Once you have a clear research question and rationale, the next step is translating the question into a feasible study design and then making sure that you can get the appropriate funding, resources, and approvals in place to run a well-executed project. There is nothing more frustrating than reaching the end of a research study only to realise that a key variable was not measured, an important confounder was overlooked, the study was too underpowered to answer the question you set out to investigate, or you were way over budget. Careful planning at this stage helps avoid these pitfalls by forcing you to think through how the question will be operationalised, how data will be collected and analysed, and what practical constraints might influence the quality or interpretability of the results.
This section will guide you through the six key steps in planning a well-designed study:
The following guidelines for Step 1 apply primarily to designing epidemiological studies. Other study types including qualitative studies, simulation modelling, economic analyses, systematic reviews, and meta-analysis require slightly different approaches.
To choose an appropriate study design framework, you need to consider four key factors: what you are trying to achieve, how common the outcome or disease is, how common the exposure is, and whether the proposed intervention is ethical and practically feasible.
Before data collection begins, clearly define all variables that will be measured or analysed. This includes outcomes, exposures, predictors, potential confounders, and effect modifiers. Precise variable definition is essential for ensuring consistent data collection, appropriate analysis, and meaningful interpretation of results.
Outcomes
Start by defining the primary outcome(s) of interest. Specify exactly what is being measured, how it will be measured, and over what timeframe. Where relevant, provide explicit diagnostic or classification criteria rather than relying on informal or subjective definitions. For example, an outcome might be defined as veterinary-confirmed clinical mastitis occurring in dairy cows within the first 30 days of lactation, recorded in farm treatment records, or the presence of gastrointestinal parasite infection in dogs based on a positive faecal flotation test during the study period.
Exposures or Predictors
Next, define the exposure(s) or predictors you are interested in evaluating. Exposures are the factors hypothesised to influence the outcome, and should be clearly operationalised. Examples include the use of selective versus blanket dry cow therapy in the previous lactation, or the frequency of anthelmintic treatment in dogs over the preceding 12 months. Predictors may also include non-causal variables used to explain variation in outcomes, such as body condition score at calving as a predictor of reproductive performance, or the age of the dog at the time of sampling as a predictor of parasite positivity.
Confounders
Identify potential confounders a priori. Confounders are variables associated with both the exposure and the outcome that can distort observed relationships if not accounted for. For example, parity in dairy cows may be associated with both the choice of dry cow therapy and the risk of mastitis, while housing environment (such as urban versus rural settings) may be associated with both anthelmintic use and parasite exposure in dogs.
Effect Modifiers
Consider whether any variables may act as effect modifiers. Effect modifiers alter the strength or direction of the association between exposure and outcome across different subgroups. For example, season may modify the relationship between pasture exposure and parasite burden in grazing livestock, or immune status may modify the effect of vaccination on disease incidence in companion animals. Identifying potential effect modifiers in advance allows for planned stratified analyses or interaction terms.
For all variables, specify units of measurement, categorisation rules, thresholds, and how missing or ambiguous data will be handled. Decisions about coding, transformation, and grouping should be made before data collection wherever possible, rather than retrospectively during analysis.
In some studies, particularly exploratory or hypothesis-generating work, the aim may be to identify patterns, generate new questions, or map practices rather than to test a specific causal hypothesis. In these cases, variable selection may be broader and more flexible, but variables should still be clearly defined and measured consistently. Being explicit about the exploratory nature of the study helps frame interpretation appropriately and positions the findings as a foundation for future, more targeted research.
The choice of data sources and measurement methods should be driven by the study design and the type of question being asked. Different study types place different demands on data quality, consistency, and feasibility, and these should be considered explicitly during the planning stage.
For survey-based studies, data collection typically relies on questionnaires administered online, by mail, by phone, or in person. Key considerations include question wording, response options, survey length, and mode of delivery. Surveys should be drafted with the target population in mind, using appropriate language and avoiding technical terms where possible. Draft questionnaires should be aligned with the analysis plan to ensure that all variables required for analysis are captured. Pilot testing is particularly important for surveys to identify unclear wording, respondent burden, and patterns of missing data.
For observational studies using existing records or administrative data, measurement involves data extraction rather than direct data collection. This requires the development of data extraction templates and clear definitions for each variable. Researchers should assess data completeness, consistency, and accuracy early, as well as the extent of missing or misclassified data. Pilot extractions from a small subset of records can help identify data quality issues and refine extraction protocols.
For field-based observational studies, data may be collected through direct observation, clinical examination, or sample collection. Standardised observation protocols, data collection forms, and measurement procedures are essential to ensure consistency across observers, sites, and time points. Training and calibration of data collectors should be planned where multiple people are involved.
For laboratory-based studies, measurement is governed by laboratory protocols and assay methods. Researchers should specify the tests to be used, sample handling and storage procedures, quality control measures, and any thresholds for classification or interpretation. Pilot testing or validation of laboratory methods may be required, particularly when using new assays or working in unfamiliar settings.
For interventional or clinical trials, data sources often combine several of the above approaches. Measurement plans should clearly distinguish baseline, intervention, and follow-up data, specify timing and frequency of measurements, and outline procedures for monitoring compliance and adverse events.
For qualitative studies involving interviews or focus groups, data sources include audio recordings and transcripts. Measurement in this context refers to the use of interview guides rather than fixed instruments. Interview guides should be structured enough to address the research question while allowing flexibility to explore emerging themes. Pilot interviews help refine question flow, identify sensitive topics, and assess the time required for data collection and transcription.
Develop an analysis plan before data collection begins, outlining the statistical, qualitative, or mixed-methods approaches that will be used and the software required. The primary purpose of the analysis plan at this stage is to ensure that the data being collected are appropriate in content, structure, and format to support the intended analyses.
Specify how each research question will be addressed analytically, including the main comparisons, models, or themes of interest. For quantitative studies, this may involve identifying outcome measures, exposure variables, confounders, and planned regression or descriptive analyses. For qualitative studies, this may include the intended coding approach, thematic framework, or analytic lens. Mixed-methods studies should clearly describe how quantitative and qualitative components will be integrated.
Critically, use the analysis plan as a check on data collection. Confirm that variables are being captured at the correct level (for example continuous versus categorical), with appropriate response options, units of measurement, and timing. Data collected in the wrong format, or without sufficient detail, cannot be easily fixed during analysis and may prevent key questions from being answered.
Clearly specify who will be included in the study and how participants will be identified, approached, and recruited. Recruitment is one of the most common points of failure in research projects, so this section should be treated as a practical feasibility exercise rather than a theoretical one.
Eligibility and sampling
Start by defining inclusion and exclusion criteria in operational terms that can be applied consistently by anyone involved in recruitment. Avoid criteria that require subjective judgement or information that is difficult to verify at the recruitment stage. For example, specifying “dogs aged 6 months to 5 years attending participating clinics” is more feasible than criteria requiring detailed historical information that may not be available at first contact.
Determine the sample size or sampling strategy and assess feasibility realistically. For quantitative studies, confirm that the required sample size can be achieved within the recruitment window, given expected response rates and attrition. For example, if an online survey is expected to achieve a 10–20% response rate, recruitment efforts must reach five to ten times the required sample size. For qualitative studies, define a target range (for example 15–25 interviews) and assess whether access to participants is sufficient to reach thematic saturation.
Setting, locations, and timing
Specify exactly where recruitment and data collection will occur and over what period. This may include veterinary clinics, farms, shelters, laboratories, or online platforms. Recruitment timing should account for seasonal effects, clinic or farm workload, school holidays, and industry cycles, all of which can substantially affect participation. Build in longer recruitment periods than you think you need, particularly for voluntary studies.
Incentives and participant support
Incentives should be chosen to match the burden of participation and the characteristics of the target population. Practical examples include:
Incentives should compensate for time and inconvenience rather than act as coercion, and must be clearly described and approved through ethics review. In many studies, even small incentives substantially improve recruitment and follow-up completion.
Recruitment channels
Choose recruitment channels based on where the target population already engages, rather than where it is easiest to advertise. Common effective approaches include:
Passive approaches alone (for example a single social media post or flyer) rarely generate sufficient recruitment. Plan for repeated contact and follow-up.
Project advertisements
Recruitment advertisements and participant information should be drafted early and tested informally before submission for ethics approval. Effective recruitment materials are concise, visually clear, and written in plain language.
They should clearly state:
Avoid overstating benefits or using technical language. If the study involves multiple steps or follow-up, this should be made explicit to reduce dropout later.
Check out the STROBE guidelines for more detailed information on best practice for reporting different research study designs.
Effective study design includes anticipating what could go wrong and planning in advance how those risks will be managed. Identifying risks early allows mitigation strategies to be built into the study design, timeline, and budget, rather than reacting once problems arise.
Recruitment risks
One of the most common risks in research projects is failing to recruit sufficient participants within the available timeframe. This may occur due to overestimated response rates, gatekeeper fatigue (for example clinic staff or farm managers), competing demands on participants’ time, or poor alignment between recruitment channels and the target population.
Mitigation strategies include:
Access and permissions risks
Delays or failure to obtain access to sites, data, animals, or records can halt a study even when ethics approval is in place. This is common in studies relying on third-party data, commercial operators, or multiple organisations.
Mitigation strategies include:
Seasonal and timing constraints
Seasonality, production cycles, breeding seasons, or weather conditions can limit when data can be collected. Academic calendars, staff availability, and ethics review cycles can also affect timelines.
Mitigation strategies include:
Data quality and completeness risks
Poor data quality, inconsistent measurement, or high levels of missing data can render a dataset unusable. This risk is particularly high in multi-site studies or studies involving self-reported data.
Mitigation strategies include:
Equipment and technical failures
Failure of equipment, loss of samples, software issues, or data corruption can disrupt or invalidate data collection.
Mitigation strategies include:
Ethics and regulatory delays
Ethics approvals, amendments, or additional regulatory requirements can take longer than anticipated and delay project start dates.
Mitigation strategies include:
Budget and resourcing risks
Underestimated costs, staff turnover, or unanticipated expenses can compromise study delivery.
Mitigation strategies include:
Explicitly documenting risks and mitigation strategies strengthens study feasibility, improves funding applications, and provides a practical framework for managing uncertainty once the project is underway.
Once the study design is clear, the next step is to identify who and what will be required to deliver the project. This step focuses on specifying personnel, expertise, infrastructure, and materials, rather than recruiting or securing them. The specific resources you need will vary greatly depending on the type of research you are conducting.
Simple surveys and questionnaires
Studies based on surveys or questionnaires are typically lower cost but still require careful planning. Common requirements include:
Observational studies using existing data
Studies that analyse clinical records, production data, or administrative datasets often appear straightforward but still require significant resourcing. Common requirements include:
Field-based observational studies
Studies involving direct observation or sample collection from animals in the field require additional logistical planning. Common requirements include:
Laboratory-based studies
Laboratory studies require careful consideration of facilities, safety, and staffing. Common requirements include:
Intervention studies and clinical trials
Interventional studies are among the most resource-intensive. Common requirements include:
Qualitative studies and interviews
Studies involving interviews, focus groups, or other qualitative methods have distinct resourcing needs. Common requirements include:
Mixed-methods studies
Mixed-methods projects combine two or more of the study types above and require careful coordination. Common requirements include:
Once you know what resources are required to run the project, you can start translating the study design into a realistic budget. The aim at this stage is to determine what is needed to deliver the study from start to finish, and to justify those costs clearly.
Developing a detailed and defensible budget is critical, both for internal planning and for funding applications. Reviewers use the budget to assess feasibility, realism, and whether the scope of work is appropriately matched to the requested funding. In the New Zealand context, under-budgeting is one of the most common reasons otherwise strong proposals are viewed as high risk.
Budgets should be broken down into clear categories, with a brief explanation of why each cost is necessary.
Personnel costs
Personnel time is often the largest component of a research budget and the area most scrutinised by reviewers.
Typical inclusions are:
Indicative New Zealand cost ranges (very approximate):
A common budgeting mistake is underestimating the total hours required for data cleaning, management, and analysis, particularly for observational or multi-site studies.
Travel and fieldwork costs
Travel and fieldwork costs are often underestimated in New Zealand due to geographic spread and rural access.
These may include:
Fieldwork costs can escalate quickly for rural, farm-based, or wildlife studies and should include a buffer for weather delays, repeated visits, or failed recruitment attempts.
Equipment and consumables
Equipment and consumables vary widely by study type.
Common inclusions are:
Indicative costs:
Software and data management
Many projects require paid software or platform access, even if analysis itself is conducted in free environments such as R.
Typical costs include:
Funders increasingly expect clear data management planning, and these costs are rarely questioned when justified.
Participant costs and incentives
Where appropriate, budgets should include costs associated with participant involvement.
These may include:
Indicative amounts often range from $20–$50 per participant for low-burden studies, increasing for longer interviews, repeated visits, or invasive procedures. Even modest incentives can substantially improve recruitment and retention and are often viewed positively by ethics committees.
Transcription, laboratory, and external services
Many projects rely on outsourced services.
Common examples include:
These costs should be estimated carefully, as reviewers often compare them against expected sample sizes and timelines.
Contingency costs
Including a contingency line is good practice, particularly for field-based or externally dependent projects. A contingency of around 5–10% of total direct costs is common, although some funders restrict or prohibit explicit contingency lines. Where this is the case, risk should be managed by conservative cost estimates elsewhere in the budget.
If your project is administered through a university, overheads (also called indirect costs) will usually apply. These costs contribute to institutional services that enable research to occur, including administration, finance and contracting, IT systems, library access, facilities, and compliance and governance support.
In Aotearoa New Zealand, overhead rates vary by institution, project type, and funder, but commonly fall in the range of 30–60% of eligible direct research costs, with rates of around 45–55% being typical. What that means is that if you need $10k to cover personnel time, then you will need apply for $20k – $25k total funding to account for what the university will take for overheads. Some universities apply overheads to most direct costs, while others calculate them primarily as a percentage of salary costs, excluding items such as major equipment or subcontracted services.
Funder rules differ substantially. Some funding schemes cover overheads in full, others cap the amount that can be claimed, and some do not allow overhead recovery at all. Industry-funded or contract research often attracts higher overhead rates than charitable or small-grant funding schemes.
It is essential to check both institutional overhead policies and funder rules early in the planning process. Overheads can significantly affect the total project cost and the amount of funding available for direct research activities. Even when overheads are not paid by the funder, universities may still require them to be calculated and approved internally, which can influence whether a project is financially viable.
Budgets should be reviewed carefully for internal consistency and realism. Under-budgeting is a common problem, particularly for personnel time, data management, laboratory work, and ethics or compliance-related activities. Reviewers are often sceptical of budgets that appear unrealistically low for the proposed scope of work.
If your research involves collecting data from animals and/or humans, you will almost certainly need to obtain ethics approval before starting any data collection. Ethics approval helps ensure that the study is scientifically justified, that risks are minimised, and that participants (human or animal) are treated appropriately. It is also usually a requirement to have ethics approval if you plan to publish the findings in a reputable peer-reviewed journal.
The timing of ethics approval relative to funding applications varies. Some researchers choose to seek ethics approval before applying for funding, as this can provide reviewers with greater confidence in the feasibility and readiness of the project. Others wait until funding has been secured to avoid investing time and effort in an application for a project that may not proceed. Both approaches are acceptable, but it is important to understand the expectations of the funder and to factor ethics review timelines into project planning.
Most larger research universities will have their own Animal Ethics Committee and Human Ethics Committee that allow staff and students to submit applications for review at no cost. If you are working outside of a university environment, then your options are either (1) to work with a collaborator at a university who can submit the application on your behalf or (2) apply through one of the independent ethics committees.
In Aotearoa New Zealand, independent ethics options include:
Always check the website of the relevant ethics committee for up-to-date information on submission deadlines, meeting dates, and expected review timelines, as these vary between committees and can change over time.
In addition to formal ethics approval, animal health research may require a range of permits or regulatory approvals depending on the species involved, study activities, and location.
Relevant permits and approvals may include:
These requirements should be identified early in the planning phase and factored into both timelines and budgets. Failure to obtain the correct permits can delay or prevent data collection, even when ethics approval has been granted.
Once a study design is in place, it must be translated into a practical plan for action. This section outlines how deliverables, milestones, tasks, and timelines can be used to operationalise a research project and move it from concept to completion.
Deliverables
Deliverables are the tangible outputs the project will produce. These should be concrete, measurable, and clearly linked to the study objectives. Defining deliverables early helps clarify the scope of work and provides a basis for evaluating progress and success.
Examples of common deliverables include cleaned and well-documented datasets, interim or final reports, thesis chapters, manuscript drafts, policy briefs, technical summaries, or stakeholder presentations. Each deliverable should have a clear purpose and intended audience.
Milestones
Milestones are the key checkpoints or decision points that must be reached in order to produce each deliverable. They mark meaningful progress through the project rather than day-to-day activity.
Typical milestones include ethics approval, completion of pilot testing, achievement of recruitment targets, completion of data collection, analysis sign-off, submission of reports or manuscripts, and delivery of stakeholder outputs. Milestones should be associated with target dates or timeframes and used to monitor whether the project is progressing as planned.
For grant applications, milestones are often linked to payment and reporting. Be cautious about the number of milestones you include because this can drastically increase the administrative work for both you and your funder.
Tasks
Tasks are the specific activities required to reach each milestone. These are the operational steps that translate the study design into action.
Tasks may include activities such as developing data collection instruments, recruiting participants, conducting interviews or fieldwork, entering and cleaning data, running analyses, drafting reports, and coordinating stakeholder engagement. At this stage, it is also important to clarify how the project will be managed day to day, including roles and responsibilities, meeting schedules, reporting formats, decision-making processes, and expectations for data sharing, file naming, version control, and document management.
Timelines
Timelines bring deliverables, milestones, and tasks together into a coherent schedule covering the full lifecycle of the project, from planning and approvals through to analysis and reporting.
When constructing timelines, identify task dependencies and critical path activities where delays would directly affect the project end date. Timelines should explicitly account for known constraints such as seasonal effects, academic calendars, staffing availability, and funding or ethics review cycles. Overly optimistic timelines are a common weakness in funding applications and can undermine confidence in project feasibility.
In addition to scheduling, timelines should incorporate success metrics or key performance indicators (KPIs) that will be used to track progress. These may include recruitment targets, data completeness thresholds, adherence to analysis deadlines, stakeholder engagement milestones, or publication and dissemination outcomes.
A common mistake at this stage is underestimating how long research activities will take. Recruitment often proceeds more slowly than expected, approvals take longer than planned, and analysis and write-up frequently expand beyond initial estimates. In addition, research does not occur in isolation from the rest of life—illness, caring responsibilities, competing work demands, and other unforeseen events can all disrupt timelines. Building realistic time buffers into tasks, milestones, and overall timelines is essential for maintaining project momentum and avoiding unnecessary stress or compromised outcomes.
Applying for research funding to support your project can be a daunting and sometimes discouraging process. Preparing a competitive application often requires substantial time and effort, and depending on the funding scheme, it may take weeks or months to receive an outcome, frequently with a low probability of success. This is a normal part of the research process rather than a reflection of project quality or researcher ability.
Funding strategy should be tailored to the topic area, project budget, and your career stage. For postgraduate students or small-scale projects, departmental or faculty-based scholarships and grants are often the most appropriate starting point. These typically support modest budgets, often in the range of $2,000–$3,000, and are designed to fund pilot work, thesis expenses, or limited data collection.
Many external competitive funding schemes offer specific early-career categories, commonly defined as researchers within 7–8 years of completing a PhD. Others provide explorer or pilot grants intended to support proof-of-concept studies or preliminary data generation. At the other end of the spectrum are full research programme grants, which can range from several hundred thousand dollars to multi-million-dollar investments. These larger schemes are usually beyond the scope of most small or early-stage projects and are included here primarily for awareness.
In Aotearoa New Zealand, potential funding sources include:
A key complicating factor in the New Zealand funding landscape is that many competitive grants require co-funding, often in the range of 20–60%, as a condition of eligibility. This co-funding may be provided as direct financial support or as in-kind contributions such as access to data, staff time, facilities, or equipment. The purpose of these requirements is to demonstrate industry or end-user support and to provide confidence that the research has a realistic pathway to uptake and impact.
Securing co-funding typically involves engaging early with industry organisations, professional bodies, or other stakeholders to explore opportunities for collaboration.
Strategies for maximising the chance of funding success include:
The planning phase brings together the conceptual and practical elements of a research project. By carefully designing the study, identifying required resources, developing a realistic budget, securing necessary approvals, and mapping deliverables, milestones, and timelines, this phase translates a research question into a feasible, well-structured project plan. Investing time in thorough planning reduces the risk of avoidable problems during implementation, strengthens funding and ethics applications, and provides a clear roadmap for delivering meaningful and credible research outcomes. With this foundation in place, the next phase shifts focus to project implementation, where the plan is put into action and data collection and day-to-day project management begin.
Advancing Animal Care through Shared Learning