This paper presents a general inference engine based on interval type-2 fuzzy logic systems (IT2 FLS) to contrast statistical hypothesis on means. Fuzzy statistics is an useful approach which offers new answers for imprecise and uncertain statistical problems. In this way, we propose an extension of the Type-1 FLS reported by Figueroa and Soriano to an Interval type-2 FLS for involving uncertainty in the definition of the FLS. The proposed inference engine attempts to use uncertain hypothesis tests to obtain a fulfilment degree of Ho and after using the IASCO type reduction algorithm, we obtain a crisp information criteria about the hypothesis being tested, useful for many analysts who need to get information about samples.