ABSTRACT
Objective
This study aimed to study the characteristics of the items of Brazilian Loneliness Scale UCLA (UCLA-BR) and define its the cutoff points to the Brazilian context by the Item Response Theory.
Methods
It was used the Politonic Graded Response model on the 20 items of UCLA-BR answered by 1,438 participants in six states in Brazil. The discrimination capacity and item difficulties were estimated as well as the scale cutoff points.
Results
UCLA-BR showed to have good capacity to provide information about loneliness. Sixteen items had high or very high discrimination capacity and four showed moderate discrimination. The cutoff points were defined.
Conclusions
UCLA-BR showed good psychometric indexes. The Item Response Theory made possible the analysis of psychometric qualities of instruments.
Loneliness; Item Response Theory; cutoff point; UCLA-BR