Skip to content
Your knowledge of your community has valueGoloka ContributorsContributor Sign In
Goloka
Sign inBook a Call

Trust & Methodology

Provenance,
quality control.
Know why to
trust the data.

Before we collect anything, we agree with you what valid evidence looks like, how it should be collected and how it will be checked.

Talk to our teamExplore how Goloka works
A Goloka field contributor with an evidence record documenting methodology, permission, quality checks and provenance

Methodology that fits the question

Not every study needs
the same design.

We agree on the research approach around what you are actually trying to learn.

Who

Target population

Define the people, groups or participants the study needs to reach.

Where

Geography

Set the locations, communities or areas the research should cover.

Selection

Sampling approach

Decide how participants, locations or observations should be selected.

Questions

Research instrument

Design the questions, prompts or tools used to collect the required evidence.

Evidence

Evidence requirements

Define the information or supporting proof that needs to be captured.

Validation

Quality controls

Set the validation and review checks used to assess the quality of submissions.

Permission

Consent

Ensure participants understand the study and consent matches the intended use.

Output

Delivery format

Agree how the final data or findings should be structured and delivered.

Collection

Fieldwork approach

Define how and when evidence should be collected.

Quality assurance

We build quality checks
around risks in the study.

The right combination depends on what you are collecting.

A field contributor recording responses on a tablet. We define the quality standard upfront, then validate each submission against what the study requires.

We define the quality standard upfront, then validate each submission against what the study requires.

  • Participant qualificationConfirm the right people take part.
  • GPS checksVerify where data was collected.
  • Timestamp checksConfirm when submissions were made.
  • Media reviewCheck photos, video or audio evidence.
  • Validation rulesFlag incomplete or inconsistent responses.
  • Duplicate detectionIdentify repeated or copied submissions.
  • Human reviewReview evidence beyond automated checks.
  • RechecksVerify unclear or questionable submissions again.
  • Specialist reviewAdd expert review where needed.

Consent

Taking part in one study does not automatically mean someone has agreed to every future use of their data. Repeat research, AI-related data and other uses are handled according to the permissions required for that engagement.

A participant’s consent covers the current study; repeat research, AI evaluation and AI training or reuse require permissions matched to each use

Data provenance

Understand where the
data came from.

Where relevant, we document how the information was collected, when it was collected, its source and the quality checks applied.

Illustrative observation record GLK-02418: field contributor, Kaduna, collected 12 March 2026 at 10:42 using structured observation, GPS and photo evidence, then human reviewed, accepted and delivered

Responsible fieldwork

No data point is worth
unnecessary risk.

We consider the safety of contributors and respondents when designing fieldwork. Where a location or subject creates unnecessary risk, we change the approach or tell you the assignment is not feasible.

A Goloka researcher reviewing fieldwork plans and evidence in an office

Safe to collect. Can appropriate fieldwork proceed?

Use the agreed method, safeguards and evidence requirements to collect responsibly.

Safer methods. Is physical collection appropriate?

Where fieldwork creates unnecessary risk, use remote interviews, existing data, local respondents or alternative evidence.

Not feasible. Should the assignment proceed?

We adapt the research method to the level of risk so evidence can be collected safely and responsibly.

Do not deploy where the risk cannot be reduced to a level that allows the work to proceed safely and responsibly.

Reach out to our team

Have a
methodology or
data-governance
question?

Tell us the geography, issue and type of change you want to monitor.

Talk to our team