Author(s):
Fernandes, Ananda Maria 
; Oliveira, Armando Mónica ; Batalha, Luís Manuel da Cunha ; Grácio, Joana Catarina Gonçalves ; Viegas, Ricardo
Date: 2013
Origin: Repositório Científico da Escola Superior de Enfermagem de Coimbra
Subject(s): anxiety; pain; child; scale;
Description
A functional look at the concept validity of the CAPS's pain and anxiety faces scales.
Introduction and Aims
The CAPS is a set of two face scales, composed of 5 drawn faces each, for evaluating pain intensity and anxiety in children. The goal of the present study was to assess whether the two sets of faces actually address distinct constructs. Rather than looking at convergent validity, it did so by examining the rules whereby the inner features of faces in each set are combined by subjects while issuing pain and anxiety judgments.
Methods
The study rested upon the methodology of Information Integration Theory. Faces in each set were divided into upper- and lower half-faces (grossly, "eyebrow-eyes" and "nose-mouth"), which were factorially combined to produce 25 pain- and 25 anxiety-faces.
Two groups of children (9-11 years old) participated in the main study: children without a regular experience of pain ("pain-free"; n=23), and children undergoing a post-operatory period ("acute-pain"; n=21). Both groups evaluated the pain faces as to pain intensity and the anxiety faces as to anxiety/fear in two separate tasks. The faces were randomly presented on a computer screen and answers given on a 600 pixels horizontal graphical rating scale.
An additional group of pain-free children (n=20) took part in an ancillary study, in which the judgment dimension was reversed between the two sets of faces (pain judged from "anxiety faces", an anxiety/fear from the "pain faces).
Results
Parallel plots and a statistically null interaction were found for the "anxiety faces". Rightward convergence in the factorial plots and a significant interaction were found for the "pain faces". These outcomes were replicated in both groups.
Functional measures derived from the integration models revealed that "eyes-eyebrows" got the most importance in the pain set, while "mouth-nose" got the most importance in the anxiety set. This was replicated in both groups.
Reversing the judgment dimension (ancillary study) didn't change the integration rules or the patterning of relative importance in neither group.
Conclusions
A robust difference in the way facial features were combined, as in their estimated relative importance, was found among the two sets of faces. This difference cannot be accounted by the judgment dimension, as shown in the ancillary study. No differences among the "pain-free" and "acute-pain" groups were found as regards these findings. Outcomes do support the claim that the CAPS' anxiety and pain faces scales address distinct constructs.
Acknowledgments: Work supported by grant PTDC/PSI-PCO/107910/2008, from the Portuguese Foundation for Science and Technology
