GCRI · Labs
Protocols and research design
Good research is designed before it is run. This page sets out how Labs, a platform of The Global Centre for Risk and Innovation (GCRI), expects protocols to be written: a hypothesis stated in advance, a method chosen to test it, rules for research involving people and an independent review before work starts. These rules apply to every experiment, testbed and challenge on the platform. Information and enquiries are open today.
Protocols and research design in five parts.
State the question and the test before the data.
- 01
Question and hypothesis
What is being tested, and what result would count against the hypothesis.
- 02
Method and controls
The design, the comparator, the variables and the controls, with the reasons for each choice.
- 03
Analysis plan
Assumptions, tests, handling of multiple comparisons and of missing data, decided in advance. Significance thresholds are never the only evidence of importance.
- 04
Stop conditions
When the experiment will be paused or stopped, for safety or for futility.
- 05
Deviations
Any change after work starts is recorded with its reason, not folded silently into the protocol.
Research with people needs extra care.
Informed and voluntary
Participants know what is asked, why, the risks, and that they can withdraw.
Ethics review
Protocols involving people are reviewed by a competent ethics body before recruitment begins.
Minimal personal data
Only the data the question needs, protected and kept no longer than necessary.
Protected and Indigenous knowledge never enters public examples or model training without the consent of its holders.
Independent review before and after.
Reviewer independence is assessed from team, institution, data, environment, funding, conflicts and method. Reviews record approval, required changes or rejection with reasons, and dissent is preserved as a record. A reviewer with a declared conflict never signs the relevant acceptance.
Plan for reproduction from the start.
- Name the data sources, their owners and the permissions for each use.
- Freeze the data cut and record the environment and code version for every run.
- Decide in advance what will be shared, under which licence, and what stays restricted.
- Plan an independent reproduction attempt for any result that will be relied on.
Where protocols live.
Protocols are designed to be written in the experiment templates of the NXS-LAB (research, experiments and simulations), so the protocol, runs and reviews stay linked. The methods they use have canonical pages in the STEM foundation of Labs. Topics are discussed on research protocols.
પ્રશ્નો અને જવાબો.
Must every protocol be preregistered?
Does GCRI provide ethics approval?
Can exploratory work be done?
Where do I send a protocol question?
Bring a method, a dataset or a question.
Information and enquiries are open today. An enquiry receives a reference and names the next step; applications, payments, compute allocations and registrations open only when their team, terms and review capacity exist.
કાનૂની દસ્તાવેજો.
આ સાઇટ પરના નકશા, ડેટા અને આકૃતિઓ સંદર્ભ અને સામાન્ય માહિતી માટે દર્શાવવામાં આવ્યા છે. તે સલાહ, આગાહી અથવા ચેતવણી નથી; માહિતીની મર્યાદાઓ અસ્વીકરણો અને ડેટા સૂચનાઓમાં દર્શાવેલી છે.