Building: Pinnacle
Room: 2F-Harbourside Ballroom
Date: 2016-07-02 11:00 AM – 12:30 PM
Last modified: 2016-05-22
Abstract
Introduction
Likert-type scales are commonly used as the base for self-report tests. Thus, response styles have an important role in the scores and validity of the instruments. Response styles are the person differential functioning it can cause variance in total scores without being related to the intended construct. One of the goals of anchoring vignettes is to control responses styles, in order to increase the comparability of the scores.
Objective
The aim of this study was to explore SENNA’s anchoring vignettes in a Brazilian sample of students.
Design
The sample was composed by a total of 24605 Brazilian students from the Rio de Janeiro state. The students fulfilled SENNA 5th (70 items), 10th and 12th (104 items) grades. T-student/ANOVA tests were used to compare between the different reference groups.
Results
A total of 58.2% of them were women, 6.0% were in 5th grade, 58.9% in 10th and 35.1% in the 12th grade. For gender, vignettes scores showed small differences (maximum Conhe’s d = 0.27) appeared in 14 of the 15 vignettes. For grade, differences appeared in 8 vignettes (p>0.05), the post hoc tests showed that those differences were majorly between 5th and 10th, and between 5th and 12th grades. For the region of the state, differences appeared for 14 of the vignettes (p>0.05). For race, differences appeared in 11 of the vignettes (p>0.05).
Conclusion
The gender and the grade had small differences, those appeared especially between Elementary School and High School students. The race and the region of state here the school was localized also showed differences. This brings attention for this references groups when analyzing and using vignettes in SENNA.