions are mostly used to ask a respondent to clarify why they gave a certain answer.

Mixed method research requires the collection and analysis of both qualitative and quantitative data.

Mixed methods designs are useful to gather a more comprehensive understanding of a problem, as they allow data triangulation.

Triangulation entails comparing the qualitative and quantitative findings of your primary research with the existing literature.

The principle behind triangulation is that, getting the same result when using different methods, increases confidence in the findings.

Creswell and Plano Clarke (2018, pp.65-68) identify 3 core mixed methods designs:

 Convergent Design

The convergent design is used when the researcher wants to combine the results of both qualitative and quantitative data analysis.

• Data collection: Both qualitative and quantitative data are collected at the same time and have equal value for addressing the research question.

The researcher has to compare the qualitative and quantitative data and identify if the findings are convergent (the two datasets show similar finding) or divergent (different/conflicting findings).

If the results are divergent, the data has to be analysed more in depth to identify the reason why.

• Strengths: Efficient, because both datasets can be collected at the same time. Allows the comparison between the participants’ point of view on an issue, gathered by asking open-ended questions, and the researcher’s point of view, which influences the response options provided on the questionnaire.
• Challenges:It is difficult to merge meaningfully numerical and text based data. It may be necessary to explain contradictions in the results.

 Sequential Explanatory Design

The purpose of this design is to use qualitative data to interpret quantitative results.

Data collection: QUANTITATIVE → QUALITATIVE

Two- phase process: To begin, the quantitative data is collected and analysed, then, a follow up interview is set up with selected individuals from the same sample group to examine more in depth in their answers.

This method is useful when:

• The research problem is mainly quantitative in nature, therefore, it makes more sense to start with the quantitative data collection.
• The researcher already knows the variables of the problem.
• The survey respondents are willing to be contacted for follow-up interviews.
• The researcher has time to conduct research in two phases.

 Sequential Exploratory Design

The purpose of this design is to use qualitative data to identify variables that will then be tested quantitatively. The advantage of first conducting qualitative research is that the variables generated from this exercise will be specific to the culture or setting that you are analysing, instead of being generic.

Data collection: QUALITATIVE → QUANTITATIVE

Three-phase process:

1.Data Collection.

2.Data analysis to identify variables.

3.Testing variables.

This method is useful when:

• The research problem is qualitative in nature.

OR

• The available literature on the research problem was conducted in other Countries/ environments. In this case, it may be necessary to develop variables that reflect the culture or setting that is being investigated. The researcher wants to assess if the findings of the qualitative data can be generalised (how many people share the same opinions?).
• The researcher has time to conduct research in 3 phases: qualitative, development, and quantitative.

References

Creswell, J.W. and Plano Clarke, V.L. (2018) Designing and conducting mixed methods research. 3rd edn. London: Sage Publications.