Community & Knowledge Hub
Practical guidance for questionnaire authors, field teams and data reviewers, from translation, calculations and household rosters to high-frequency checks, back-checks and nonresponse coding.
- EditorialWritten by the GlassFisch team on established survey and fieldwork practice.
- ReviewedEach entry names its source and is reviewed before publication.
- Community QuestionsModerated questions from registered users are coming soon.
GlassFisch is being prepared for a controlled pilot. Articles describe the product as documented today, and what is available can differ between environments and organizations. Each article shows its availability.
Showing 8 of 8 entries
Designing high-frequency checks for survey data
What to check while fieldwork is running, how to set thresholds you can defend, and why every flag is a signal for review rather than proof of an error.
Fieldwork quality control: back-checks, spot checks and supervision
How survey teams verify interviews: accompanied interviews with respondent consent, unannounced spot checks, back-checks in person or by phone, and fair follow-up.
Not asked, don't know, refused or missing? Coding nonresponse in survey data
Why survey data must distinguish questions that were not asked from answers that are missing, don't know or refused, and how to design and code each one.
Changing a questionnaire during fieldwork: version control principles
How to change a questionnaire after fieldwork has started without losing comparability: numbered versions, stable codes, change logs and version-aware analysis.
Calculations in survey questionnaires: blanks, text and rounding
Why a survey calculation comes out empty or wrong: unanswered inputs, numbers stored as text, zero denominators, early rounding and when the formula is evaluated.
Required questions and range checks: preventing errors without forcing bad data
How to use required answers, hard limits and soft warnings so that interviewers catch real mistakes without forcing answers or pressuring respondents.
Designing household rosters that analyse cleanly
How to design household rosters and repeat groups: clear membership rules, line numbers, stable person IDs, counts that match and data linked to the household.
Managing survey translations: a practical workflow
A product-neutral workflow for multilingual surveys: stabilize the source, translate as a team, keep codes identical, review after edits and test each language.
Availability labels
- In development
- Documented in the current GlassFisch product and being prepared for the planned controlled pilot. Not generally available yet.
- Planned controlled pilot
- Intended for organizations in the planned controlled pilot, once it begins. Not generally available.
- Partial
- Only part of what the article describes is supported so far. The article says which parts.
- Planned
- Not built yet. Details may change.
- General guidance
- Product-neutral survey and fieldwork practice that does not depend on a GlassFisch feature.
Community Questions
Registered GlassFisch users will be able to ask practical questions about survey design, fieldwork and data quality. Every question will be reviewed by a moderator before anyone else can see it.
Questions cannot be submitted yet, and nothing on this page is a member post. Sign-in, secure storage, moderation and community guidelines come first.