In the translation of Clinical Outcome Assessments, care needs to be taken around the presentation of numerical scales, units of measurement and expressions of time. These three elements pose some of the biggest challenges in COA localisation because of the vast differences of interpretation across cultures. Different cultures have varying interpretations of rating scales, use different systems of measurement and perceive time differently. Failure to adapt these properly can lead to misinterpretation of items and inconsistencies of data.

This article examines each of the three challenges of rating interpretation, units of measurement and expressions of time.

Numerical Scales and Rating Interpretation

It is common for COAs to contain numerical scales to measure aspects such as pain levels, quality of life or treatment effectiveness. I wonder if there is much consideration given in COA development to the influence of cultural differences on how study participants use numerical scales.

For example:

·       In some East Asian cultures, people tend not to opt for extreme ratings, e.g. 0 or 10

·       By contrast, in Western cultures, participants are usually comfortable using the full range of the scale, including extreme ratings

·       In Latin American and Southern European cultures, there is frequently a preference for higher values on a scale.

To avoid differences in responses, solutions could be to:

·       Use anchored scale labels so that each point is clearly understood, e.g. “0 = No pain; 10 = Worst Pain”

·       Provide examples to illustrate to participants the scale’s intent

·       In COA development, conduct pilot testing across cultures to make sure the scale functions consistently across cultures.

Units of Measurement

COAs frequently reference weight, height, distance, units of currency and temperature, all of which require adaptation in translation. This includes adapting pounds for kilograms and miles and feet for metric measurements. COAs that have a US source usually present temperatures in Fahrenheit. Units of currency are sometimes used in items to do with mental agility (e.g. To a caregiver in a dementia COA: do they recognise the difference between a cent, a nickel and a dime?), or fine motor skills or mobility: ‘To what extent can you pick a penny up from the ground?’). In some of these cases, I have seen instances where units of currency have not been localised. 

Even when the non-localised terms are understood, such as ‘cent’, it can be off-putting and confusing for people in areas where ‘cents’ are not used. More seriously, however, are the appearance of Fahrenheit, miles, feet, pounds measurements. If a person is not familiar with these measurements, it will not mean much to them. Apart from risking the quality of data elicited from participants where measurements have not been adequately localised, it also demonstrates a lack of respect to the target populations.

Expressions of Time and Recall Periods

There are differences in how different cultures perceive time. This means that recall periods, such as ‘in the past week’ may be interpreted differently where attitudes towards time differ across cultures. Western cultures tend to perceive time in a linear manner, while others see it in a more flexible or cyclical way. Therefore, a phrase like ‘in the past week’ typically refers to the most recent 7-day period in Western cultures. By contrast, in some African cultures, the perception of time is more fluid, which makes fixed recall periods less intuitive. Similarly, a phrase like ‘a few days ago’ in Middle Eastern cultures may not have fixed boundaries, which leads to a variability in interpretation.

A solution would be to use specific timeframes, such as ‘in the past 7 days’. In development, cross-cultural interpretations of recall periods should be user tested.

Time expressions such as ‘yesterday’ or ‘recently’ do not always carry fixed meanings in some cultures. This variability across cultures can lead to inconsistencies in COA responses. In some Asian and African cultures, ‘yesterday’ may be used to refer to a past event beyond the preceding 24 hours.

The translation process needs to pay special attention to expression of time and recall periods. Similarly, the cognitive debriefing process needs to dig deeply into these expressions and not just gloss over them. It should be the in-country debriefer’s task to provide evidence of how these terms are understood by native speakers of the language in question.

Final Remarks

Cultural differences in rating scale use, measurement standards, recall periods and perception of time can all lead to data inconsistencies if not adequately localised.

COA developers need to incorporate cross-cultural validity at an early stage to find out how scales, unit conversions, the phrasing of recall period and expressions of time are used and understood across cultures. Gaining a deeper understanding of these issues and addressing them properly in localisation will lead to more accurate global research findings. However, usually, these issues may only be discovered at the Linguistic Validation stage and that is only if cognitive debriefing is performed to an adequate level. If you are a sponsor or a developer of a COA and these issues have never been reported at the Linguistic Validation stage, then start asking questions of your provider. To discover these issues so late in the COA’s life is a shame, if not futile.

Thank you for reading,

Mark Gibson

Leeds, United Kingdom, March 2025