Vaccine–autism linkAtlas

Position A

Agency cohort tables on childhood vaccines and autism are read as showing similar diagnosis rates across vaccination-status groups, with large schedule-wide effects treated as incompatible with multi-year clinical patterns those tables summarize.

Falsification · This account would be weakened if well-controlled studies showed large, reproducible autism rate differences by vaccination status after confounder adjustment.

• Multiple large cohort and case-control studies report no population-level causal link.
• Health agencies summarize autism research as not supporting vaccine causation.
• Autism diagnosis rate changes track diagnostic criteria and awareness shifts across eras.

• Assumption (moderate): Key terms in the A account are used in a stable operational sense across cited materials.
• Assumption (weak): The cited corpus for the A account is treated as sufficiently complete for comparative evaluation.
• Assumption (moderate): Counter-materials against the A account have been considered when stating the position.

Position B

Agency cohort tables on childhood vaccines and autism are read as showing clustered diagnosis rates near vaccination-schedule windows, with average population-null effects treated as compatible with multi-year clinical patterns those tables summarize.

Falsification · This account would be weakened if timing clusters disappeared in analyses that properly separate diagnosis age from exposure timing.

• Dissenting clinicians emphasize regressive onset near schedule points in case series.
• Component and combination schedule changes are treated as under-studied relative to single-antigen trials.
• Parental reports are treated as signal generation that population averages can miss.

• Assumption (moderate): Key terms in the B account are used in a stable operational sense across cited materials.
• Assumption (weak): The cited corpus for the B account is treated as sufficiently complete for comparative evaluation.
• Assumption (moderate): Counter-materials against the B account have been considered when stating the position.

Position A

Methods guidance on vaccine-outcome studies is treated as privileging population comparisons for average claims, with case chronologies cast as thinner designs when denominators and base rates are the main question in the review corpus.

Falsification · This account would be weakened if validated subgroup biomarkers showed vaccine-triggered autism pathways invisible to cohort designs.

• Rare-outcome causal questions typically require large denominators to separate signal from chance.
• Methods guidance warns that unblinded case series are vulnerable to recall and selection effects.
• Negative population results constrain how large a universal effect could be.

• Assumption (moderate): Key terms in the A account are used in a stable operational sense across cited materials.
• Assumption (weak): The cited corpus for the A account is treated as sufficiently complete for comparative evaluation.
• Assumption (moderate): Counter-materials against the A account have been considered when stating the position.

Position B

Methods guidance on vaccine-outcome studies is treated as under-weighting case chronologies for subset probes, with population nulls cast as thinner designs when uncommon timing clusters are the main question in the review corpus.

Falsification · This account would be weakened if case series signals failed replication under blinded exposure assessment.

• Heterogeneity arguments hold that averaging across genotypes can hide real subgroup effects.
• Detailed clinical chronologies can surface mechanisms later tested at scale.
• Historical drug withdrawals show case recognition sometimes preceded epidemiology.

• Assumption (moderate): Key terms in the B account are used in a stable operational sense across cited materials.
• Assumption (weak): The cited corpus for the B account is treated as sufficiently complete for comparative evaluation.
• Assumption (moderate): Counter-materials against the B account have been considered when stating the position.