Level of agreement between self‐administered and interviewer‐administered CPQ8–10 and CPQ11–14

Abstract

The aim of the present study was to assess the psychometric properties and level of agreement between the self-administered and interviewer-administered Child Perceptions Questionnaire (CPQ) for children between 8 and 10 years of age (CPQ(8-10) ) and between 11 and 14 (CPQ(11-14) ) years of age. A randomized cross-over study was carried out, involving 180 children (Group 1 - 90 children between 8 and 10; Group 2 - 90 children between 11 and 14 years of age) in the state of Minas Gerais, Brazil. All children completed both administration modes of the CPQ; half of each group received interviewer-administered mode first [Subgroup A (CPQ(8-10) n = 45) and Subgroup C (CPQ(11-14) n = 45)], and the other half performed the self-administered mode first [Subgroup B (CPQ(8-10) n = 45) and Subgroup D (CPQ(11-14) n = 45)]. Test-retest reliability of each mode of administration was tested on 60 children (30 for CPQ(8-10) ; 30 for CPQ(11-14) ), who were not included in the other analyses. The level of agreement between scores on the self-administered and interviewer-administered versions of the CPQ(8-10) and CPQ(11-14) was established using the intraclass correlation coefficient (ICC). The order of presentation of both instruments was tested considering the four subgroups (A, B, C and D). The calculation of effect size proposed by Cohen (1992) was used to test the clinical significance of the findings. Both the self-administered and interviewer-administered versions of CPQ(8-10) and CPQ(11-14) demonstrated acceptable psychometric properties. Agreement between the administration modes for the CPQ(8-10) and CPQ(11-14) was 0.90 and 0.88 (ICC), respectively. With the exception of the functional limitation subscale, the scores of the subscales and overall score on the CPQ(8-10) were significantly higher in the group of children who responded to the interviewer-administered measure first. With the CPQ(11-14) , statistically significant differences were found only for the emotional well-being subscale. Both administration modes of the CPQ(8-10) and CPQ(11-14) demonstrated satisfactory psychometric properties and a high level of agreement. Although statistically significant differences were observed for oral symptoms, emotional well-being and social well-being, with the first administration of the interviewer-administered version, the effect of the order of administration had small to medium effects on the CPQ scores.