Developing a Qualitative Research Proposal

tOriginal Editor - Mariam Hashem

Top Contributors - Mariam Hashem, Kim Jackson, Ewa Jaraczewska and Tarina van der Stockt

Introduction

A research proposal is a document that describes the idea of the research, why it's important, and the research method. The format can vary widely among different higher education settings, different funders, and different organizations.

background and context

The title of your research proposal can be different from the publishing title. You can consider it to be more like a working title that you can revisit later after finishing the research proposal and amend it if needed.

The title should have some keywords of what your research encompasses such as:

• The patient population e.g. women who had breast cancer
• Methods e.g. quantitative research, feasibility study or a pilot randomized control trial, a systematic review
• An intervention

Example: experiences of pregnant women using in-depth focus groups

Word count is similar to writing an abstract and can vary from a proposal to another.

Tips:

• Pin down your key points
• Use the filter approach by starting off broadly then refine it down to our research question.

For example, a qualitative study in Ireland by Algeo looking at the facilitators and barriers for women with Breast Cancer in returning to work. The author started by looking at Breast Cancer in terms of the statistics nationally and worldwide. Then moved to issues that women with breast cancer experience, highlighting policies and guidelines. To narrow it down, she made it more specific to work and its importance and returning to work after Breast Cancer. The proposal looked at national and cultural differences in return to work in Breast Cancer and highlighted a gap in an Irish context.

When thinking of the research proposal, it's your tool to sell the research to probably an ethics committee or a research funder so you want to show them why your research is important to be done. Here are some prompting questions to help with writing the background:

• how much do we know about the problem?
• What are the gaps in our knowledge?
• How would new insights contribute to society? Or to clinical practice?
• Why is this research worth doing?
• And who might have an interest in this topic?

Defining your research question: SPIDER tool

Research question(s) give your research a clear focus and guide on the design, research methodology, and data collection. It should be focused and researchable whether through primary or secondary sources. It should be feasible to answer within a given timeframe and it should be specific enough for you to be able to answer thoroughly.

The SPIDER tool is a question format tool and helpful method to structure the research questions. It's similar to the PICO tool which is used for experimental studies:

• P: population
• I: intervention
• C: comparison
• O: outcome.

For qualitative research and mixed methods, we can use the SPIDER Tool:

• S: sample. who is your population? Who do you want to study?
• P and I: the phenomenon of interest. what exactly are we trying to explore? it could be anything like an event, an occurrence, or an object.
• D: the design
• E: the evaluation.
• R: research type.

An example of a research question: what are women with Breast Cancers' experiences of attending a lymphedema clinic?

S = sample is women with Breast Cancer

PI=Phenomenon of interest is the lymphedema clinic. What are their experiences of this phenomenon ( lymphedema clinic)?

D= design which could be a questionnaire, interviews, focus groups, a case study, an observation

E = evaluation, evaluating the experience.

R =research type using a qualitative approach or mixed methods. This something that you need to decide further down the line but will be based on your research question.

After formulating the question(s) you need to look at the way you're going to answer it. Answering the question(s) will depend on the question, the design and research type.

Data collection technique

Each design method has its pros and cons and the selection depends on the question, the participants, and the time-scale. For example, looking at the experiences of somebody who's had severe trauma or exploring a sensitive topic, a one to one interview is probably the most appropriate method to respect privacy.

Interviews:

Interviews are typically one to one between the participant and the researcher. They can be structured, semi-structured, or unstructured. The most common approach is probably semi-structured.

A structured interview is led by the researcher using a set list of questions that all must be answered and there's not a huge amount of flexibility to deviate elsewhere. An unstructured interview is a complete opposite in which the researcher can start off by saying something like ''tell me about your experiences at this clinic'' so it's completely by the participant who can deviate to different ways and the researcher can prompt by ''tell me more, tell me more''. Whereas the semi-structured interview is somewhere in between. There is a list of some questions to give a structure to the interview with flexibility for the participant to deviate or go in-depth into their answers.

Pros & Cons:

• Privacy as it calls for trust between the participant and the researcher with a level of transparency but you do need to be able to build rapport wich needs some practice.
• More appropriate for discussing sensitive topics
• Interviews can be more in-depth than focus groups and you can get a real sense by digging in deep when you're interviewing someone one to one, as opposed to snippets here and there from a focus group.
• Interviews are time-consuming. However, there is less potential for bias, unlike a focus group in which some participants might dominate the dynamics of the group and people can be led by what others are saying and might not speak about something that comes naturally to them so a one-to-one interview cuts out that bias.
• Interviews can be expensive to run than focus groups due to the time required to run them. Focus groups, on the other hand, can vary between five to fifteen participants. Ideally, eight participants is a good number to be able to manage the conversation while obtaining a number of detailed views. Fifteen or more participants can limit the depth of details and the insight provided on the chosen topic but it can also be a handy way to offer a number of different perspectives in a time-efficient manner.

Focus groups can be more appropriate for identifying group norms, patterns, and opinions. There are lower average speaking times, therefore it's not as in-depth as the interviews. The group dynamics need to be managed which needs some experience from the reseacher.

The trustworthiness of the findings can be reduced compared to interviews due to the bias when participants tend to provide a desired answer or to be led by others.

Participant observation: where the researcher gets close enough to participants that they observe with or without participation. Participants can be communities or groups where the research aims to gain familiarity with this specific cohort of people.

For example, observing at a religious group, an occupational group, or a community through intensive involvement in their own environment over an extended period of time that could be with participation, direct or indirect observations.

This method can be time-consuming and requires a conscious effort to be objective and gain trust to build a rapport with the group you're observing. The trust is needed to be built to encourage the participants to feel comfortable and natural to eliminate bias.

Participant recruitment and sample

There is no magic number as to how many people you should recruit in qualitative research. The sample sizes is usually smaller than quantitative research and will depend on a number of variables and some considerations.

When writing a research proposal, it's not good enough to give a vague number instead you need to provide justification and rationale on to how you came about to choose the number of participants that you aim to recruit.

Considerations:

Study design: there are various methods and approaches to collect your data but not all methods will be relevant to the study question(s). For example, interpretive phenomenological approach or analysis (IPA) calls for smaller sample sizes (3-5 participants)

However, there are cases where some participants in IPA creep up to the double (up to 10-12 people). The reason for the smaller sample size with IPA is to allow for the in-depth collection of data and this is different from observing themes or patterns, instead, it's looking at for example, underlying meanings, words, how many things are repeated, and the tone of the language. On the other hand, if we look at thematic analysis it's been suggested to consider between 10-50 participants for participant-generated data by Braun and Clark in 2013. For grounded theory, Moore suggested between thirty and fifty participants, whereas Creswell suggested twenty to thirty.

The principles of data saturation. While our sample should be sufficient enough to describe the phenomenon of interest in detail, having a very large sample is a risk of repeating obtained data to the point of limiting the emerging themes, information, and ideas. This is referred to as data saturation. There is a need to balance the sample size to the amount of information we aim to get from the study.

Quality over quantity. Recruiting the right participants by setting up clear criteria who meet the right criteria. On some occasions, researchers might not be able to recruit enough participants or reach the planned sample size. However, this is better than recruiting people who don't necessarily match those criteria as this might compromise the results.

Setting a research timeline: Gantt chart

Other considerations such as costings and budgeting