STUDY / OBJECTIVE
Set prospective criteria for cases and retain failures and missing observations.
01 / STUDY
Fix the question and outcome definition
Before recording a claim, specify the subject, deadline, outcome space and scoring rule. “Something will change” is difficult to score; a predefined event within a fixed period is more inspectable. Use fictional material for practice rather than collecting sensitive personal information.
02 / STUDY
Retain all outcomes
Keep matches, non-matches, unscorable cases and missing observations. Missingness is neither success nor a reason for silent deletion. Report raw counts and define the denominator. Preserve the original plan when rules change, documenting the reason and separating later exploration from planned analysis.
03 / STUDY
Specify a comparison baseline
A common event can be guessed correctly often. Evaluation therefore needs a baseline, such as always choosing the most frequent outcome. Sample selection, size, information leakage and retrospective rewording affect interpretation. A small teaching table demonstrates documentation, not validation of a mantic system.
WORKED EXAMPLE
Reporting ten fictional records
Ten fictional records contain six matches, two non-matches and two missing outcomes. This demonstrates counting, not an experiment on any system.
Report all categories—6, 2 and 2—out of ten, without dropping failures.
Among eight observed outcomes, six match: 6/8. State separately that two outcomes are missing.
If missing cases are treated first as all non-matches and then as all matches, the total proportion ranges from 6/10 to 8/10. Their actual outcomes remain unknown.
Counts, denominator and missingness are transparent; these proportions have not been compared with a baseline, sampling uncertainty or independent data.
PRACTICE / REFLECTION
Exercises and answers
01Can ten selected successful cases establish a success rate?
Read the explanation
No. The number of all attempts and selection procedure are missing. A complete sample frame, failures, missingness and criteria are needed to define an interpretable proportion.
02How should a claim refined after seeing the outcome be classified?
Read the explanation
Preserve the original and revision time, labeling the revision as retrospective interpretation or exploration. Do not score it as a prior prediction.
A distinction to retain
Preregistration improves transparency but does not guarantee a sound design or correct conclusion; scoring rules also require scrutiny.
Unit output
A prospective scoring plan, an all-outcome table of ten fictional cases and a baseline-and-limitations statement.
REFERENCES
Reading references
What is Preregistration?
Explains planned versus exploratory work; fictional course cases are exercises in documentation.References reviewed: 2026-10-24. Examples, record templates and explanations are editorial teaching material. Fictional inputs are not historical events or efficacy evidence. Citations imply no affiliation.
Personal study record; not a grade, assessment or credential.
Connected study paths
Fate systems: structure and interpretationEditorial study material · 2026-10-24