Formative or Reflective: On the Nature of the OHIP Items

Objectives: The OHIP-49 is an instrument for measuring Oral Health Related Quality of Life (OHRQoL). Because this is presumed to be a latent variable, the OHIP-49 is often analyzed as if it were based on a reflective measurement model with the underlying assumption of local independence, which requires the OHIP-49 to consist solely of items that reflect OHRQoL. This assumption is untenable, however, because the OHIP-49 also contains items that describe symptoms or side effects. Clearly, as these are determinants of OHRQoL rather than reflections, they underpin a formative measurement model and not a reflective measurement model. The objective is to distinguish the reflective items from the formative items in the OHIP-49. Methods: A method described by Fayers et al. (1997) was used, which requires a general QoL score to which each item of the OHIP-49 can be cross-tabulated. Participants, who were randomly selected for an epidemiological study conducted by the Netherlands Organisation for Applied Scientific Research (TNO) and by the Academic Centre for Dentistry Amsterdam (ACTA), received both the OHIP-49 and a single QoL-item. Results: Data was gathered from 675 participants. Preliminary results show that most items in the OHIP-49 are formative in nature. Conclusions: Given the formative nature of the OHIP items, the assumption of a reflective measurement model and therefore local independence is rebutted. This implies that methods such as factor analysis and Cronbach's alpha are not appropriate for the analysis of OHIP data.