This study has two objectives: One is to demonstrate the Likert-type response scale generation for the Turkish version of the WHOQOL. The other is to show the effect of level of education of the subjects to translate their perception into a rating on a Visual Analog Scale (VAS). WHOQOL questions have four types of response scales: frequency, evaluation, capacity, and intensity. The WHOQOL cross-cultural response scale methodology is based on the VAS. Fifty-one low-level educated subjects from the initial study population (n = 228) were excluded by applying two special exclusion criteria, because of the reliability problems. Those subjects with 8 years and less education were more likely than those with 11 years' (OR = 0.25%; 95 CI 0.11-0.56) and 15 years' and more education (OR = 0.11; 95% CI 0.04-0.29) to be excluded from the study. After translating anchor points into Turkish, the VAS were prepared for each of the response scales (intensity, capacity, frequency, and evaluation). A list of descriptors—16 for frequency, 15 for capacity and intensity, 16 and 14 for evaluation scales—were compiled from dictionaries and the relevant literature. Each descriptor was placed on a 10 cm fresh line in a random order in every scale. The study subjects were asked to place a mark on a 10 cm line for each descriptor, according to where they think the descriptor lies in relation to the anchor points. In order to select the intermediate descriptors for each scale, mean distances for each descriptor were calculated and target intermediate descriptors were found for each scale by applying WHOQOL response scale methodology. The intra-class correlation coefficients (ICC), which indicate the interrater reliability in the current setting, were calculated in the included study population (n = 177) for each education category separately. ICCs were found as 0.39 for 5 years', 0.46 for 8 years', 0.66 for 11 years', and 0.79 for 15 years' and more education categories. Categories of at least 11 years' education were found to show sufficient interrater reliability. The mean and the variabilities of the target descriptors produced similar results with the initial 15 WHOQOL centers. On the other hand, the obtained results indicate that visual analog scale methodology should be used with caution on subjects educated less than 9 years in the Turkish context.

No abstract