IDENTIFYING THE DIRECTIONS OF THE RELATIONSHIPS OF ASSOCIATION BETWEEN THE CONSTRUCTS AND THEIR INDICATORS: AN EMPIRICAL CASE

• DOI: 10.22533/at.ed.2162110126

• Palavras-chave: Confirmatory tetrad analysis; causal indicators; reflective indicators; composite variable; latent variable.

• Keywords: Confirmatory tetrad analysis; causal indicators; reflective indicators; composite variable; latent variable.

• Abstract:

  There are two types of models for measuring a construct. A construct is a latent variable when the measurement indicators are influenced by it. In this case, the indicators are called reflected or effect indicators. On the other hand, a construct can be called a composite variable when it is the indicators that condition its behavior. These indicators are called formative or causal. There is disagreement in the literature about the nature of indicators for measuring various constructs. Furthermore, in most empirical work, indicators are assumed to be reflective. The direction of the linear relationship between indicators and their constructs influences the parameter estimates of structural models. An empirical study with categorical data is used to assess the direction of linear relationships. Although the theoretical framework of some constructs used advocates the use of causal indicators, tests of statistical significance pointed out that all indicators in the model are reflected.

• Número de páginas: 26