Separate a crude association from matched comparison evidence

Define the host evidence

Host condition, exposure history and sampling can differ between comparison groups. A crude association can therefore mix within-stratum response differences with unequal group composition. Stratification and standardization make a comparison population explicit, but neither automatically removes every confounder or establishes a causal protective mechanism.

Connect mechanism and inference

The original low/high-baseline risks are 0.1/0.4 in group A and 0.05/0.3 in group B. B has a lower supplied risk within each stratum. If the high-stratum fractions are hA and hB, crude risks are 0.1(1−hA)+0.4hA and 0.05(1−hB)+0.3hB. The labels are fictional and do not represent an actual treatment or vaccine.

Using equal low/high weights for both groups yields standardized risks 0.25 and 0.175, difference−0.075. The target weights define an estimand; different legitimate target populations can give different averages. Unmeasured differences, selection and inaccurate outcome measurement can still undermine causal interpretation.

Community and immune-response evidence should be connected to the mechanism actually measured. A correlation between a marker and lower crude risk is not proof that changing the marker would cause protection. This record supports denominator and alternative-explanation reasoning, not medical advice or an efficacy claim.

Reconstruct the original record

At hA=0.2, hB=0.8, crude A risk 0.16 and B risk 0.25 give B−A=+0.09, reversing the within-stratum ordering. The equal-weight standardized contrast remains 0.175−0.25=−0.075. The arithmetic reversal arises from different stratum composition.

Predict and investigate

Predict first: Can a group have lower risk in both strata but higher crude risk overall?

Reconstruct both crude mixtures and the fixed-standard comparison. Name the target population and one unmeasured factor that would keep this from being a randomized causal estimate.

Defend a changed case

Apply the mechanism to a changed boundary

Use equal high-stratum fractions 0.5 in both groups. Find crude risks and compare with the standardized contrast.

Retain the first response, units, uncertainty and any prior feedback exposure in the notebook before comparing this public explanation.

Compare a worked response

The crude risks equal the half/half standardized risks 0.25 and 0.175, difference−0.075. Agreement follows from matched weights, not proof that every causal assumption is satisfied.

Retain bounded evidence

Preserve original assumptions, first responses, source identity and feedback exposure in the Microbiology evidence workspace. These are fictional nonclinical records, not personal medical advice, actual patient data, diagnosis, treatment or an invitation to collect biological samples. Actual practical and independent learner competence require separate evidence.

Primary mechanism context, checked 31 October 2026. Original teaching and numerical records; no source clinical cases or procedures copied.