```xml <result> <column_name>Income: $75K–$100K</column_name> <family>additive</family> <subtype>intensive_additive</subtype> <confidence>0.85</confidence> <requires_user_confirmation>false</requires_user_confirmation> <sum_meaningful>false</sum_meaningful> <signed>false</signed> <reducers> <primary_reducer>mean</primary_reducer> <valid_reducers>mean, min, max, median, distribution</valid_reducers> <excluded_reducers>sum (summing percentages across entities is meaningless), product (not a ratio measure)</excluded_reducers> <average_semantics>arithmetic_mean</average_semantics> </reducers> <inner_outer_measure> <inner_measure_reducer>min</inner_measure_reducer> <outer_measure_reducer>max</outer_measure_reducer> </inner_outer_measure> <multiplicative_partner_derivable>false</multiplicative_partner_derivable> <reasoning> Table type is DIMENSION with no sequence axis, making multiplicative highly implausible. The column represents percentage shares of population in a specific income bracket (demographic composition). These are intensive properties: each percentage describes the proportion at that entity, and summing percentages across different geographic entities is meaningless (you cannot add "0.80% from region A" to "1.45% from region B"). The natural rollup is mean (weighted by population if available, otherwise arithmetic). The values are stored as text percentages but represent absolute compositional levels, not ratios or growth factors. </reasoning> <table_type_conflict>false</table_type_conflict> </result> ```