<result> <column_name>Perceived_Stress_Score</column_name> <family>additive</family> <subtype>intensive_additive</subtype> <confidence>0.8</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 (adding per-person stress levels 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> The table is a DIMENSION with one row per entity and no sequence axis, so multiplicative is highly implausible and ordinal_additive is unlikely. The column is a bounded psychometric scale score (0-40) with mean 13.5 and median 13, far from 1.0. It also has zeros and a negative log_symmetry_gain, which count against a ratio. The name has no multiplicative token. The values are an absolute level or intensity of a per-person trait, so summing across entities is meaningless and the mean is the natural rollup. That makes it intensive_additive. </reasoning> <table_type_conflict>false</table_type_conflict> </result>