Copie un prompt en Claude, ChatGPT o Gemini. Pegue su documento al final y ejecute.
Pegue un documento para obtener una evaluación de calidad con puntuación, evidencia y prioridades de revisión.
6,100 caracteres
You are an expert M&E data specialist with deep experience designing and reviewing survey instruments. Score the response options in the survey I will provide using the rubric below. Focus only on the answer choices attached to each question, not on the question wording itself, the introduction, the skip logic, or the overall instrument structure.
SCORING RUBRIC - Survey Response Options Quality
Score each dimension 1-5 using these criteria:
DIMENSION 1: Exhaustiveness
- Score 5: All elements present. Response options cover the full range of plausible answers for each closed question. An "other (specify)" or "none of the above" option is included where the closed list cannot reasonably anticipate every case. Categorical lists (occupations, languages, geographies) reflect the actual population. No question forces a respondent into an inaccurate category for lack of an option.
- Score 4: Most elements present. Closed lists are nearly complete; one or two questions miss a plausible category or omit an "other" option where one is warranted.
- Score 3: Closed lists cover the main cases but leave gaps in several questions. "Other" is used inconsistently across analogous questions.
- Score 2: Multiple closed questions force respondents into inaccurate categories. "Other" is missing where the population is heterogeneous.
- Score 1: Absent or inadequate. Response lists routinely fail to cover plausible answers across many questions, so the data will systematically misclassify respondents.
DIMENSION 2: Mutual Exclusivity
- Score 5: All elements present. Response options are non-overlapping for every closed question. Numeric ranges have clean boundaries (e.g., 0-9, 10-19, 20-29, not 0-10, 10-20). Categorical options do not duplicate the same underlying meaning under different labels. No question forces a respondent to choose between two equally valid options.
- Score 4: Most elements present. Options non-overlapping in nearly every case; one or two questions have an overlapping range or a near-duplicate category.
- Score 3: Mutual exclusivity holds for the majority of questions but is broken in several places (overlapping ranges, near-duplicate categories, or ambiguous boundaries).
- Score 2: Multiple questions have overlapping options that will force inconsistent respondent choices and create coding ambiguity.
- Score 1: Absent or inadequate. Response options overlap routinely, making the data unanalyzable without recoding.
DIMENSION 3: Scale Balance
- Score 5: All elements present. Likert and rating scales are balanced (equal numbers of positive and negative options). A neutral midpoint is included where appropriate to the construct. Label spacing is roughly equal across the scale. Scale direction is consistent across analogous questions (or deliberately reversed for response-set checks, with a note).
- Score 4: Most elements present. Scales balanced and consistent; one scale uses an unusual direction or omits a midpoint where one would help.
- Score 3: Scales are workable but unbalanced in one or two places (more positive than negative options, missing midpoint, or inconsistent direction).
- Score 2: Multiple scales are unbalanced (e.g., three positive options against one negative option), biasing responses toward agreement.
- Score 1: Absent or inadequate. Scales are systematically unbalanced, with no symmetry between positive and negative endpoints across the instrument.
DIMENSION 4: Label Clarity
- Score 5: All elements present. Option labels are clear, parallel in structure across the scale (e.g., "Strongly disagree, Disagree, Neutral, Agree, Strongly agree"), and unambiguous in meaning. Labels are concrete rather than abstract where possible (frequency labels name a period, agreement labels are verbal not numeric only). Two analysts would code the same option the same way.
- Score 4: Most elements present. Labels clear and parallel for nearly every question; one or two scales mix verbal and numeric labels or use one ambiguous term.
- Score 3: Labels are interpretable but not parallel across the scale. Some labels are abstract ("somewhat", "very") without anchoring.
- Score 2: Multiple labels are ambiguous or non-parallel. Respondents will interpret the same option differently.
- Score 1: Absent or inadequate. Labels are inconsistent or unclear across the instrument, undermining comparability.
DIMENSION 5: Don't-Know and Skip Handling
- Score 5: All elements present. "Don't know", "prefer not to say", and "not applicable" options are offered where appropriate to the question. These options are distinguished from substantive responses (e.g., set off visually, coded separately). Their use is consistent across analogous questions. Their role is documented for analysts (treated as missing, excluded from denominators, or reported separately).
- Score 4: Most elements present. These options offered in most appropriate cases; treatment in analysis briefly noted.
- Score 3: "Don't know" or "not applicable" offered inconsistently across analogous questions. Analyst treatment not stated.
- Score 2: These options are missing where the question demands them (e.g., factual questions with no escape route), forcing respondents to guess or skip.
- Score 1: Absent or inadequate. No "don't know", "prefer not to say", or "not applicable" anywhere in the instrument. Respondents will fabricate answers or break off.
OUTPUT FORMAT:
Return your assessment as a table followed by a summary:
| Dimension | Score (1-5) | Evidence | Priority Revision |
|-----------|-------------|----------|-------------------|
| Exhaustiveness | | | |
| Mutual Exclusivity | | | |
| Scale Balance | | | |
| Label Clarity | | | |
| Don't-Know and Skip Handling | | | |
**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 list the specific questions where the problem appears, with corrected response options.
SURVEY TO SCORE:
[Paste your survey here]
Scoring Criteria
Exhaustiveness
5Excellent
Options cover the full range. "Other (specify)" or "none of the above" used where needed. Categorical lists reflect the population.
4Good
Closed lists nearly complete; one or two questions miss a category or omit an "other".
3Adequate
Closed lists cover main cases but leave gaps in several questions. "Other" used inconsistently.
2Needs Improvement
Multiple questions force respondents into inaccurate categories. "Other" missing where the population is heterogeneous.
1Inadequate
Lists routinely fail to cover plausible answers across many questions.
Mutual Exclusivity
5Excellent
Options non-overlapping. Numeric ranges have clean boundaries. No near-duplicate categories.
4Good
Non-overlapping in nearly every case; one or two questions with an overlapping range or near-duplicate category.
3Adequate
Holds for the majority of questions but broken in several places.
2Needs Improvement
Multiple questions with overlapping options that force inconsistent choices.
1Inadequate
Options overlap routinely; data unanalyzable without recoding.
Scale Balance
5Excellent
Equal positive and negative options. Neutral midpoint where appropriate. Consistent direction across analogous questions.
4Good
Balanced and consistent; one scale with unusual direction or missing midpoint.
3Adequate
Workable but unbalanced in one or two places.
2Needs Improvement
Multiple unbalanced scales, biasing responses toward one end.
1Inadequate
Systematically unbalanced across the instrument.
Label Clarity
5Excellent
Labels clear, parallel across the scale, and unambiguous. Concrete anchors where possible.
4Good
Clear and parallel in nearly every question; one or two scales mix verbal and numeric labels or use one ambiguous term.
3Adequate
Labels interpretable but not parallel across the scale. Some abstract terms without anchoring.
2Needs Improvement
Multiple ambiguous or non-parallel labels.
1Inadequate
Labels inconsistent or unclear across the instrument.
Don't-Know and Skip Handling
5Excellent
"Don't know", "prefer not to say", and "not applicable" offered appropriately. Distinguished from substantive responses. Analyst treatment documented.
4Good
Offered in most appropriate cases; analyst treatment briefly noted.
3Adequate
Offered inconsistently across analogous questions. Analyst treatment not stated.
2Needs Improvement
Missing where the question demands them.
1Inadequate
No "don't know", "prefer not to say", or "not applicable" anywhere.
Score Interpretation
Total (out of 25)
Band
Next Step
22-25
Strong
Response options are ready for fielding. Pair with question wording, skip logic, and pretesting work to complete the instrument.
17-21
Adequate
Tighten the flagged response-option sets before fielding. Most likely fix: add "other (specify)" or rebalance one Likert scale.
11-16
Needs Revision
Substantial revision required. Use the Revise prompt to fix exclusivity, balance, and missing-option problems before any pilot.
5-10
Substantial Revision
Response options will produce unreliable data as designed. Rebuild option sets using the Generate prompt against the question list, then re-score.