office (03) 9820 3006 or 920 3008
fax (03) 9820 3007

Step 9 - Data collection techniques

Data collection techniques

Cultural competence in:

Cultural competence in data collection techniques

Cross-national health research (health policy?

Summary Measures of Population Health (SMPH) would need to be compatible across nations and cultures to serve as comparable measures for global health policy making. To serve as inputs for global SMPH, for reasons of comparison and policy relevance, these instruments need to be cross-culturally relevant and sensitive to measuring comparable constructs across cultures (Sommerfeld et al. 1999).

Acceptability and construction of scales in cross-cultural comparison

Sommerfeld et al. (Sommerfeld, Baltussen, Metz, Sanon, & Sauerborn 1999) warn of the need to check ‘whether questioning with scale response patterns are acceptable and known to a particular culture. Attitude measures such as the Likert and semantic differential scales are 'culture-specific, emic instruments’ which largely depend upon a subject's interpretation of the measures.

In Likert scales, the measure characteristics (number of items, number of scale points) can affect responses, as scale endpoints or midpoints may be avoided for cultural reasons or individuals choose extreme points for that same reason. In Asian cultures, for example, an even number of scale points appears more relevant than an uneven number. In semantic differential scales certain adjective pairs may be considered irrelevant by the respondent.

Utility measures based on statistical or objective (aleatory) probability confront subjective or personal considerations of probability. In populations with low levels of formal education, another problem may arise from using numbers and thus translating lay knowledge, attitude and experience into a single numerical value. A quantified valuation is based on lay or folk interpretations of numbers, a lay epidemiology. Sommerfeld’s question is how far people are able to think in, and use, numbers, i.e., how far is a population’s relative degree of numeracy developed? Does a population understand and accept a numerical expression of risk and uncertainty?’