Questionnaires about mental health are often translated from English and used around the world on the assumption that they measure the same thing everywhere. A study from Malaysia suggests that this assumption does not always hold.
The study, published in The Open Psychology Journal, was led by Vie Cheong Thong with colleagues at Universiti Malaysia Sabah in Kota Kinabalu. It tested the Malay version of the Mental Health Knowledge Schedule, a short questionnaire created by Evans-Lacko and colleagues in 2010 to measure knowledge about mental health and stigma-related attitudes.
Mental health literacy, the knowledge and beliefs that help people recognise, manage, or prevent mental disorders, is described by the authors as a key factor in reducing stigma and encouraging timely help-seeking. Measuring it, they note, is fundamental to developing effective education and anti-stigma programmes.
The researchers surveyed 203 university students aged 18 and over who could read both Malay and English. Each completed the Malay version, then a measure of psychological mindedness, and finally the original English version, after a pre-test with 20 bilingual students. They checked how consistently the questions worked together, whether the expected structure held up, and whether scores related to a similar measure.
The results were poor. Both the knowledge and attitude sections showed very low internal consistency, and many individual questions did not fit well with the rest.
Statements that were worded in reverse, where agreeing indicates less knowledge, performed especially badly. Some of these items showed no meaningful link, or even a negative link, with the concept they were supposed to measure.
Statistical tests of the questionnaire’s structure also suggested a poor fit. There was only partial support for the questionnaire relating as expected to another measure of interest in mental health.
The authors suggest several reasons. These include ambiguities in the language, cultural differences in how mental illness is understood, and overlap between knowledge and attitudes, which the questionnaire treats as separate.
Reverse-worded items appear particularly vulnerable to misunderstanding when translated. The researchers also argue that the questionnaire’s structure may not be stable outside Western cultural settings.
Rather than confirming the questionnaire, the study sets out the barriers to using it reliably in Malaysia. The authors say this points to wider problems with directly adapting scales from one culture to another.
The study has limits, including a modest sample of university students. The authors also acknowledge that asking people to complete both language versions may have introduced priming effects, with answers to one version shaping answers to the other.
The findings suggest that before judging how much a population knows about mental health, researchers need to be confident that translated questions carry the same meaning in a new language and culture.
