psychology dissertation inclusion and exclusion criteria turn an abstract target population into a reproducible decision rule. They state who or what can enter a study, who or what cannot, and why. Well-designed criteria protect participants, align the sample with the research question, reduce avoidable bias, and let readers judge where the findings apply.
This guide shows how to develop, justify, apply, and report criteria for participant research, secondary data studies, systematic reviews, scoping reviews, and meta-analyses. It also separates eligibility decisions from sampling, withdrawal, and analysis exclusions, which are often confused in dissertation methods chapters.
What a Pyschology Dissertation inclusion and exclusion criteria do
Inclusion criteria are the characteristics that every eligible case must meet. Exclusion criteria identify circumstances that disqualify a case that might otherwise appear eligible. A case may be a person, interview, organisation, document, database record, or published study. The unit depends on the design.
Criteria should translate the population named in the research question into observable rules. For example, “university students experiencing academic stress” is not yet operational enough. A protocol might define current enrolment, minimum age, language needed for the validated measure, and a specified score range on a named screening instrument. Each rule needs a scientific, ethical, or practical justification.
Reporting guidance treats eligibility as a core method, not an administrative detail. The STROBE cohort checklist asks researchers to report eligibility criteria plus the sources and methods used to select participants. It also recommends reporting numbers at each stage, reasons for non-participation, and, when useful, a flow diagram.
Eligibility is not the same as sampling
| Decision | Question answered | Psychology example |
|---|---|---|
| Eligibility criteria | Who or what may enter? | Adults currently enrolled at a university and able to complete questionnaires in a validated study language |
| Sampling strategy | How are eligible cases approached or selected? | Stratified sampling by year of study |
| Withdrawal rule | When may an enrolled participant leave? | A participant may stop at any time without giving a reason |
| Analysis exclusion | When is collected data omitted from a specific analysis? | A prespecified failed attention check or missing primary outcome |
Eligibility is decided before enrolment or record inclusion. Sampling is the route used to recruit among eligible cases. Withdrawal occurs after enrolment and must respect participant rights. Analysis exclusions happen after data collection and can create serious bias if invented after results are visible. Keep these decisions separate in the protocol, participant flow record, and dissertation.
Table of Contents
Derive criteria from the research question
Begin with the constructs, population, context, timeframe, and design in the research question. Then ask what must be true for the case to provide interpretable evidence. Do not start with a convenient recruitment pool and reverse-engineer a question around it.
Map each criterion to a defensible reason
A useful eligibility rule normally has one or more of four foundations:
- Scientific fit: the case represents the population or phenomenon required by the question.
- Measurement validity: the measures can capture the intended constructs for that case.
- Ethical protection: participation would be appropriate under the approved consent and risk procedures.
- Feasibility: the research team can complete the protocol reliably within the stated setting and period.
Feasibility alone is a weak reason for excluding a group when reasonable accommodations could make participation possible. A dissertation should explain why a criterion is necessary and consider how it changes representativeness. The CIOMS international ethical guidelines emphasise scientific and social value, protection of participants, and reasoned conditions for involving potentially vulnerable groups. Local institutional guidance and ethics approval remain essential because requirements vary by setting and study.
Write criteria as observable decisions
A criterion should let two trained screeners reach the same answer from the same information. Replace vague adjectives with a source, threshold, timeframe, or verification process.
| Vague rule | More reproducible rule | Why it is stronger |
|---|---|---|
| Young adults | Aged 18 to 25 years on the consent date | Defines the boundary and date |
| High stress | Score at or above the prespecified threshold on the named validated scale | Links the decision to a measure |
| Regular social media user | Self-reported use on at least five days per week during the past three months | Defines frequency and recall period |
| Good English | Able to understand the consent information and complete the English-language instrument without translation | Ties language to the actual task |
| Complete records | Contains the exposure, outcome, age, and prespecified covariates required by the analysis plan | Names required variables |
Thresholds should not appear from nowhere. Cite the measure manual or validation evidence, explain a clinically or theoretically meaningful boundary, or identify the threshold as a pragmatic protocol choice. Where a continuous variable can remain continuous, a cutoff may discard information and narrow generalisability without improving the design.
Common types of participant criteria
Not every study needs every category. Use only the rules necessary for the question and protocol.
- Population characteristics: age range, education stage, occupation, caregiving role, or another directly relevant characteristic.
- Phenomenon or exposure: experience of bereavement, use of a service, exposure to a workplace condition, or a score on a screening measure.
- Setting and timeframe: enrolment at participating institutions or experience during a defined period.
- Communication and consent: ability to understand the approved information and provide the form of consent required by the protocol.
- Protocol compatibility: access to required equipment, capacity to attend sessions, or no conflicting intervention when scientifically necessary.
- Safety: a condition that would make a task unacceptably risky, based on the actual study procedures rather than a broad diagnostic label.
A diagnosis should not be used as a proxy for inability to consent, unreliability, or risk. Assess the relevant capability or risk directly when possible. If translated materials, accessible formats, remote participation, breaks, or supported consent can remove a barrier, discuss those options with the supervisor and ethics committee before excluding people.
Example: sleep and working memory survey
Suppose a cross-sectional dissertation asks whether sleep quality is associated with working memory among undergraduate students. Inclusion might require current undergraduate enrolment, age 18 or older, and ability to complete the validated instruments in an available language. An exclusion might cover a documented protocol conflict, such as participation in a concurrent sleep intervention if that would alter the focal association.
Excluding every participant who uses medication, reports any mental health condition, or works night shifts would usually require strong justification. Those characteristics may be measured as covariates, described as sample features, or examined in sensitivity analyses instead. The choice depends on the causal reasoning and available sample, not on a desire for a perfectly uniform group.
Example: qualitative lived-experience interview
For interviews about adjustment during the first year after international relocation, eligibility might require having moved to the study country within the previous 12 months, being at least 18, and being able to take part in an interview using one of the study languages. The researcher should define “moved” and the reference date. Recruitment should seek variation relevant to the question rather than treating demographic diversity as statistical representativeness.
A qualitative criterion should support information-rich accounts without predetermining the themes. “Participants who experienced successful adjustment” would bias a study intended to explore the full range of adjustment. A neutral temporal or experiential definition is more defensible.
Example: intervention or experimental study
An online attention-training experiment might include adults within the approved age range who can use the required device and complete baseline assessment. Exclusion criteria could address previous exposure to the exact experimental task or a safety issue caused by the stimulus. The current CONSORT reporting guideline provides the relevant framework for randomised trials, while specialised extensions may apply to social and psychological interventions.
Eligibility should be assessed before random assignment. Failure to finish the intervention later is attrition, not retrospective ineligibility. Analyse and report it according to the prespecified plan.
Criteria for reviews and evidence syntheses
In a systematic review, criteria define which reports and studies contribute evidence. They should be established before screening begins and applied consistently at title and abstract screening, full-text screening, and data extraction.
The PRISMA 2020 checklist includes reporting recommendations for eligibility and study selection. Its flow diagram maps records identified, included, and excluded, with reasons for full-text exclusions. These tools support transparent reporting but do not design the review question for you.
Build criteria from a framework
For intervention questions, Population, Intervention, Comparator, Outcomes, and Study design can structure eligibility. Qualitative questions may be better served by Sample, Phenomenon of Interest, Design, Evaluation, and Research type. Choose a framework that fits the question rather than forcing every review into one template.
| Domain | Example inclusion rule | Possible exclusion reason |
|---|---|---|
| Population | Adults providing unpaid care to a relative with dementia | Paid professional caregivers only |
| Phenomenon | Psychological experiences of caregiver burden | Physical workload only, without psychological outcomes |
| Design | Peer-reviewed qualitative studies using interviews or focus groups | Editorial, protocol, or purely quantitative survey |
| Timeframe | Studies published from 2000 through the final search date | Published before the justified start date |
| Language | Languages the review team can reliably screen and analyse | Full text unavailable in a supported language |
Language and publication-status restrictions can introduce bias, so state and justify them. “Peer reviewed only” is not automatically superior if the question concerns emerging practice or publication bias. Align sources with the dissertation’s purpose and resources.
A scoping review may use broader criteria to map concepts, populations, and evidence types. A meta-analysis may need additional rules about effect-size information, independence, and comparable outcomes. Do not exclude a relevant study solely because an effect size is not printed; it may be calculable or obtainable from authors.
Criteria for secondary data and records
In a secondary data dissertation, the case may be a participant record, assessment occasion, or organisation. Define the source dataset, observation window, required variables, duplicate handling, and whether linked records must meet a matching-quality threshold.
Separate eligibility from missing-data treatment. A record lacking the primary outcome may be ineligible for the primary analysis, but a record with one missing covariate may still contribute under a prespecified missing-data method. Blanket complete-case eligibility can silently change the target population and reduce precision.
Also distinguish data-quality checks from implausible-value decisions. Define valid ranges from the instrument or data dictionary before inspecting associations. Record correction, recoding, and exclusion decisions in a reproducible log.
Prevent avoidable selection bias
Narrow criteria can improve internal interpretability while weakening external relevance. Broad criteria can improve applicability while increasing heterogeneity or practical complexity. There is no universally correct width. The aim is a justified match between the research question, design, risks, and intended inference.
Ask who becomes invisible under each rule. Age, disability, language, location, digital access, diagnosis, pregnancy, and comorbidity restrictions may remove groups that experience the phenomenon differently. An exclusion should not be copied from an unrelated study merely because it looks conventional.
Use a criterion audit
- State the target population in one sentence.
- List every proposed criterion and its information source.
- Write a one-sentence justification for each rule.
- Identify groups disproportionately removed by the rule.
- Consider accommodation, stratification, measurement, or sensitivity analysis as alternatives.
- Check that the remaining sample can still answer the research question.
- Review the final criteria with the supervisor and ethics process before recruitment or screening.
This audit does not remove all bias. It makes trade-offs visible and helps the researcher avoid criteria based only on convenience or untested assumptions.
Plan screening before data collection
Create a screening form with one item per criterion, permitted response options, the information source, and instructions for ambiguous cases. Pilot it on a small set of hypothetical or real cases that are not part of the final analysis when appropriate.
For participant studies, collect only information needed to determine eligibility. Protect screening data under the approved data management plan. Explain what happens to information from people who are screened but not enrolled.
For reviews, train at least two screeners when the protocol requires independent decisions. Calibrate interpretations on a sample of records, document disagreements, and use a predefined resolution process. A reason such as “wrong population” should correspond to a written rule, not an intuitive judgement.
Create a transparent decision hierarchy
A case may fail several criteria. For a flow diagram, choose a consistent hierarchy for the primary exclusion reason. For example: duplicate, wrong population, wrong phenomenon or intervention, wrong outcome, wrong design, unavailable full text. The hierarchy prevents the reported reason from depending on which criterion a screener notices first.
Keep an audit trail without exposing confidential details. Record the date, decision, coded reason, screener, and resolution of uncertainty. For participant research, do not include identifying information in the dissertation flow chart.
Handle ambiguity and protocol changes
Real cases expose gaps in apparently clear criteria. A participant may be between education stages, a record may span the time boundary, or a review paper may include a mixed-age sample. Anticipate these cases with rules such as a required subgroup result, a percentage threshold, or author contact.
If a criterion must change after screening or recruitment starts, do not quietly rewrite the method. Record what changed, when, why, which cases were affected, and whether ethics or protocol amendments were needed. Update the preregistration or protocol record where applicable. Consider a sensitivity analysis if the change could alter conclusions.
Common mistakes and repairs
| Mistake | Why it matters | Repair |
|---|---|---|
| Repeating the target population as the only criterion | Screeners cannot apply an abstract label consistently | Define observable characteristics, sources, and boundaries |
| Using exclusion as the logical opposite of every inclusion rule | Creates redundant lists and hides special disqualifying conditions | State all required inclusion conditions, then list distinct exclusions |
| Choosing cutoffs after seeing the data | Allows results to influence sample composition | Prespecify thresholds and justify them |
| Calling attrition an exclusion | Conceals participant flow and may bias analysis | Report eligibility, enrolment, withdrawal, completion, and analysis separately |
| Excluding missing data without a plan | Can change the analytic population unpredictably | Define required variables and missing-data methods in advance |
| Using broad safety exclusions | May be unfair and reduce applicability | Link each restriction to a concrete procedure and risk |
| Giving no full-text exclusion reasons in a review | Readers cannot reproduce selection | Use a coded log and report a flow diagram |
A dissertation-ready reporting template
Adapt the following structure rather than copying it unchanged:
Cases were eligible when they met all of the following criteria: [criterion, operational definition, and source]; [criterion, definition, and source]; and [criterion, definition, and source]. Cases were excluded when [distinct disqualifying condition] or [distinct condition]. These rules were selected because [scientific, measurement, ethical, and feasibility rationale]. Eligibility was assessed by [who] using [form or procedure] before [enrolment, extraction, or analysis]. Ambiguous cases were resolved by [process]. Participant or record flow and reasons for exclusion were documented using [method].
In the dissertation, place the complete criteria in the methods chapter. Report recruitment or search sources, screening procedure, numbers at each stage, and reasons for exclusion in the results or study-flow section. Discuss how the rules affect transferability or generalisability in the limitations.
Final quality checklist
- Every criterion follows from the research question or protocol.
- Each rule is observable, bounded, and consistently applicable.
- Thresholds and timeframes have a stated rationale.
- Eligibility, sampling, withdrawal, attrition, and analysis exclusions are separate.
- Unnecessary demographic, diagnostic, language, and access restrictions have been challenged.
- Screening information is collected and stored ethically.
- Ambiguous cases and protocol changes have a documented process.
- Participant or record flow can be reported transparently.
- The criteria match the approved ethics application, protocol, and analysis plan.
Frequently asked questions
How many inclusion and exclusion criteria should a dissertation have?
There is no correct number. Include every rule necessary to define the analytic population and protect the protocol, but remove redundant or unjustified restrictions. A short, precise list is stronger than a long list copied from previous studies.
Should exclusion criteria be the opposite of inclusion criteria?
Usually not. If inclusion requires age 18 or older, “under 18” need not be repeated as an exclusion. Reserve the exclusion list for distinct disqualifying conditions, such as prior exposure to an experimental task or a protocol-specific safety issue.
Can language be an inclusion criterion?
Yes, when valid consent, interviewing, or measurement is only available in specified languages. State the practical and measurement reason, consider translation or interpretation where feasible, and discuss how the restriction affects the population represented.
Can incomplete questionnaires be excluded?
Do not use a blanket rule unless it is justified. Define which variables are required for each analysis, distinguish participant withdrawal from item nonresponse, and apply the prespecified missing-data plan. Report the number and reasons for analysis exclusions.
When should review eligibility criteria be finalised?
Finalise them in the protocol before formal screening. Pilot screening may reveal ambiguous wording, which can be clarified transparently before or early in screening. Log any later amendment and its effect on included records.
Where should criteria appear in a dissertation?
State them in the methods chapter, alongside recruitment or search procedures. Report the number screened, excluded, included, completed, and analysed in the results or flow section. Reflect on the resulting boundaries of inference in the discussion.
Conclusion
Strong inclusion and exclusion criteria are concise decision rules, not decorative methods text. They connect the research question to the cases that produce evidence, make screening reproducible, protect participants, and show readers where conclusions apply. Draft them before selection begins, justify every restriction, record decisions consistently, and report the resulting flow without hiding attrition or post hoc exclusions.
Get ethical support with your eligibility plan
If you need help refining psychology dissertation criteria, seek feedback that preserves your authorship and follows your institution’s academic integrity and research ethics requirements. A responsible review can test alignment among your question, sampling plan, measures, consent process, and analysis without fabricating data or making eligibility decisions on your behalf.
