;   Medical Translation Insight: Why measure translation quality? - ForeignExchange Translations

Why measure translation quality?

Why measure translation quality?Everybody talks about quality: Oh, we deliver great quality! We don't sacrifice on quality. We're all about quality. The problem with these statements is that they're impossible to verify and thus are meaningless.

Some of that cannot be avoided; language is inherently subjective. What may be a perfect translation to one person may be incorrect for the next person.

The subjectivity comes into play when we use words like "perfect" or "wrong". How often have you been in situations where somebody tells you "this is a bad translation – it reads as if it was done by a machine"?

Usually, when you dig deeper, it comes down to a particular turn of phrase or a word that should have been used differently. So, now you are arguing about whether the text should use "car" or "automobile", or maybe the active voice or the passive voice. In the absence of clear guidelines, it is very difficult to resolve these kinds of arguments.

It is easy to see where these disagreements come from. Each of us is a product of our unique background. Inevitably, we evaluate the world around us in the context of our experiences and beliefs regarding culture, society, and language.

Translation, at its core, aims to achieve equivalence between the source and target texts, yet that equivalence is difficult to achieve.

WHAT IS QUALITY?
I know, I know - we all think that we can spot poor translations. But are we as adept at recognizing a good translation?

In order to answer "yes" with any kind of conviction, we need to overcome the subjectivity mentioned earlier and form a common understanding of what quality actually is.

ISO 8402 defined quality, in part, as "The characteristics of a product that bear on its ability to satisfy stated and implied needs".

Translation quality really is the fulfillment of the requirements that have been defined and agreed upon with the client.

Quality exists at every level of the clinical translation process:

  • The quality of a price proposal is the prime criterion for a client to award a project.
  • The quality of processes determine how well projects are executed and complex tasks managed.
  • Etc.

QC ALONE IS NOT ENOUGH
Every once in a while, I hear people say "We don't need quality measurements – we already perform thorough quality-control steps".

In fact, most translation processes include multiple QC steps. The same is true for medical writing, printing, design work, and so on. So, why do we still have quality problems? And why is there no firm sense of how good the quality is?

The same is true for quality metrics. In and of themselves, they don't improve quality. However, when used as part of a quality-improvement system, they play an important role in:
  • Increasing productivity
  • Reducing costs
  • Enhancing customer satisfaction
  • Improving communications
  • Increasing product safety
So, if you are tired of arguments about "car" vs. "automobile", track objective quality metrics!

Like this post? Subscribe to Medical Translation Blog via email or RSS!

6 Comments:

  1. Anonymous said...
    I found this approach very interesting. Although not believable for those who still think that their translations are better than others and are not willing to accept it.
    Changing views and methods; and renewing yourself are implied in this and if this does not change, new generations of translators would not be considered for any kind of job which requires a high standard "quality".
    Anonymous said...
    Quality these days seems antithetical to the standard ways agencies now work. The projects are always rushed and priced low. I receive requests for thousands of words in complex medical articles and detailed medical reports, including illegible handwriting, to be turned around in 24 hours or less. Quality is out of the question under these conditions.
    Renato Beninatto said...
    Congratulations on the blog!

    The quality discussion is a long one. What matters at the end is what the client wants and expects. We wrote Buyer-Defined Translation Quality here at Common Sense Advisory based on interviews with 24 large scale buyers of language services in multiple industries. Two key findings associated to the discussion so far: (1) The buyer trusts the LSP for quality; (2) On-time delivery trumps quality.

    As to price, another long discussion. But the fact is that there are companies that charge $0.42 per word for German-English translation, and there are companies that charge $0.10 for the same job. The ones that charge more are perceived to provide better quality.
    Anonymous said...
    Renato: "on-time delivery" trumps quality. Well, let's turn it around: If the customer is unable to:
    1. schedule the translation phase (so that the best and not the "available" resources can be booked)
    2. prepare the translation phase (eg. terminology, "clean documents", etc.)
    3. foster the translation phase (quick turnaround of queries, etc.)
    THEN
    "on-time delivery" _sinks_ quality.
    80 per cent of my work is done under those circumstances.
    So: Instead of measuring the translation quality, let's measure the "project quality" :-) (yeah, I know, the customer is always right, but, no, they very often aren't and they don't listen).
    dtishin said...
    All the components of translation service quality can be very rarely correctly evaluated by the client (especially when translating from the client's mother tongue). What the client can measure very well is the pure service components (meeting deadlines, friendliness of the PM etc.) and the appearance (layout) of text. The other components (i.e. meaning, terms, style, grammar, spelling, typos, regional standards) are less obvious.
    This is why translation quality means not only compliance with clients's style guides and glossaries (which is obligatory because client pays money), but also compliance with the rules of the language and existing practice of usage in the subject field.
    Practice shows that all of these quality criteria can be successfully implemented into a translation quality metric system that helps measure quality as objectively as possible.
    We at our translation company use 3 different systems (ATA Framework for Standard Error Marking, SAE J 2450 and a proprietory system, which took us about 6 months to develop). There are error classes and types with definitions, requirements, guidelines and procedures for translation quality testers, forms for putting in results etc.
    In January 2009 we also made a research comparing the three translation quality metric systems mentioned above (now it is available in Russian only but I think we will translate the Powerpoint report into English too).
    The result is, for a year we have been measuring translation quality both at the stage of recruiting ant testing translators into project teams (80-90% translators get a "Fail" TQI<86 acc. to our system), and at the stage of delivery to client (selective (batch) translation quality measurement).
    Vic Uzumeri said...
    As several prior comments have suggested, it is very difficult to enforce 'quality' in an inherently subjective arena. It is even harder to find someone who is really qualified to make a fair quality assessment.

    That is one reason that I argue for greater use of visual, graphical (even video) elements in cross-cultural communications. For example, there may be nuances that are lost while watching a foreign-language film with subtitles. However, it would be much harder to understand the story if you ONLY had the subtitles.

    New imaging, video and web technologies promise to make it much easier to add visual elements to even the most technical materials.

Post a Comment





 

Services | Resources | Company | Contact Us | Blog | Home

(c) Copyright 2010, ForeignExchange Translations, Inc.