LQE stands for Linguistic Quality Evaluation, which can be defined as the process where external reviewers linguistically assess translations by completing a scorecard. They are usually carried out in a dedicated platform, where reviewers include any changes along with comments when required.
As we mentioned, LQEs are usually carried out by “external reviewers,” i.e., those that are hired by clients in order to evaluate the quality of the Language Service Provider (LSP) work.
LQEs are crucial since they provide an overall assessment of a translation to measure its quality in the most consistent and objective possible way. Moreover, they give translators clear and structured feedback. In the process, reviewers include justifications for the errors pointed out. If the translator disagrees, he/she can refute the error by providing the reasons why he/she considers that the correction is inaccurate or that the error category (preferential, minor, major, etc.) for the change is unfair.
The scope of reviews can be either full review or sampling/sample review:
Generally, the review scope is determined by:
Once the review process has finished, the translation is given a score: Pass or Fail. Although the Pass percentage varies, it is usually between 97% and 98.5%. What is more, there may be different score categories: Pass/Pass Strong or Fail/Fail Hard.
Besides score categories, there are different labels to refer to error categories. Such labels may differ from one client to another, but they usually include:
Depending on the degree of penalty, there may be different levels of error severity, that go from highly visible errors that cause considerable usability issues or mistakes that might confuse or mislead the reader, to those that don’t affect the meaning of the text but that would lessen stylistic quality or fluency.
Nowadays, there are dedicated platforms where LQEs can be carried out. In said platforms, we can usually find the name of the project, its identification number (that helps to easily find the review), the translation score, the reviewer’s name, the time he/she took to review the text, comments related to the corrections made by the external reviewer and some space for the LSP to accept or rebut the changes.
Once the review process is finished, translators are given the opportunity to rebut any corrections they deem inaccurate or unfair in what is frequently called the ‘rebuttal process’. This stage is key since the LSP can comment on the changes made and request a second round of review. As a result, the reviewer may agree and change the category of some corrections to “preferred/neutral” and thus improve the final score, which might change a failing score to a passing one.
If a failing score is obtained, clients can send the project to ‘rework’ so that the corrections are implemented in the delivered bilingual files, and they can also request that the revisions are implemented globally in the entire file or in all files (for cases where the LQE consisted of revising only a percentage of the translation).