A latent trait approach to measuring HIV/AIDS related stigma in healthcare professionals: application of mokken scaling technique

Researchers have described HIV/AIDS related stigma across populations and across domains of interpersonal interaction [1, 3, 18, 23, 24, 28]. Of the many forms of HIV/AIDS related stigma that have been described, one of forms with the greatest potential for lasting harm is the stigma by healthcare professionals towards people living with HIV/AIDS (PLWHA). The negative attitudes compromise the quality of care to PLWHA, and can affect the willingness of PLWHA to access health settings in which they are the subject of stigmatising responses from staff [6, 20, 29].

Dealing with the attitudes of [future] healthcare professionals is central to the management of this form of HIV/AIDS stigma [13, 43]. World Medical Association – declaration of Geneva – clearly lays the foundation for the non-stigmatising attitudes and behaviors expected of the healthcare professionals: “I will NOT permit considerations of age, disease or disability, creed, ethnic origin, gender, nationality, political affiliation, race, sexual orientation, social standing or any other factor to intervene between my duty and my patient” [42]. The management strategies to ensure non-stigmatising attitudes and behaviors may include post-qualification training or the integration of the stigmatising attitude issues in the educational curriculum of healthcare professionals – regardless of their ethnicity, nationality, race, religious beliefs, etc. – during their initial training. The relative merits of these strategies are subject to empirical investigation, and there is no reason to believe that they are not complimentary. However, whichever strategy is adopted, the measurement of change in stigmatising attitudes is a key to the assessment of the effectiveness of the intervention. There is therefore a strong case for robust measures of HIV/AIDS related stigma, developed for [future] healthcare professionals. Monash University, because of its campuses in various parts of the world, provided a good opportunity to develop a HIV/AIDS stigma scale to measure stigmatising attitudes of healthcare students of the Australian and the Malaysian campuses. The admission criteria of the health programs are identical in both campuses. Therefore, the students’ pool consisted of the individuals who deemed to have the same levels of intellectual abilities, but coming from different social and cultural backgrounds.

A number of HIV/AIDS stigma scales have been developed to measure stigmatising attitudes towards PLWHA [2, 15, 34]. The approach to scale development has tended to rely on classical test theory, and assumed that each item (question) measured the true score (level of stigma) with error for each person [32]. Good items to include in a stigma scale were selected on the basis of their pooled reliability, or in combination with Principal Component Analysis (PCA) according to their loading on a single dimension [22]. The approach makes assumption about the normality Gaussian nature of the distribution of the responses to each item.

Mokken scale analysis (MSA) takes a different approach. It is a nonparametric hierarchical scaling technique related to Guttman scaling, and falls under the umbrella of nonparametric item response theory (IRT) [32, 40]. The point of departure from classical test theory is the underlying assumption that the probability of a person responding in a particular way to an item depends on their personal latent trait (i.e., how stigmatising their attitudes towards PLWHA actually are), and on the characteristics of the item itself (i.e., how demanding or difficult an item is in terms of eliciting a negative response towards PLWHA) [9]. Thus, MSA orders people according to their probability of responding in a stigmatising manner (i.e. their latent trait) – the monotone homogeneity (MH) assumption. It also orders items according to the probability of being answered in a stigmatising manner independent of the person answering the question – the double monotonicity (DM) assumption. If the MH and DM assumptions both hold, then a Mokken scale is established that can order people along a latent trait of stigmatising attitudes and order the items in the scale on their “difficulty”. Mokken scales also make no distributional assumptions about the underlying data, other than that the data are capable of being ordered by item and by person.

The advantage of MSA is that it can be used to develop unidimensional cumulative scales that are usually shorter than scales developed using other approaches, while retaining acceptable psychometric properties [27, 32, 33]. Recently, Nyblade et al. hinted at the lack of brief, simple and standardised tools measuring HIV/AIDS-related stigma as one of the barriers to scaling up stigma reduction programs in health services [19]. Thus, Nyblade and her colleagues developed an 18-item measurement tool out of which five of its items meant to measure attitudes towards PLWHA [19]. Usually it takes more time and resources for creating shorter measurement tools that would retain their acceptable psychometric properties. For instance, the HIV-Knowledge Questionnaire was a 45-item measurement tool [4] that was made shorter – 18 items – for the ease of administration while retaining its psychometric properties [5]. The Brief HIV-Knowledge Questionnaire was found to be suitable for use in clinical, educational and public health settings [5]. Moreover, some of the commonly used HIV/AIDS measurement tools might have been decontextualized as the dynamic nature of HIV/AIDS stigma is under constant change [19]. AIDS Attitudes Scale (AAS), for example, was first developed in 1992, using classical test theory approach [12]; and was further validated in 1997 [10]. Since then, except for the development of an alternative form of the scale for use in general public [11], the scale has not undergone further validations. While AAS has strong psychometric properties [12], it only measures ‘empathy’ and ‘avoidance’ as the two domains of AIDS-related stigmatising attitudes. Recently developed HIV/AIDS measurement tools tend to emphasise on other domains such as HIV positive individual’s rights to fair treatment by their family members and the members of the society [19, 36].

The aim of this study was to develop a short measure of HIV/AIDS related stigma, applying Mokken scale analysis technique, for use among healthcare professionals (in training) that had sound psychometric properties.