Methods to identify differential Item functioning in health context. A systematic Review

Erika Margarita Arias

 

The standards of regulation in health interventions have generated in researchers and practitioners, a growing interest in the design and use of tests that are adequate for the collection of information on aspects, such as, the quality of life, level of functioning, emotional state, and adjustment to the health-disease process of patients. Additionally, epidemiological studies are become interested on the quality of these instruments as support in diagnosis, monitoring and investigation of several illness.

Differential Item Functioning, DIF (Holland & Thayer, 1988) is a technical approach to identify the presence of systematic error in the measurement of latent characteristics or constructs, when we applied the same test in different populations or groups.  The DIF´s presence in the items can seriously affect the validity of the tests when they are to be used in different populations, leading to the measurement being affected by other variables, skills, or traits that are not the object of measurement.

Under what is known as ” DIF third generation” (Zumbo, 2007), the use of methods for identification of differential item functioning in applied studies is characterized by approach from an ecological model in which it is included the effect of environmental factors on the processes that are involved in responding to the items, with the management of these factors (Zumbo , et al, 2015).

In this sense, this study seeks to make a systematic review to identify these contextual factors that can produce DIF and are of special interest in massive scale measurements such as those performed in transcultural epidemiological studies.