```xml <result> <column_name>Trust (Government Corruption)</column_name> <family>additive</family> <subtype>intensive_additive</subtype> <confidence>0.92</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 trust scores across entities is meaningless), product (not a ratio)</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> Gate prior: DIMENSION table with no sequence axis makes multiplicative highly implausible. The column is a trust/corruption score - an intensive measure of institutional quality at the entity level. Values are bounded [0, 0.505], centered away from both 0 and 1 (mean=0.138, median=0.105), representing a level or intensity rather than an extensive quantity. Summing trust scores across countries is meaningless; averaging is the natural aggregation. No multiplicative signals present (not centered on 1.0, no ratio naming). Classified as intensive_additive with high confidence. </reasoning> <table_type_conflict>false</table_type_conflict> </result> ```