Abstract
M.A.
This study serves to expand the work of A.C. Faul on scale development in social work to
incorporate the demand for multiculturalism. Ecometrics – the measurement of ecosystems – is a
steadily growing field in South Africa. To date, however, scale development has assumed that the
ecometrics will be practiced in a monocultural context. This is obviously not the case in South
Africa. Consequently, the research goal is to design a process model for the development of social
work scales for multicultural use in South Africa. As a secondary objective, the study aims to test
this model in practice, through the development of a multicultural scale that accurately measures
the social health of military employees/families.
A number of issues underlying the technical aspects of multicultural scale development are first
explored, including issues of the characteristics of ecometrics; the meaning of the term culture;
the emic-etic debate; cultural equivalence; and bias, fairness and standards in ecometrics.
Thereafter, a process model for the development of multicultural ecometric scales is introduced
and five main phases are described: analysis, design, development, evaluation, and diffusion &
adoption. Each of these phases is further decomposed into main moments and steps, each of
which is described at both theoretical and technical levels.
In order to test this process model in the real world, a new multicultural, multilingual,
multidimensional, systems-oriented, salutogenic scale was created, called the Military Social
Health Index. In the analysis phase, the need for the scale was analysed and the innovation
requirements determined and contracted with the client. A theoretical framework – family
resilience theory – was identified and explored, resulting in the development of an assessment
model that underlies the scale. The cross-cultural comparability of the constructs was assessed
and each construct was operationally defined, using facet maps.
In the design phase, a multicultural, multilingual team of social workers generated close to 200
items, as well as instructions, using a multifocus approach, in which items were generated in four
languages simultaneously (English, Zulu, Setswana and Afrikaans). Only items that could be
expressed equivalently across languages were accepted. This resulted in an initial instrument,
comprising 175 items (plus 16 demographic items), covering seven constructs, in four languages,
at an average reading level of Grades 6-7.
During the development (or field testing) phase, the instrument was reviewed by a group of social
workers for content relevance, translation equivalence, item formulation, etc. Thereafter it was
reviewed by focus groups of soldiers in the SANDF. Finally, the instrument was subjected to an
analysis of linguistic equivalence. In response to each review, changes were made to the
instrument.