Case study protocol planning makes a research project easier to conduct and easier to defend. Before arranging interviews, you need a clear account of what the case is, which evidence can answer the research question and how you will handle information that challenges your explanation. A protocol puts those decisions in one working document.

This guide develops an illustrative study of employee onboarding in a regional logistics company. The organisation and all examples are fictional. The purpose is to show how nine protocol sections connect, rather than provide a form to copy without methodological thought.

Contents  What a case study protocol should do  •  1 Define the purpose and question  •  2 Identify the case and its boundaries  •  3 Explain case selection and access  •  4 Specify participants and recruitment  •  5 Link questions to evidence sources  •  6 Prepare fieldwork procedures  •  7 Set out the analysis approach  •  8 Address ethics and data management  •  9 Define quality checks and changes  •  Test the protocol before using it

What a case study protocol should do

A research proposal explains why a study deserves to be undertaken. A protocol explains how you will conduct it consistently. Your proposal may describe an interest in onboarding; your protocol specifies which onboarding programme, which period, which employees and which records fall within the investigation.

Baxter and Jack’s case study methodology paper provides a useful foundation for defining and bounding a case. The practical sections below turn those design decisions into instructions for fieldwork. They are suggestions for an applied organisational study, not a compulsory structure for every qualitative tradition.

If you are still deciding whether the method fits your question, first read The Case HQ’s guide to the role of case studies in academic research. A protocol cannot rescue a research strategy that does not fit the question.

1 Define the purpose and question

Start with one sentence describing what you want to understand. In the illustrative study, the purpose is to understand how employees experience a newly introduced onboarding programme and why experiences differ between two operating sites.

The main question becomes: how does the company’s onboarding programme help or hinder new employees in becoming able to perform their roles during their first three months?

This wording directs attention towards experience, processes and context. It does not promise to establish that onboarding causes a particular productivity improvement. A causal effectiveness claim would require an appropriate design and stronger treatment of alternative explanations.

Record two supporting questions: which parts of the programme employees find useful, and how supervisors influence its implementation. These guide data collection without creating an unmanageable list of separate studies.

2 Identify the case and its boundaries

The case is the onboarding programme within one company, implemented across two sites. The employees are participants who provide evidence about that programme. They are not automatically separate cases.

Set a time boundary, such as employees who joined between January and June in the chosen study year. Specify that the investigation covers the first three months of employment. Exclude recruitment decisions and later promotion processes unless they become relevant contextual evidence.

Write down the geographical and organisational boundaries as well. Agency workers may experience a different induction process; decide whether they are included and explain why. Scope decisions made before fieldwork are easier to defend than exclusions introduced because some accounts are inconvenient.

3 Explain case selection and access

State why this organisation offers a useful setting. Perhaps it recently standardised onboarding while allowing supervisors discretion over local delivery. That combination creates an opportunity to examine the relationship between a common programme and different implementation conditions.

Access is a separate matter. Permission from a company director does not mean every employee has freely agreed to participate. Describe who will circulate invitations, how volunteers will contact the researcher and how managers will be prevented from seeing individual participation decisions where feasible.

Also identify access dependencies. If training records cannot be released, what part of the question remains answerable through interviews and observations? Recording this early helps you distinguish a practical adaptation from a fundamental change in design.

4 Specify participants and recruitment

Use eligibility criteria connected to the question. New employees should have experienced enough onboarding to discuss it; supervisors should have delivered or overseen the programme. Human resources staff can explain intended design, but their account cannot substitute for employee experience.

A provisional recruitment plan might seek variation across sites, job families and onboarding experiences. Avoid claiming that a small purposive sample statistically represents the workforce. Explain the kinds of experience you need to understand and review gaps as recruitment develops.

Describe a starting range only if it serves planning, and explain how analytical adequacy will be assessed. A fixed interview count is not, by itself, proof of saturation or sufficient understanding. Keep a recruitment log recording eligibility, invitation routes, participation and relevant gaps without exposing identities unnecessarily.

5 Link questions to evidence sources

For each supporting question, identify evidence that could answer it. Employee interviews reveal how support was experienced. Supervisor interviews explain local delivery choices. Training schedules describe planned activities. Observations may show how introductions and task demonstrations actually happen.

A compact evidence map could read as follows:

Research issue	Main evidence	What it cannot establish alone
Employees’ experience of support	Individual interviews	Whether every employee shares that experience
Consistency of delivery	Schedules and supervisor accounts	Whether scheduled activities occurred as intended
Learning during real work	Observations and employee examples	Long-term organisational effectiveness

The point is complementarity. Multiple sources do not automatically confirm each other, and agreement between documents written by the same department may provide less independence than it appears.

6 Prepare fieldwork procedures

Describe how each interview will begin, how consent will be confirmed and how recordings will be transferred into secure storage. Use a common set of core questions while allowing follow-up questions appropriate to each person’s experience.

A suitable opening is: describe a recent moment when you needed help to complete a task during your first weeks. Follow with questions about what happened, who was involved and what difference the help made. This invites a concrete account rather than a general satisfaction rating.

Pilot the guide before full collection. If participants interpret onboarding as payroll registration while you mean workplace learning, revise your opening explanation. Keep a dated change log so later readers can see which version was used.

The UK Data Service’s tools and templates can support the organisation of research records. Adapt relevant resources to your institution’s procedures rather than treating an online template as ethics approval.

7 Set out the analysis approach

Name the analytical approach and explain why it fits. For this example, an applied thematic analysis might examine forms of support, barriers to access and differences between intended and experienced delivery. Specify whether initial concepts guide coding and how unexpected issues will be retained.

Begin with familiarisation, then code relevant material while keeping sufficient context around each extract. Develop provisional interpretations and revisit the underlying accounts. Compare sites only after examining the circumstances within each site; otherwise superficial differences can become premature explanations.

Keep analytical notes that distinguish what a participant said from your interpretation. If one employee reports that frequent supervisor checks felt supportive while another describes similar checks as intrusive, investigate the circumstances instead of forcing both accounts into a single positive category.

8 Address ethics and data management

Explain the risks particular to this setting. Newly recruited employees may worry that criticism affects probation decisions. Removing names may not be enough when a quotation identifies a rare role, shift or incident.

Plan how quotations will be reviewed for indirect identification, who will access the recordings and how the identity key will be protected separately. Set retention and deletion arrangements that follow applicable institutional requirements and participant information.

The UK Data Service guidance on anonymising text explains why contextual details need careful treatment. Preserve analytical meaning while reducing identification risk; indiscriminately removing context can make the findings difficult to interpret.

If AI transcription or analysis is proposed, assess the actual service, data handling and permission requirements before uploading recordings. Institutional approval for one system does not automatically cover another provider or processing location.

9 Define quality checks and changes

Choose checks that fit your methodological position. For this organisational study, you might maintain source identifiers, examine contradictory accounts and document why themes change. An independent colleague could challenge whether claims exceed the available evidence.

Do not add every technique labelled rigour to the protocol. Intercoder agreement, participant feedback and reflexive interpretation serve different purposes and are not universally interchangeable. Explain what each selected check contributes.

Record changes to boundaries, recruitment or interview questions with the date, reason and consequence. If fieldwork reveals that temporary workers receive an entirely separate programme, decide whether including them changes the case. Make that decision visible rather than quietly expanding the study.

Test the protocol before using it

Give the document to a colleague and ask them to explain who would be interviewed, which records would be requested and how the evidence would answer the question. Confusion reveals decisions that remain implicit.

Then rehearse one evidence pathway: take a proposed claim about supervisor availability and identify the interviews, observations or documents needed to assess it. If the evidence would only show that a policy exists, narrow the claim or strengthen the collection plan.

For further support with the broader writing process, see The Case HQ’s guide to writing case studies for research. Researchers seeking structured instruction can explore Designing and Conducting Case Study Research, while educators teaching these decisions can review the paid Certified Research Methodology Instructor programme.

Your finished protocol should be detailed enough to guide action and concise enough to use during fieldwork. Begin with the boundaries and evidence map, test them against the research question, and develop the remaining procedures around those decisions.