**3. Results**

### *3.1. Phase 1: Cross-Cultural Adaptation of the RIFL-P*

Three key adaptations were made during development of the Brazilian Portuguese RIFL-P version. First, the term "parent" was replaced by "caregiver". Second, in Portuguese, "mind reading" was translated as "thought reading" (*leitura do pensamento*) in order to preserve conceptual equivalence. Third, definitions of the scoring rubrics of the adapted RIFL-P version were expanded to include a specific rubric for score 3 ("Sometimes true/I partially agree") and descriptions of scores 1 and 5 of the Likert scale gained one more expression (i.e., 1 = "Not at all true/I totally disagree"; 5 = "Very true/I totally agree"). The adapted RIFL-P coding sheet and manual, in Brazilian Portuguese, are presented as Appendices A and B respectively.

#### *3.2. Phase 2: Testing the New Measure*

#### 3.2.1. Coder Training

Additional guidelines for Brazilian coders (i.e., examples of culture-specific behaviors) were elaborated for five of the 11 items to adapt the measure, both culturally and psychometrically, for use with the local population. Three items were coded infrequently (1, 3, and 8), meaning that what was average for those behaviors for Canadian mothers was not average for Brazilian mothers. In order to ensure the underlying normal distribution on all items in the adapted RIFL-P measure, the scoring rubrics for those three items were softened in the Portuguese Brazilian version.
