## Relation to sufficient statistics::Completeness (statistics)

### ::concepts

''X''::''g''    Which::first    ''s''::function    Lehmann::random    ''p''::''t''    Scheff::''t''

Relation to sufficient statistics For some parametric families, a complete sufficient statistic does not exist. Also, a minimal sufficient statistic need not exist. (A case in which there is no minimal sufficient statistic was shown by Bahadur in 1957. {{ safesubst:#invoke:Unsubst||date=__DATE__ |\$B= {{#invoke:Category handler|main}}{{#invoke:Category handler|main}}[citation needed] }}) Under mild conditions, a minimal sufficient statistic does always exist. In particular, these conditions always hold if the random variables (associated with Pθ ) are all discrete or are all continuous.{{ safesubst:#invoke:Unsubst||date=__DATE__ |\$B= {{#invoke:Category handler|main}}{{#invoke:Category handler|main}}[citation needed] }}

Completeness (statistics) sections
Intro  Definition  Relation to sufficient statistics  Importance of completeness  Notes  References

 Relation to sufficient statistics PREVIOUS: Definition NEXT: Importance of completeness << >>