¬¬Five Reasons Why Translation Quality Measurements Are Important

Posted by Alexandra on August 24th, 2022

Translation quality measures refer to a language service provider’s (LSP’s) use of metrics to determine the quality of translation output. Translation agencies are required to develop robust quality control systems to qualify for independent certifications awarded to LSPs with a proven track record of delivering quality translation and of other services.

For businesses looking for a language company to collaborate with, they need to look out for relevant quality certifications held by LSPs, such as ISO 17100:2015. This ISO standard is awarded to language service providers who meet the internationally recognised standard requirements for translation services.

This article explains why translation quality measures are so important and should by default provided by your language service supplier.

#1: An objective model for quality assessment

By developing a reliable system of translation quality metrics, language service providers can offer an objective model for assessing the quality of translation output. For example, agencies often use the Multidimensional Quality Metrics (MQM) framework, which specifies a Quality Score metric and a scoring algorithm based on an extensive range of issue types, including:

• Accuracy: The target text does not accurately reflect the source text, allowing for any differences authorised by specifications.
• Design: There is a problem relating to design aspects (vs. linguistic aspects) of the content.
• Fluency: Issues related to the form or content of a text, irrespective as to whether it is a translation or not.
• Locale convention: The text does not adhere to locale-specific mechanical conventions and violates requirements for the presentation of content in the target locale.
• Style: The output text has stylistic problems compared to the original.
• Terminology: A term (domain-specific word) is translated with a term other than the one expected for the domain or otherwise specified.
• Verity: The text makes statements that contradict the world of the text.

This is an example of a scoring mechanism that allows language service providers to objectively measure the quality of their translation output.

#2: Minimum benchmarks for every project

Implementing a data-driven system for quality assessment allows language service providers to set minimum benchmarks for every project. For example, the MQM framework discussed in the previous section defines four categories for translation issues:

1. None: Items that need to be noted for further attention or fixing, but which should not count against the translation – i.e. “preferential errors”.
2. Minor issues: Do not impact usability or understandability of the content, even if an alternative translation is considered more “correct”.
3. Major issues: They have an impact on usability or understandability of the content, but which do not render it unusable.
4. Critical issues: These render the translated content unfit for use.

From here, an agency might set minimum thresholds that allow for a certain percentage of issues flagged as “none” and, potentially, a smaller percentage of issues flagged as minor with a minimum Quality Score results for all content.

This benchmarking guarantees an objective level of quality for all content.

#3: Track performance over time

With a system capable of measuring the translation quality of output content, language service providers are able to track the performance projects over time. This is particularly important for long-term projects where customers require a consistent level of quality over an extended period of time or a large body of work.

#4: Flag quality declines

By tracking performance over time, agencies can quickly identify any drops in quality and put measures in place to restore standards. This could happen for a variety of reasons, such as the agency expanding, hiring new language experts, adopting new technology and a range of other variables that can impact output content.

In most cases, any negative impact is negligible and can be rectified without affecting the quality of output content. The key point here is that, with the right quality process, potential impact can be measured and interventions can be made before any issues affect the quality of the output.

#5: Identify opportunities for improvement

Aside from identifying decline in quality, a robust translation quality measurement system helps language service providers to optimise translation processes to improve results. For example, when testing new software solutions designed to make internal processes more efficient and when setting specific goals, such as reducing the time it takes to produce output content that meets quality requirements.

With quality metrics clearly defined, the result of such changes can be measured in terms of translation quality, whilst ensuring the implementation of solutions designed to have only a positive impact.

To learn more about translation quality measurement and about the QA processes we follow at translate plus, please visit the translate plus website.