ABSTRACT
AIM: Construct stability over time is required for reliable inference, but evidence regarding the longitudinal invariance of negative symptoms is still limited. Thus, we examined the longitudinal invariance of the negative dimension using the positive and negative syndrome scale (PANSS) in an antipsychotic-naïve first-episode schizophrenia sample at baseline and after 10 weeks. METHODS: Our study was conducted at a specialized early intervention service. PANSS ratings were analysed for 138 patients, and two different models were specified and tested: a unidimensional and a two-correlated factor solution. RESULTS: The unidimensional model fulfilled criteria for longitudinal invariance, whilst the two-correlated did not. CONCLUSION: Our study provides support for the PANSS negative unidimensional model use to evaluate negative symptoms' longitudinal change following first-episode schizophrenia.