Statistical Linguistic Quality Assessment: Leading the Way
Over a decade ago, OmniLingua Worldwide worked under General Motors’ direction to help establish the SAE J2450 metric for the Automotive industry. This successful endeavor led OmniLingua to drive application of this metric to the Life Sciences industry. In fact, we chaired the Standards Committee through its initial years.
How It Works
The J2450 metric applies rules of Statistical Process Control to compute a weighted, numeric translation quality based on the categories, “wrong term, wrong meaning, omission, structural error, misspelling and punctuation error.” Independent linguists use this metric to assess the quality of our translations, and we share the scores directly with you. Unlike other translation companies that rely solely on internal quality inspections, our J2450 scores tell you exactly the level of quality you are receiving.
How we use SAE J2450 to ensure quality translation service:
- We use it to test incoming translators and continue improvement of their overall quality through periodic performance reviews.
- We perform independent J2450 assessments on random samples of all qualified translation projects using a statistically meaningful sample size, which ensures 95% confidence levels.
- We use control chart analysis with the J2450 data to evaluate the stability and quality levels by client, linguistic team, language and other subsets.
- We reference specific J2450 assessments on projects as part of our root cause investigations to identify correlations of possible areas for improvement.
A key service that OmniLingua performs in order to provide you with confidence in the linguistic quality of your translation and localization projects.