Therefore, the aim of the present manuscript is to propose a method and an SPSS syntax, for the comparison of quantitative evidences of content validity, by means of confidence intervals for the difference between coefficients V.
Estimacion de intervalo de confianza professional#
university education), between culturally or socioeconomically distinct groups (e.g., Peruvians and South Africans), or according to the degree of thematic or experiential expertise derived from different professional practices (eg, educational psychologists vs. For example, between men and women, between subjects with differentiated years of formal education (technical education vs. This comparison occurs when there is the necessary justification to suppose that the evaluation of the content of the items is associated with the variations related to specific groups. Indeed, the review of previous research suggests that the situation least raised in obtaining quantitative evidence of content validity is the comparison of groups of judges with respect to this type of evidence. The content validity estimated in the previous studies rests on the assumption that the results do not interact with differences between groups of experts however, it is plausible to suppose the opposite if a rigorous analysis of the characteristics of the subjects that served as expert judges is made. In the current content validity reports that used the V coefficient ( Domínguez & Villegas, 2012 Domínguez, Villegas, Yauri, Mattos, & Ramírez, 2012 Freiberg, de la Iglesia, Stover, & Fernández, 2014 Gómez, Sainz de Baranda, Ortega, Contreras & Olmedilla, 2014 García & García, 2013 García, Merino & Valero, 2015 Medrano, Liporace & Pérez, 2014 Merino & Valero, 2014 Palao, Manzanares & Ortega, 2015 Sánchez-Alcaraz & Parra-Meroño, 2013 Vallejo-Medina, Granados & Sierra, 2014), to date, the variability of expert judgment related to a fixed factor that differentiates two (or more) groups of experts has not been raised. This procedure is an important advance to generate intervals in the level 100 (1 - α) % to conclude in the statistical and practical significance of the coefficient V. This method seems to fit very well the estimates of content validity by means of coefficients, since it is reasonable to find negative asymmetric distributions in validity judgments, especially when the instruments have rigorously sampled the content domain of the construct in this situation, validity judgments will present denser distributions at the higher extremes of the rating. The introduction of this method in Spanish language seems to be located in Escurra (1988), and later, Merino and Livia (2009) disseminated the asymmetric confidence interval (CI) approach, developed by Penfield and Giacobbi (2004). Among the methods to quantify the validity of content, obtained through the participation of expert judges, the use of the coefficient V ( Aiken, 1980, 1985), also known as V of Aiken, seems to have increased in psychological research or in fields that involve psychological constructs applied in various disciplines.
0 kommentar(er)
