Propuesta para el cálculo del índice dm para la validez de los ítemes

Abstract

The aim of this paper was to propose a program in MS Excel for quick calculation of Dm index, which could be used in psychometrics to evaluate the item validity of psychological instruments. Validity is a process of accumulating evidence to support the interpretation and use of psychological test scores, with scores relations with other external variables an important source of validation. A variant is called criterion-related validity, which seeks to demonstrate the accuracy of the hypothesized relationships between measurements made with other measurements, using correlations or other measures that indicate some kind of relationship. This was considered the Index Dm, which is expected to be significantly correlated with those external criteria that are theoretically relevant, so achieving specify prior, as a hypoth esis, what direction it will take and what the correlation coefficient is expected. Dm index indicates a measure of the absolute deviation of the observed correlations regarding hypothesized values. In this index, we then consider the number of external criteria, the correlation coefficient between the item and the external criterion, the squared distance between the coefficient of correlation and .50 by the sign of the correlation coefficient between the item and the criterion external. In social sciences rarely expect a correlation greater than .50, although this constant researcher can vary depending on the background you have about the test to investigate and the relationship you have with other constructs. The range of this indicator ranges from 0 to 1, with the result 1 when there is perfect and criterion validity is obtained when each observed correlation is hypothesized correlations like. As such, the proper use of these methods and their lack of implementation in traditional statistical packages justifies its construction as a tool for researchers in Psychology, and specifically in Psychometrics, discipline which is common to use correlations as an indicator absolute validity, without considering other conceptual aspects. One advantage of this program is that it provides a quick calculation of a useful procedure when you do not have the number of data required for more sophisticated procedures such as factor analysis. Furthermore, it gives more relevance to the findings of previous form as defined a network of concepts expressed by measurements, which correspond to the instrument you want to work, all expressed by correlations. It exemplified the use of the tool with data presented in research presented at an event, where we explored the validity of the items of School Burnout Inventory (SBI) in a sample of adolescents, using external criteria the complete Magallanes Anxiety Scale (MAS) and State-Trait Depression Inventory (STDI). The latter has four scales: Eutimic State and Eutimic Trait (both positive affect) and Dysthymia State and Dysthymia Trait (both negative affect). In this regard, being a measure of stress burnout is hypothesized positive direction between items of School Burnout Inventory (SBI) and Magallanes Anxiety Scale (MAS), subscales Dysthymia State and Dysthymia Trait (both negative affect) from State-Trait Depression Inventory, and negative between the subscales Eutimic State and Eutimic Trait (both positive affect) from State-Trait Depression Inventory. Data have been entered, the calculation of Dm index from the expression is performed automatically, and finally to the index and the level that has this indicator. You can display the graph to see the differences between the hypothesized and observed correlations. The findings presented by the tool match background research. The main limitation of the tool is that it can be used Microsoft Office must be installed on the computer, so it is recommended to interested researchers expand the scope of this tool to other formats or by some programming strategy that is wider dissemination and outreach.