Psychology dissertation data management plan decisions should be made before collection begins, then reviewed whenever the project, risks, or sharing arrangements change.
A data management plan, often shortened to DMP, explains what research data you will create or obtain, how you will organise and protect them, who may access them, how changes will be documented, and what will happen after the dissertation. It connects research quality with participant protection. It is not merely a paragraph promising to use a password.
Requirements vary across universities, funders, countries, and study types. Use your institution’s current template, ethics approval, information-security rules, consent wording, and applicable law as the controlling documents. This guide provides a practical psychology framework, not legal advice.
Importance of having a Psychology Dissertations data management plan
Psychology projects frequently combine sensitive human data with multiple working files. A modest interview study may create contact details, consent records, audio, transcripts, field notes, a coding project, quotations, and a re-identification key. A survey may generate recruitment logs, raw exports, scoring syntax, cleaned datasets, codebooks, output, and figures. Without a plan, it becomes difficult to know which file is authoritative or who can see identifiable material.
Good management supports integrity because another authorised researcher can understand how raw inputs became reported results. It supports continuity because a damaged laptop or accidental overwrite does not destroy the project. It also supports ethical review because promises about confidentiality, retention, deletion, and sharing can be translated into concrete actions.
The UK Data Service research data guidance describes management across the lifecycle as organising, protecting, documenting, preserving, and, where appropriate, sharing data. Its guidance links these practices with research quality, integrity, and reproducibility.
Start with the rules that apply to your project
Collect the relevant requirements before choosing software or storage. Check your dissertation handbook, ethics decision, participant information and consent materials, funder conditions, data provider licence, collaboration agreement, repository rules, and institutional retention schedule. Resolve conflicts through your supervisor, ethics body, research office, information-security team, or data-protection contact.
A funder template is not automatically the right template for every student. The current NIH data management and sharing plan guidance, for example, applies to covered NIH-supported work and has its own required format. It is useful for seeing how data types, repositories, timing, privacy, and justified sharing limits connect, but it does not replace a student’s local instructions.
Create a requirements register containing the source, exact obligation, responsible person, deadline, and evidence of compliance. Record the version or access date because policies change. Do not copy a generic retention period or promise unrestricted sharing without checking whether it matches consent, law, licences, and institutional policy.
Table of Contents
Define what counts as research data
Inventory information by function, not only by file extension. Research data can include observations, recordings, images, physiological signals, questionnaire responses, test scores, derived variables, coded extracts, analytic outputs, and the documentation or code needed to interpret them. Administrative records such as contact details and signed consent forms also need protection even when they are not analysed.
| Data group | Psychology example | Management question |
|---|---|---|
| Administrative | Contact list, scheduling log, consent record, compensation details | Can these be stored separately from research responses? |
| Raw research data | Survey export, audio recording, eye-tracking file, observation sheet | How will the original be preserved as read-only? |
| Processed data | Scored scale, anonymised transcript, cleaned dataset | Which documented steps connect it to the raw source? |
| Analysis materials | Syntax, scripts, codebook, qualitative coding project, audit notes | What is needed to reproduce or understand decisions? |
| Outputs | Tables, figures, model objects, quotation list | Which version supports the submitted dissertation? |
| Documentation | README, data dictionary, protocol, provenance log | Could an authorised reader understand the files without guessing? |
For secondary research, inventory both the received data and your derived materials. The licence or data-use agreement may forbid redistribution even when your code and variable definitions can be shared. See the secondary data analysis guide for additional design and reporting considerations.
Write the core sections of the plan
A practical plan should be specific enough to guide action. Replace statements such as “data will be secure” with named locations, approved controls, access roles, backup responsibility, and review points. The following elements can be adapted to most institutional templates.
| Plan element | What to specify | Evidence to retain |
|---|---|---|
| Data description | Types, formats, approximate volume, sources, and sensitivity | Data inventory and variable or file list |
| Collection and quality | Capture method, validation, transcription, scoring, and quality checks | Protocol, pilot record, and change log |
| Organisation | Folder structure, naming convention, master-file rule, and versioning | README and naming guide |
| Documentation | Codebook, metadata, decision trail, software, and dependencies | Data dictionary, syntax, and environment information |
| Security and access | Approved storage, encryption, permissions, transfer, and incident route | Access register and periodic review |
| Ethics and privacy | Consent, minimisation, identifiability, confidentiality, and disclosure risk | Approved ethics documents and risk assessment |
| Retention and disposal | What is retained, for how long, by whom, and how disposal is verified | Retention schedule and destruction record |
| Sharing and preservation | Files, timing, repository, licence, access level, and justified restrictions | Deposit record, persistent identifier, and data availability statement |
| Roles and resources | Owner of each task, support services, costs, and review dates | Responsibility table and dated revisions |
The UK Data Service’s practical research data management resource covers these connected areas, including planning, security, backup, formats, version control, file naming, documentation, and sharing. Use recognised guidance to inform decisions, but state the controls actually approved for your project.
Classify data by sensitivity and identifiability
Security should follow risk. Begin by asking whether a person is directly or indirectly identifiable and what harm could follow unauthorised access, alteration, loss, or disclosure. Direct identifiers include names and contact details. Indirect identifiers, combinations of demographic details, rare experiences, voices, images, locations, or free-text narratives can also reveal identity.
Pseudonymisation and anonymisation are not interchangeable. Pseudonymised data retain a route back to a person through separately held information. The UK Information Commissioner’s Office guidance explains that pseudonymised data remain personal data for a party holding the additional information. Although this is UK guidance, the conceptual distinction is useful internationally; apply the legal definition relevant to your location.
In a longitudinal wellbeing study, participant codes may be needed to connect waves. Store the linking key separately with more restricted access, not beside the response file. In an interview study, replacing names may be insufficient if a rare occupation, small organisation, and unique event identify the speaker when combined. Review the entire disclosure context before labelling a file anonymous.
| Risk level | Example | Typical control direction |
|---|---|---|
| Identifiable and sensitive | Clinical interview audio linked to contact details | Approved restricted storage, strong access control, separate identifiers, secure transfer, and limited retention |
| Pseudonymised | Participant-coded symptom scores with a separate key | Protect both files, restrict the key more tightly, and document who may reconnect them |
| De-identified but disclosure-prone | Detailed free-text accounts from a small community | Assess combinations and context, redact carefully, and consider controlled access |
| Effectively anonymous or aggregate | Non-disclosive summary statistics | Verify disclosure risk and preserve documentation explaining derivation |
Do not collect identifiers “just in case.” Data minimisation reduces risk and management burden. Align the plan with the approved psychology dissertation ethics materials and update ethics approval before making material changes where required.
Create an organised folder and file system
Choose one approved master location and a predictable structure. Separate administrative, raw, processed, analysis, output, and documentation files. Keep raw data unchanged after import where technically feasible. Conduct cleaning and transformations in copies generated by recorded procedures.
A simple project structure might include folders for administration, raw data, processed data, code, qualitative analysis, outputs, documentation, and deposit materials. Permissions can differ by folder. A transcriber may need access to selected recordings but not the participant contact list; a statistical adviser may need a de-identified analysis dataset but not consent records.
Use meaningful, safe file names
Agree a convention before collection. A name can include the project code, content, study wave, participant code, date in year-month-day order, and version or status. For example, psydmp_w02_p014_transcript_v01_2026-08-15.txt communicates more than final_interview_new.txt. Never place participant names, diagnoses, or other sensitive details in file names because names may appear in synchronisation logs and recent-file lists.
Use short consistent terms, valid characters, and dates that sort chronologically. Document every abbreviation in the README. Avoid relying on “final,” because files often continue to change. Define milestone states such as raw, cleaned, analysis-ready, submitted, and archived.
Control versions and provenance
Version control means knowing what changed, when, why, and by whom. For quantitative work, preserve syntax that transforms the raw export into the analysis dataset. For qualitative work, record transcript corrections, codebook changes where the method uses a codebook, memo development, and the relationship between extracts and source material.
Do not overwrite the only copy of a dataset while recoding missing values. Create scripted transformations or a documented change log. The goal is a provenance trail from source to result, not a collection of unexplained duplicates.
Build documentation that makes the data interpretable
A README should state the project title, responsible contact or office, folder contents, naming rules, data relationships, software, access conditions, and update history. A quantitative data dictionary should define variable names, labels, types, units, permitted values, missing-value codes, scoring rules, derived variables, and relevant instruments.
For qualitative data, document the sampling context, data-generation method, transcription convention, anonymisation decisions, analytic approach, and relationships among recordings, transcripts, memos, codes, and reported extracts. Avoid placing identifying explanations in documentation intended for public release.
Documentation should also distinguish planned from exploratory decisions. Link the plan to your preregistration, where applicable, and to your data analysis workflow. A DMP explains stewardship; it does not replace an analysis plan.
Choose secure storage, access, transfer, and backup
Use institutionally approved systems appropriate to the data classification. Consumer email, personal messaging, an unapproved personal cloud account, or an unencrypted portable drive may breach policy even if convenient. Confirm where servers are located, whether backups are provided, who administers accounts, and what happens when student access ends.
Apply least-privilege access: each person receives only the data required for their task and only for the necessary period. Use individual accounts rather than shared passwords, enable approved multifactor authentication, review permissions, and remove access promptly when roles end. Record external processors such as transcription services and confirm that the ethics and contractual arrangements cover them.
Encryption can protect data in storage and transfer, but it does not repair excessive access or poor key management. Follow institutional support on encryption rather than inventing a personal system. Never send the password or key through the same channel as the protected file.
Separate synchronisation from backup
A synchronised folder may reproduce an accidental deletion or corrupted file across devices. A backup should allow restoration from an earlier state. Define what is backed up, frequency, location, retention of versions, responsible person, and restoration test. The plan is incomplete until recovery has been tested with non-sensitive material or an approved procedure.
Keep enough independent copies to survive a device or location failure, following institutional policy. Do not turn “offsite” into permission to place sensitive data on a personal device. Your approved service may already provide managed replication and backup.
Plan quality assurance without changing raw data
Quality management begins during psychology dissertation data collection. Define range and logic checks, duplicate detection, device calibration, scoring verification, transcription review, observer training, and audit procedures that fit the design. Record corrections rather than silently editing source files.
For a survey, retain the untouched platform export, a scripted cleaning file, a data-quality report, and the analysis-ready dataset. For behavioural coding, keep the coding manual version, coder assignments, adjudication decisions, and reliability outputs. For reflexive qualitative analysis, preserve a coherent trail of materials and decisions without misrepresenting the method as mechanically reproducible.
Align consent, sharing, and preservation
Open data is not the same as responsible data. Decide what can be shared, with whom, for what purposes, under which access conditions, and for how long. The APA data-sharing guidance connects sharing for verification with participant confidentiality, informed consent, proprietary rights, and other ethical constraints.
Plan sharing when consent materials are written, not after collection. Explain foreseeable reuse accurately without promising that re-identification is impossible. For sensitive qualitative or clinical data, appropriate outputs may include a metadata record, code, measures where licensing permits, non-disclosive summaries, or a controlled-access dataset rather than open raw files.
Select a recognised subject or institutional repository where appropriate. Check accepted formats, access controls, preservation policy, persistent identifiers, licensing, costs, and withdrawal or correction procedures. A project website or temporary file-sharing link is not necessarily a preservation repository.
The FAIR principles describe data and metadata that are findable, accessible under stated conditions, interoperable, and reusable. “Accessible” does not require unrestricted public access. Sensitive data can be discoverable through metadata while access remains controlled. Respect community governance and collective interests as well as individual privacy, particularly in cross-cultural or Indigenous research.
Set retention and secure disposal rules
State which files will be retained, the rule that determines the period, the start date, custodian, approved location, review point, and disposal method. Different materials may have different schedules. Contact lists may be deleted after follow-up ends, while de-identified analysis data, syntax, and documentation may be retained under institutional or publication rules.
Deletion should be verifiable and appropriate to the medium. Moving a file to a recycle folder may not remove backups, synchronised copies, or exported files. Include paper records, portable media, recording devices, transcription platforms, local download folders, and collaborators’ copies. Ask the responsible technical service how managed systems implement retention and secure destruction.
Assign responsibilities and review the plan
Even a solo dissertation involves shared responsibility. The student may organise files and maintain documentation; the supervisor may approve access and verify alignment; institutional services may manage storage and backup; the ethics body may approve changes; and a repository may preserve deposits. Name roles, not assumptions.
Review the plan at approval, before collection, after a pilot, when staff or tools change, before analysis, before submission, and before sharing or disposal. A version history should record the date, change, reason, approval needed, and person responsible. Update related consent, protocol, and ethics documents when a change affects participants or approved handling.
Common data management mistakes
- Keeping names and research responses in the same broadly accessible file.
- Editing raw data directly without syntax or a change log.
- Calling coded data anonymous while retaining the linking key.
- Using personal storage because it is easier than the approved system.
- Assuming synchronisation is a tested backup.
- Putting sensitive information in file names or free-text notes.
- Promising open sharing that consent, licences, or disclosure risk do not allow.
- Choosing a retention period without an authoritative source.
- Leaving disposal until after institutional access has expired.
Record genuine constraints in the dissertation limitations where they affect interpretation or reproducibility. Do not conceal them behind generic assurances.
Frequently asked questions
How long should a psychology dissertation data management plan be?
Follow the required template or word limit. Length matters less than operational detail. A concise plan can be adequate if it identifies data, controls, responsibilities, timelines, and justified sharing or retention decisions clearly.
Is a password-protected laptop enough?
Usually not as a complete plan. You must consider approved storage, encryption, account security, access, transfer, backup, incident response, retention, and disposal. Use the controls required for your project’s risk classification.
Should consent forms be stored with survey or interview data?
Normally, identifiable consent and contact records should be separated from research responses with tighter access, subject to local policy. Document how participant codes link records and who may access the key.
Can I share pseudonymised psychology data publicly?
Not automatically. Pseudonymised data may remain personal and may still permit identification through combinations or external information. Check consent, disclosure risk, law, ethics approval, institutional policy, and repository access options.
Do I need a DMP for a systematic review?
A review may not collect participant-level data, but it still creates search records, screening decisions, extraction files, appraisal data, synthesis code, and outputs. A proportionate plan improves version control, transparency, backup, licensing, and reproducibility.
Where should the plan appear in the dissertation?
Follow programme requirements. Relevant decisions may be summarised in the methodology or ethics section, with a full approved plan in an appendix or repository where permitted. Avoid publishing security details or identifiable information. The appendices guide explains how to place supporting material safely.
Conclusion
A strong psychology dissertation data management plan turns ethical and methodological commitments into a working system. Inventory the data, classify risk, define organisation and provenance, use approved security and backup, document transformations, align consent with sharing, and assign retention and disposal responsibilities. Review the plan as the project changes.
Psychology Dissertation Help can ethically review the clarity and completeness of your plan, file logic, documentation outline, and reporting structure. Any support should preserve your responsibility for the research, follow institutional rules, and never involve transferring confidential participant data through an unapproved channel.
