Ethical Standards

AI Prompt Templates

Copy a prompt into Claude, ChatGPT, or Gemini. Paste your document at the bottom and run.

Paste a document and get a scored quality assessment with evidence and revision priorities.

6,175 characters
You are an expert in research and evaluation ethics. Score the ethical dimensions of the deliverable I will provide using the rubric below. The deliverable may be a MEL plan, evaluation ToR, inception report, survey, FGD/KII protocol, sampling plan, evaluation report, case study, or any deliverable involving human participants or sensitive data.

SCORING RUBRIC - Ethical Standards
Score each dimension 1-5 using these criteria:

DIMENSION 1: Informed Consent Architecture
- Score 5: All four elements present. All applicable consent elements present and operationalized for the deliverable type (purpose, voluntariness, right to refuse or withdraw, confidentiality terms, contact information, recording consent, data use permissions). Context-appropriate format chosen (verbal vs. written, accessible language, literacy considerations). Consent process documented (when consent is obtained, by whom, how it is recorded). Consent for special populations addressed where relevant (children with parental and assent, persons with cognitive disabilities, displaced populations, illiterate respondents).
- Score 4: At least three of four elements present. Consent elements operationalized; context fit or special populations partial.
- Score 3: Standard consent elements named but not operationalized for the specific deliverable. Format or special population considerations missing.
- Score 2: Generic consent statement. No operationalization, no special population considerations.
- Score 1: No consent architecture described.

DIMENSION 2: Confidentiality, Privacy, and Data Protection
- Score 5: All four elements present. Data handling protocols specified (where data is stored, who has access, transmission security). Anonymization or pseudonymization plan for analysis and reporting (specific methods, who can re-identify if anyone). Retention and disposal policies stated (how long data is kept, what is destroyed when). Data breach response or incident reporting addressed.
- Score 4: At least three of four elements present. Data handling and anonymization clear; retention or breach response partial.
- Score 3: Confidentiality stated as a principle but operational details thin. Storage or anonymization mentioned without specifics.
- Score 2: Confidentiality promised in passing. No data handling specifics. No retention plan.
- Score 1: No confidentiality, privacy, or data protection considerations.

DIMENSION 3: Do-No-Harm and Risk Mitigation
- Score 5: All four elements present. Risk assessment for participants conducted (what could go wrong, for whom, how severe). Procedures for sensitive disclosures specified (what to do if a participant discloses abuse, trauma, illegal activity, or distress during data collection). Cultural sensitivity considerations addressed (language, gender of interviewer, religious or social context, taboo topics). Referral pathways defined (where participants can seek support if needed, how the team facilitates referral).
- Score 4: At least three of four elements present. Risk assessment and disclosure procedures present; cultural fit or referrals partial.
- Score 3: General "do no harm" commitment stated. Some risk considerations addressed but referrals or disclosure procedures weak.
- Score 2: Generic do-no-harm language with no specific assessment or procedures.
- Score 1: No do-no-harm or risk consideration.

DIMENSION 4: Power, Positionality, and Conflicts of Interest
- Score 5: All four elements present. Researcher or evaluator positionality acknowledged (who we are relative to participants, the program, and stakeholders). Power dynamics between researcher and participants addressed (with concrete strategies, e.g., language register, physical setup, validating refusals). Conflicts of interest disclosed (relationship with implementer, donor, government, or program staff that could affect findings). Mitigation strategies for identified COIs named (independence safeguards, external review, separation of roles).
- Score 4: At least three of four elements present. Positionality and power addressed; COI disclosure or mitigation partial.
- Score 3: Positionality or power dynamics mentioned but not operationalized. COI disclosure boilerplate or absent.
- Score 2: Power dynamics or COI noted as concepts but not addressed. No mitigation.
- Score 1: No consideration of power, positionality, or conflicts of interest.

DIMENSION 5: Ethical Oversight and Accountability
- Score 5: All four elements present. Ethical approval obtained where required (IRB, ethics committee, government clearance, sectoral protocol) or explicit reasoning for why approval is not needed. Accountability mechanisms specified (how participants or stakeholders can raise concerns, who handles complaints, what response timelines are committed). Compliance with named ethical codes or standards (CESS, AEA Guiding Principles, OECD-DAC, donor-specific ethics requirements). Documentation trail for ethical decisions (what was considered, what was decided, why).
- Score 4: At least three of four elements present. Approval and accountability addressed; codes or documentation trail partial.
- Score 3: Ethical approval mentioned but unclear if obtained. Accountability mechanism vague. No specific codes referenced.
- Score 2: No ethical approval claim. Accountability mechanism absent.
- Score 1: No ethical oversight or accountability described.

OUTPUT FORMAT:
Return your assessment as a table followed by a summary:

| Dimension | Score (1-5) | Evidence from Document | Priority Revision |
|-----------|-------------|------------------------|-------------------|
| Informed Consent Architecture | | | |
| Confidentiality, Privacy, and Data Protection | | | |
| Do-No-Harm and Risk Mitigation | | | |
| Power, Positionality, and Conflicts of Interest | | | |
| Ethical Oversight and Accountability | | | |

**Total: X/25**
**Band:** Strong (22-25) / Adequate (17-21) / Needs Revision (11-16) / Substantial Revision (5-10)
**Single Most Important Revision:** [One specific sentence]

For any dimension scored 1 or 2, add a brief explanation and a concrete revision example.

DOCUMENT TO SCORE:
[Paste your deliverable here]

Scoring Criteria

Informed Consent Architecture
5Excellent

All four elements present. All applicable consent elements operationalized for the deliverable type. Context-appropriate format chosen. Consent process documented. Special populations addressed.

4Good

At least three elements. Consent operationalized; context fit or special populations partial.

3Adequate

Standard consent elements named but not operationalized. Format or special population considerations missing.

2Needs Improvement

Generic consent statement. No operationalization.

1Inadequate

No consent architecture described.

Confidentiality, Privacy, and Data Protection
5Excellent

All four elements present. Data handling protocols specified. Anonymization plan defined. Retention and disposal stated. Breach response addressed.

4Good

At least three elements. Data handling and anonymization clear; retention or breach response partial.

3Adequate

Confidentiality stated as principle but operational details thin. Storage or anonymization without specifics.

2Needs Improvement

Confidentiality promised in passing. No data handling specifics.

1Inadequate

No confidentiality or data protection considerations.

Do-No-Harm and Risk Mitigation
5Excellent

All four elements present. Risk assessment conducted. Procedures for sensitive disclosures specified. Cultural sensitivity addressed. Referral pathways defined.

4Good

At least three elements. Risk and disclosure procedures present; cultural fit or referrals partial.

3Adequate

General do-no-harm commitment stated. Risk considerations addressed but referrals or disclosure procedures weak.

2Needs Improvement

Generic do-no-harm language with no specific procedures.

1Inadequate

No do-no-harm or risk consideration.

Power, Positionality, and Conflicts of Interest
5Excellent

All four elements present. Positionality acknowledged. Power dynamics addressed with strategies. COIs disclosed. Mitigation strategies named.

4Good

At least three elements. Positionality and power addressed; COI disclosure or mitigation partial.

3Adequate

Positionality or power dynamics mentioned but not operationalized. COI disclosure boilerplate or absent.

2Needs Improvement

Power dynamics or COI noted but not addressed. No mitigation.

1Inadequate

No consideration of power, positionality, or COI.

Ethical Oversight and Accountability
5Excellent

All four elements present. Ethical approval obtained or reasoned exemption. Accountability mechanisms specified. Compliance with named codes. Documentation trail.

4Good

At least three elements. Approval and accountability addressed; codes or documentation partial.

3Adequate

Ethical approval mentioned but unclear if obtained. Accountability mechanism vague. No codes referenced.

2Needs Improvement

No ethical approval claim. Accountability mechanism absent.

1Inadequate

No ethical oversight or accountability described.

Score Interpretation

Total (out of 25)BandNext Step
22-25StrongEthical standards are robust. Use as-is or with minor refinements.
17-21AdequateAddress flagged dimensions before fielding. Most likely fix: tighten data protection specifics and add referral pathways.
11-16Needs RevisionSubstantial revision required. Use Revise prompt to identify and fix ethical gaps before any participant contact.
5-10Substantial RevisionEthics section fails the threshold for participant-facing or sensitive-data work. Rebuild with consent architecture and do-no-harm risk assessment as the foundation.