18 - Subject Data Collection Protocol¶
Protocol for Collecting Subject Data for Cognitive Emulation¶
SCIENTIFIC BOUNDARY: This protocol governs the collection of behavioural, physiological, and self-report data from human participants for the purpose of building subject-specific cognitive emulation profiles. All collected data captures observable, statistical patterns only — it does NOT capture subjective experience, consciousness, private thoughts, or moral identity. EEG data is biometric data requiring enhanced protections. Writing samples capture linguistic style markers, not inner thoughts. Preference surveys capture stated preferences, not underlying values. This protocol must be reviewed and approved by an appropriate ethics committee before any data collection begins.
1. Scope and Purpose¶
This document specifies the data collection procedures for the Subject-Specific Cognitive Emulation extension (Document 17) of the Consciousness-Indicator Architecture. It covers the collection of five data channels: EEG recordings, behavioural traces, writing samples, preference surveys, and autobiographical memory cues.
This protocol is designed to ensure that:
- All data collection is conducted with fully informed, voluntary, and revocable consent.
- Data minimisation principles are applied at every stage.
- Participant privacy and data security are maintained throughout the data lifecycle.
- The scientific boundaries of the emulation are clearly communicated to all participants.
- Ethical review requirements are met before, during, and after data collection.
This protocol complements (and does not replace) the physical EEG safety guidelines in Document 13 and the institutional ethics review requirements described therein.
2. Consent Requirements¶
2.1 Informed Consent¶
Informed consent is the non-negotiable foundation of all data collection for subject-specific cognitive emulation. Consent must be:
-
Informed: The participant must receive a clear, jargon-free explanation of what the study involves, what data will be collected, how it will be used, who will have access, how long it will be stored, and what their rights are. The explanation must cover the scientific boundary explicitly — the participant must understand that the emulation captures statistical patterns, not their mind, consciousness, or identity.
-
Voluntary: Consent must be given freely, without coercion, inducement, or pressure. The participant must understand that they can refuse to participate or withdraw at any time without penalty or negative consequences. Refusal or withdrawal must not affect the participant's access to any services, benefits, or relationships.
-
Specific: Consent must be obtained separately for each data channel (EEG, behavioural, writing, preferences, memory cues). The participant may consent to some channels and decline others. The consent form must list each channel individually with a checkbox or signature for each.
-
Documented: Consent must be recorded in writing (paper or digital with secure signature). The consent form must include the date, the participant's name (or code), the version of the consent form, and the name of the person obtaining consent.
-
Ongoing: Consent is not a one-time event. If the study design changes, if new data channels are added, or if the data use is expanded, renewed consent must be obtained. The participant must be reminded of their right to withdraw at the start of each session.
2.2 Consent Scope¶
The consent form must explicitly address the following points:
| Topic | Required Disclosure |
|---|---|
| Purpose | "The purpose of this study is to build a statistical profile of your cognitive and behavioural patterns for research into personalized cognitive-state estimation. This is NOT an attempt to read your mind, copy your identity, or transfer your consciousness into a computer." |
| EEG data | "EEG records electrical activity on your scalp. It does NOT read your thoughts. It will be used to estimate proxies for attention, workload, and fatigue levels." |
| Behavioural data | "Your responses during cognitive tasks (choices, response times, accuracy) will be recorded. This data captures your observable performance patterns, not your internal reasoning or thoughts." |
| Writing samples | "Text you provide will be analyzed for linguistic style markers such as vocabulary diversity and sentence structure. This captures surface writing patterns, NOT your beliefs, opinions, or inner thoughts." |
| Preference surveys | "You will be asked about your preferences in several domains. This captures your stated preferences, NOT your underlying values or moral reasoning." |
| Memory cues | "You may voluntarily provide autobiographical memory cues. These are memories YOU choose to share. You do NOT have to provide any, and can withdraw any cue at any time." |
| Data storage | "Your data will be stored under a pseudonymous code (e.g., sub-001). The mapping between your code and your identity will be stored separately and securely." |
| Data retention | "Your data will be retained for [X] years, after which it will be securely deleted unless you consent to an extension." |
| Right to withdraw | "You can withdraw from the study at any time, for any reason, without penalty. If you withdraw, your data will be deleted (or anonymized, if you prefer)." |
| Right to delete | "You can request deletion of all or part of your data at any time by contacting [researcher contact]." |
| Risks | "EEG recording may cause mild skin irritation from electrodes. Extended sessions may cause fatigue. There are no psychological risks beyond those of normal cognitive tasks." |
| Benefits | "There are no direct personal benefits from participating. Your data contributes to scientific research on cognitive-state estimation." |
| Contact | "If you have questions or concerns, contact [PI name] at [email/phone] or the ethics committee at [contact]." |
2.3 Consent Revocation¶
Participants may revoke consent for any or all data channels at any time. The revocation procedure must be:
- Prompt: Revocation takes effect as soon as practically possible (within 48 hours at most; immediately if technically feasible).
- Documented: The date and scope of revocation (which channels) are recorded in the consent log.
- Complete: All data from the revoked channel(s) is securely deleted (not just deactivated or masked).
- Confirmed: The participant is informed when revocation has been completed and provided with a confirmation record.
- Non-penalizing: Revocation does not affect the participant in any way.
If revocation of one channel makes the profile non-viable, the participant is informed and offered the choice of: (a) providing alternative data, (b) having the entire profile deleted, or (c) retaining the partial profile with a documented scope limitation.
3. Data Minimisation Principles¶
3.1 Collect Only What Is Needed¶
Every data element collected must be justified by a specific, documented research question. The following data minimisation rules apply:
- Channel justification: Each data channel (EEG, behavioural, writing, preferences, memory cues) must be justified in the study protocol. Channels that are not needed for the research question must not be collected.
- Duration justification: The duration of data collection must be the minimum necessary. If 30 minutes of EEG is sufficient, do not collect 60 minutes.
- Resolution justification: Data resolution (sampling rate, text length, number of survey questions) should be the minimum that supports the research objective.
- Frequency justification: The number of sessions should be the minimum necessary. If a single session provides sufficient data, do not schedule multiple sessions.
3.2 Retain Only What Is Needed¶
- Define retention period: Before data collection, specify a retention period (recommended maximum: 2 years from last profile update).
- Aggregate when possible: Store derived features (means, variances, distributions) rather than raw data whenever the research question permits.
- Delete on schedule: Implement automatic deletion triggers based on the retention period. Review retention quarterly.
- Archive option: After retention, anonymized data (identity mapping destroyed) may be archived for future statistical analysis, subject to ethics committee approval.
3.3 Access Only Who Needs It¶
- Role-based access: Define access roles (principal investigator, data analyst, system administrator) and grant access based on role.
- Audit logging: Log every access event (who, when, what data was accessed).
- No public access: Subject data must never be made publicly available in identifiable form.
- Sharing restrictions: Data sharing with external collaborators requires separate ethics approval and participant consent.
4. Privacy Protection Measures¶
4.1 Pseudonymisation¶
Every participant is assigned a unique, opaque subject identifier at the start of the study. The identifier must:
- Be unrelated to the participant's name, date of birth, or other PII.
- Follow a consistent format (recommended:
sub-XXXwhereXXXis a sequential or random code). - Be used in all data files, logs, profiles, and metadata.
- Be mapped to the participant's real identity in a separate, secure file that is NOT included in the dataset.
The identity mapping file must be: - Encrypted at rest (AES-256 or equivalent). - Stored on a separate system from the research data. - Accessible only to the principal investigator and designated data stewards. - Deleted when the participant withdraws or the retention period expires.
4.2 Data Encryption¶
| Data State | Encryption Requirement |
|---|---|
| At rest (disk) | AES-256 or equivalent full-disk or file-level encryption |
| In transit (network) | TLS 1.2 or higher for all data transfers |
| In memory | No specific requirement, but minimize data exposure time |
| In backups | Same as at-rest encryption; backup encryption keys stored separately |
4.3 Access Controls¶
- Physical access to storage media is restricted to authorized personnel.
- Logical access requires authentication (password + 2FA for remote access).
- Data files are stored in access-controlled directories (not world-readable).
- Temporary copies (working files, analysis scripts) are deleted after use.
- Laptop/desktop machines used for data processing are encrypted and have screen lock enabled.
4.4 Secure Deletion¶
When data is deleted (due to consent revocation, retention expiry, or participant request):
- Delete all copies of the data (primary storage, backups, working copies, cache).
- Use secure deletion methods (multi-pass overwrite or cryptographic erasure).
- Verify deletion by attempting to recover the data (or by checking deletion logs).
- Record the deletion event in the audit log (date, scope, method, verification).
- Inform the participant that deletion has been completed.
5. Right to Delete Data¶
5.1 Absolute Right¶
Participants have the absolute right to request deletion of all or part of their data at any time, for any reason, without needing to provide justification. This right is guaranteed regardless of: - Whether the data has already been incorporated into a profile. - Whether the data has been used in analysis. - Whether results from the data have been published. - Whether the study is still ongoing.
5.2 Deletion Scope¶
The participant may request deletion of: - All data (full deletion): All data across all channels, including the profile, consent records, and identity mapping. - Specific channels (partial deletion): Data from one or more specific channels (e.g., delete writing samples but retain EEG data). - Specific sessions (session deletion): Data from one or more specific recording sessions.
5.3 Deletion Limitations¶
If data has been included in published results, the participant's data cannot be "un-published." In this case: 1. The raw data is deleted as requested. 2. Aggregated results (group means, statistical summaries) that do not identify the individual may be retained. 3. The participant is informed that aggregated results may remain in published material. 4. Future publications must not include any data from the withdrawn participant.
5.4 Deletion Procedure¶
- Participant submits a deletion request (verbal, written, or electronic).
- Researcher confirms the request and specifies the scope of deletion.
- Data is identified and located across all storage locations.
- Data is securely deleted using the procedure described in Section 4.4.
- Deletion is verified and documented in the audit log.
- Participant is informed that deletion is complete and provided with a confirmation record.
6. Behavioural Task Design Guidelines¶
6.1 Task Selection¶
Behavioural tasks must be selected to elicit the cognitive states and decision patterns relevant to the research question. Tasks should be:
- Validated: Use established cognitive tasks with published psychometric properties wherever possible.
- Appropriate for EEG: Tasks that produce measurable EEG correlates (e.g., attention modulation, workload variation, fatigue effects) are preferred when EEG is being collected simultaneously.
- Appropriate for the participant: Task difficulty should be calibrated to the participant's ability level. Tasks that are too easy or too difficult produce floor/ceiling effects that are uninformative.
- Ethically reviewed: All tasks must be reviewed and approved by the ethics committee before use.
6.2 Recommended Tasks¶
| Task | Cognitive Domain | EEG Correlate | Typical Duration |
|---|---|---|---|
| N-back (1-back, 2-back, 3-back) | Working memory | Theta power, workload proxy | 10-20 min |
| Sustained Attention to Response Task (SART) | Sustained attention | Alpha/beta ratio, attention proxy | 10-15 min |
| Psychomotor Vigilance Task (PVT) | Vigilance/fatigue | Theta/beta ratio, fatigue proxy | 20-30 min |
| Visual search task | Selective attention | P300, attention proxy | 10-15 min |
| Oddball paradigm | Prediction error | Gamma power, prediction error proxy | 10-15 min |
| Mental arithmetic | Executive function | Beta power, workload proxy | 10-15 min |
| Reading comprehension | Language processing | Left-hemisphere beta, workload proxy | 10-20 min |
| Multi-attribute decision task | Decision making | Response time distribution, choice preference | 15-20 min |
6.3 Task Design Principles¶
- Counterbalance conditions: If a task has multiple conditions (e.g., different difficulty levels), counterbalance the order across participants and sessions.
- Include catch trials: Include catch trials or attention checks to identify inattentive or disengaged participants.
- Record response times: Always record response times in milliseconds. Use high-resolution timing (not just seconds).
- Log trial-level data: Record data at the trial level (not just aggregate summaries) to support fine-grained analysis.
- Pilot test: Pilot all tasks before the main study to verify timing, difficulty, and data quality.
7. EEG Event Marker Alignment¶
7.1 Importance of Temporal Alignment¶
When EEG is collected simultaneously with behavioural tasks, precise temporal alignment between EEG windows and behavioural events is essential. Misalignment introduces noise and can produce spurious or misleading results. The alignment procedure must be documented and verified.
7.2 Event Marker Types¶
| Event Type | Description | Recommended Marker |
|---|---|---|
| Stimulus onset | Exact moment a stimulus appears on screen | Digital trigger (preferred) or software timestamp |
| Stimulus offset | Exact moment a stimulus disappears | Digital trigger or software timestamp |
| Participant response | Exact moment of button press, key press, or verbal response | Digital trigger (response box) or software timestamp |
| Task block start/end | Boundaries between experimental conditions | Software timestamp or manual log entry |
| Break start/end | Boundaries between recording blocks | Manual log entry |
| Adverse event | Any discomfort, distraction, or unusual event | Manual log entry with timestamp |
7.3 Hardware vs. Software Triggers¶
- Hardware triggers (digital TTL pulses sent from the stimulus computer to the EEG amplifier) provide the best temporal accuracy (sub-millisecond). Use hardware triggers whenever available.
- Software triggers (timestamps recorded by the stimulus presentation software) are acceptable but introduce timing uncertainty (typically 5-50 ms depending on the system). If software triggers are used, document the expected timing uncertainty.
- Manual log entries (experimenter writes timestamps by hand) are the least accurate and should be used only as a supplement to hardware or software triggers.
7.4 Verification Procedure¶
After each recording session: 1. Verify that event markers are present in the EEG data file (or in a separate events file). 2. Check that the number of markers matches the expected number (e.g., number of trials × events per trial). 3. Spot-check a few markers against the stimulus presentation log to verify alignment. 4. If significant misalignment is detected, document it and consider excluding affected trials from analysis.
8. Writing Sample Collection Procedure¶
8.1 Purpose and Limitations¶
Writing samples are collected to extract linguistic style markers (vocabulary diversity, sentence structure, formality register). These markers capture surface-level patterns of language production only. They do NOT reveal the writer's beliefs, opinions, emotions, personality, or inner thoughts.
8.2 Collection Guidelines¶
- Voluntary provision: Writing samples must be voluntarily provided by the participant. Do not require participants to generate writing samples as a condition of participation.
- Topic choice: Allow participants to choose their own writing topics, or provide a selection of neutral topics (e.g., "Describe your daily routine," "Describe a hobby or interest"). Do not request writing on sensitive, personal, or emotionally charged topics.
- Length guidance: Suggest a minimum of 200-300 words per sample. Shorter samples produce unreliable style metrics. Longer samples (500+ words) are preferred for robust analysis.
- Multiple samples: Collect 2-3 writing samples if possible, ideally on different topics or at different times. Multiple samples improve the reliability of style metrics.
- No editing restrictions: Do not restrict the participant's ability to edit, revise, or proofread their writing. Natural writing (with revisions) is more representative than constrained writing.
- Format: Accept plain text (.txt) files. Avoid formats that embed metadata (e.g., Microsoft Word with author information). If the participant writes on paper, transcribe it exactly (including errors, if present).
8.3 Content Restrictions¶
The following types of content must NOT be collected in writing samples: - Personally identifiable information (full name, address, phone number, email, national ID). - Confidential or proprietary information from the participant's workplace. - Content that could identify third parties without their consent. - Content that is illegal, harmful, or violates any law or regulation.
8.4 Storage¶
Writing samples must be: - Stored under the participant's pseudonymous subject ID (never under their real name). - Encrypted at rest. - Accessible only to authorized researchers. - Deleted upon request or at the end of the retention period.
9. Preference Survey Collection Procedure¶
9.1 Purpose and Limitations¶
Preference surveys capture the participant's stated preferences across defined domains. Stated preferences are self-reported and may not reflect actual behaviour (the well-known attitude-behaviour gap). Preference data captures what the participant says they prefer, not necessarily what they actually do.
9.2 Survey Design¶
- Use established instruments: Where possible, use validated preference questionnaires with published psychometric properties (e.g., Need for Cognition Scale, Rational-Experiential Inventory, Big Five Inventory).
- Clear, neutral wording: Questions must be clear, unambiguous, and neutrally worded. Avoid leading questions, double-barrelled questions, or jargon.
- Consistent response format: Use a consistent response format throughout the survey (e.g., 5-point or 7-point Likert scales). Clearly label the endpoints (e.g., 1 = "Strongly disagree," 7 = "Strongly agree").
- Manageable length: Keep the survey short (ideally under 50 items, under 15 minutes). Survey fatigue produces unreliable responses.
- Include attention checks: Embed 1-2 attention check items (e.g., "Select 'Strongly agree' for this item") to identify inattentive respondents.
9.3 Administration¶
- Self-administered: Preference surveys should be self-administered (paper or digital). The researcher should not be present to avoid social desirability bias.
- Private setting: The participant should complete the survey in a private setting where they cannot be observed by others.
- No time pressure: Allow the participant to take as much time as they need. Do not impose time limits on survey completion.
- Right to skip: The participant may skip any question they do not wish to answer. Skipped questions should be recorded as "no response" (not imputed).
- Anonymous within session: Even though the survey responses are linked to the participant's subject ID for profile building, the participant should feel that their individual responses are confidential.
9.4 Domains for Cognitive Emulation¶
The following preference domains are relevant for cognitive emulation:
| Domain | Example Items | Purpose |
|---|---|---|
| Cognitive style | "I prefer to analyze problems step by step" vs. "I rely on my intuition" | Inform decision bias parameters |
| Information density | "I prefer detailed information" vs. "I prefer summaries" | Inform workspace capacity modulation |
| Risk tolerance | "I am comfortable making decisions under uncertainty" | Inform uncertainty adjustment parameters |
| Time-of-day preference | "I am most alert in the morning" vs. "I am most alert in the evening" | Context for proxy interpretation |
| Task preference | "I enjoy puzzles and problem-solving" vs. "I prefer creative tasks" | Inform engagement estimation |
10. Memory Interview Collection Procedure¶
10.1 Purpose and Limitations¶
Autobiographical memory cues are memories that the participant voluntarily shares with the researcher. These cues provide context for the emulation (e.g., areas of expertise, life experiences that may influence cognitive patterns) but they are NOT automatically extracted from the participant — the participant must actively choose to share each cue.
Memory cues do NOT constitute "memory extraction" or "memory reading." The system stores only what the participant explicitly provides.
10.2 Collection Guidelines¶
- Entirely voluntary: The participant must not be pressured to provide any memory cues. This data channel is optional in every sense.
- Participant-initiated: Memory cues should be participant-initiated wherever possible. The researcher may provide prompts (e.g., "Are there any experiences you'd like to share that might be relevant?") but must not dictate what memories to share.
- Structured format: Collect memory cues in a structured format to facilitate processing:
- Short description (1-3 sentences) of the memory.
- Approximate time period (childhood, adolescence, early adulthood, etc.).
- Domain category (education, work, personal, health, etc.).
- Relevance note (optional): why the participant thinks this memory is relevant.
- No probing for sensitive content: The researcher must not probe for or encourage the sharing of traumatic, distressing, or deeply personal memories. If the participant begins to share such content, the researcher should gently redirect and ensure the participant is comfortable.
- Quantity guidance: Suggest 3-10 memory cues as a reasonable range. More cues provide richer context but also increase data volume and privacy considerations.
- Review and confirm: After the participant provides memory cues, allow them to review and modify or withdraw any cue before it is stored.
10.3 Storage¶
Memory cues are the most personally sensitive data channel and require the highest level of protection:
- Encrypted at rest with AES-256.
- Stored separately from other data channels.
- Accessible only to the principal investigator and explicitly authorized personnel.
- Deleted immediately upon participant request (no archival option for this channel without renewed consent).
11. Avoiding Deception and Impersonation¶
11.1 Deception Prohibition¶
The subject-specific emulation system must never be used to deceive anyone — including the subject themselves. Specific prohibitions:
- Do not present the system as the subject: The CIA system, even when conditioned on a subject's profile, must never be presented to any person as "being" the subject, "representing" the subject in an interactive context, or "speaking for" the subject.
- Do not generate deceptive content: The system must not be used to generate text, speech, or other content that could be mistaken for the subject's own communications, unless explicitly labeled as machine-generated and the subject has given specific consent for that use.
- Do not impersonate: The system must never impersonate the subject in any context — not in communications with third parties, not in automated systems, not in research presentations.
- Do not create false relationships: The system must not be used to create the appearance of a relationship with the subject (e.g., pretending to be the subject's friend, colleague, or family member).
11.2 Labeling Requirements¶
When emulation outputs are presented or shared:
- Always label as emulation: Any output generated using a subject's profile must be clearly labeled as "subject-specific cognitive emulation output" or equivalent.
- Never attribute to the subject: Emulation outputs must never be attributed to the subject. Phrases like "the subject said" or "according to the subject's profile" must be avoided if they could be misinterpreted as direct quotes or representations of the subject's views.
- Include caveats: All outputs must include the mandatory caveat from the profile (see Document 17, Section 6.1).
11.3 System Guardrails¶
The safety_validator.py module in the subject emulation extension implements impersonation guardrails:
- Prohibited phrase detection: Scans all outputs for language that could be interpreted as presenting the system as the subject.
- Identity assertion check: Flags any output that uses first-person pronouns in a way that could imply the system is the subject.
- Context awareness: When the system is used in an interactive context, it should prepend a disclaimer: "This is a research system, not [subject's name or any person]. Responses are generated based on statistical patterns, not by a person."
12. Data Storage and Security¶
12.1 Storage Architecture¶
| Data Type | Storage Location | Encryption | Access |
|---|---|---|---|
| EEG raw data | Secure research storage (encrypted NAS/SAN) | AES-256 | Authorized researchers |
| Behavioural traces | Secure research storage | AES-256 | Authorized researchers |
| Writing samples | Secure research storage (separate partition) | AES-256 | PI + authorized analysts |
| Preference surveys | Secure research storage | AES-256 | Authorized researchers |
| Memory cues | Separate encrypted storage (most restrictive) | AES-256 + separate key | PI only |
| Subject profiles (JSON) | Secure research storage | AES-256 | Authorized researchers |
| Identity mapping | Separate encrypted storage | AES-256 + separate key | PI only |
| Consent records | Separate encrypted storage | AES-256 + separate key | PI + ethics liaison |
12.2 Backup Procedures¶
- Backups are encrypted using the same standards as primary storage.
- Backup encryption keys are stored separately from both primary and backup storage.
- Backup access is logged and restricted to system administrators.
- Backup retention follows the same schedule as primary data retention.
12.3 Incident Response¶
If a data breach or security incident occurs:
- Contain the breach immediately (isolate affected systems).
- Assess the scope (which data, how many subjects affected).
- Notify the principal investigator within 1 hour.
- Notify the ethics committee within 24 hours.
- Notify affected participants within 72 hours (or sooner if required by regulation).
- Document the incident, response, and remediation measures.
- Conduct a post-incident review to prevent recurrence.
13. Ethical Review Requirements¶
13.1 Pre-Collection Review¶
All data collection must be reviewed and approved by an appropriate institutional review board (IRB), ethics committee, or equivalent body before any data is collected. The ethics submission must include:
- The study protocol (including this data collection protocol as an appendix).
- The consent form (with all required disclosures).
- A data management plan (storage, encryption, access controls, retention, deletion).
- A risk assessment (physical, psychological, privacy, and social risks).
- A justification for each data channel to be collected.
- A justification for the number of participants and sessions.
- A plan for handling consent revocation and data deletion requests.
- CVs and qualifications of all researchers involved.
13.2 Ongoing Review¶
- Adverse event reporting: Any adverse event (physical discomfort, psychological distress, data breach) must be reported to the ethics committee according to institutional reporting requirements.
- Protocol amendments: Any changes to the study protocol (new tasks, new data channels, changes to consent procedures) must be submitted to the ethics committee for review before implementation.
- Annual renewal: If the study extends beyond one year, submit an annual renewal to the ethics committee summarizing progress, any adverse events, and any protocol changes.
13.3 Post-Collection Review¶
After data collection is complete:
- Submit a final report to the ethics committee summarizing the study, participant flow, adverse events, and data handling.
- Confirm that all data has been stored, encrypted, and access-controlled according to the approved plan.
- Confirm the retention period and set up deletion triggers.
- Archive the study materials (protocol, consent forms, data management plan) for the required period.
14. Template Consent Form Outline¶
The following is a template outline for the informed consent form. It must be adapted to the specific institution, jurisdiction, and study design. Legal review is recommended before use.
[Institution Name] [Department/Laboratory]
Informed Consent Form¶
Study Title: [Full title of the study]
Principal Investigator: [Name, Title, Contact Information]
Ethics Approval Number: [Number]
Version: [Version number and date]
1. Purpose of the Study¶
[Clear, non-technical explanation of the research purpose. Must include the scientific boundary disclaimer: "This study builds statistical profiles of cognitive and behavioural patterns. It does NOT read minds, copy identities, or transfer consciousness."]
2. What You Will Be Asked to Do¶
[Description of each data collection procedure. Must address EEG, behavioural tasks, writing samples, preference surveys, and memory cues separately.]
3. What Data Will Be Collected¶
[Description of each data type, what it measures, and what it does NOT measure. Reference the scientific boundary.]
4. Risks¶
[Description of physical risks (EEG electrode discomfort, fatigue) and privacy risks (data collection, storage). Must state that there are no attempts to read thoughts or access private mental content.]
5. Benefits¶
[State clearly: "There are no direct personal benefits from participating. Your participation contributes to scientific research."]
6. Confidentiality¶
[Description of pseudonymisation, encryption, access controls, and data retention. Must specify the retention period and the participant's right to delete data.]
7. Your Rights¶
- You may withdraw at any time without penalty.
- You may request deletion of all or part of your data.
- You may skip any question or task you do not wish to complete.
- You may revoke consent for any data channel independently.
- You may contact the ethics committee with questions or concerns.
8. Compensation¶
[If applicable. State compensation amount and conditions. If no compensation, state "There is no compensation for participating."]
9. Contact Information¶
- Principal Investigator: [Name, Email, Phone]
- Ethics Committee: [Name, Email, Phone]
10. Consent Statement¶
By signing below, I confirm that:
- [ ] I have read and understood this consent form.
- [ ] I have had the opportunity to ask questions.
- [ ] I understand the purpose of the study.
- [ ] I understand what data will be collected and how it will be used.
- [ ] I understand my right to withdraw at any time without penalty.
- [ ] I understand my right to request deletion of my data.
- [ ] I understand that this study captures statistical patterns, not my mind or identity.
- [ ] I voluntarily agree to participate.
Data Channel Consent (check each separately):
- [ ] I consent to EEG recording (scalp electrical activity).
- [ ] I consent to behavioural data collection (task responses and response times).
- [ ] I consent to writing sample collection (linguistic style analysis).
- [ ] I consent to preference survey completion (stated preference recording).
- [ ] I consent to providing autobiographical memory cues (optional).
Participant Name (print): _________
Participant Signature: _____ Date: _______
Person Obtaining Consent (print): _________
Person Obtaining Consent (signature): _____ Date: _______
References¶
| Resource | Citation / URL |
|---|---|
| CIA Scientific Boundary | Document 00: docs/00_scientific_boundary.md |
| CIA Safety and Ethics | Document 04: docs/04_safety_ethics.md |
| EEG Setup and Safety | Document 13: docs/13_physical_eeg_setup_safety.md |
| Subject-Specific Cognitive Emulation | Document 17: docs/17_subject_specific_cognitive_emulation.md |
| GDPR | Regulation (EU) 2016/679 |
| HIPAA | Health Insurance Portability and Accountability Act (US) |
| Common Rule | 45 CFR Part 46 (US Federal Policy for the Protection of Human Subjects) |
| Belmont Report | The National Commission for the Protection of Human Subjects (1979) |
| BIDS / EEG-BIDS | https://bids.neuroimaging.io |