Hypothesis
Definition: A testable statement linking a proposed mechanism or relation to an observable outcome.
Why it matters here: In Capstone Proposal Design, Hypothesis is needed to reason about hypothesis, sample, method, power, risks, schedule, and data management rather than merely label the phenomenon.
Relationship: It is interpreted alongside Sample design and the lecture relation below; identify which quantities are observed, calibrated, or model-derived.
Limitation or common confusion: Compression should remove redundancy, not validity-defining caveats.
\(utility=expected information/(time+risk)\)
Sample design
Definition: The rules defining targets, comparison groups, size, and selection effects.
Why it matters here: In Capstone Proposal Design, Sample design is needed to reason about hypothesis, sample, method, power, risks, schedule, and data management rather than merely label the phenomenon.
Relationship: It is interpreted alongside Statistical power and the lecture relation below; identify which quantities are observed, calibrated, or model-derived.
Limitation or common confusion: Compression should remove redundancy, not validity-defining caveats.
Statistical power
Definition: The probability that a planned test detects an effect of specified size when it is present.
Why it matters here: In Capstone Proposal Design, Statistical power is needed to reason about hypothesis, sample, method, power, risks, schedule, and data management rather than merely label the phenomenon.
Relationship: It is interpreted alongside Risk register and the lecture relation below; identify which quantities are observed, calibrated, or model-derived.
Limitation or common confusion: Compression should remove redundancy, not validity-defining caveats.
Risk register
Definition: A ranked record of technical or scientific failure modes, consequences, and mitigations.
Why it matters here: In Capstone Proposal Design, Risk register is needed to reason about hypothesis, sample, method, power, risks, schedule, and data management rather than merely label the phenomenon.
Relationship: It is interpreted alongside Data-management plan and the lecture relation below; identify which quantities are observed, calibrated, or model-derived.
Limitation or common confusion: Compression should remove redundancy, not validity-defining caveats.
Data-management plan
Definition: A plan for formats, metadata, access, preservation, sharing, and reproducible analysis products.
Why it matters here: In Capstone Proposal Design, Data-management plan is needed to reason about hypothesis, sample, method, power, risks, schedule, and data management rather than merely label the phenomenon.
Relationship: It is interpreted alongside Hypothesis and the lecture relation below; identify which quantities are observed, calibrated, or model-derived.
Limitation or common confusion: Compression should remove redundancy, not validity-defining caveats.