Measuring Language Proficiency

What can be truly measured? I believe that listing the units by which something is measured is essential.

In language proficiency testing, there are many scales being used. Level 1 can mean many different things. I will use units of length as an analogy. MIT students got this right when they measured in smoots. In 1958, they measured the Mass. Ave. bridge between MIT and Boston to be “precisely 364.4 smoots plus or minus one ear” (see link). The bridge is painted in Smoots to this day. Oliver Smoot, the person they used to measure the bridge, went on to chair the American National Standards Institute.  Here is an April Fools’ joke on the topic, from the MIT site itself.

When we compare language proficiency scales, we should have the same level of concern for accuracy. Have these comparisons been validated? Are the units of the comparison mentioned? Saying “Level X is equal to Level Y” is meaningless unless we name the unit (scale) for level Y, and how many people have tested at the same level on both scales in order to validate this result.

Without these comparisons, it is difficult for clients to have a basis for comparison between one vendor’s marketing materials and another vendor’s marketing materials. Maybe it is time for third-party verification to become a common practice in this system.

Until this happens, at Gaucha Translations we will continue to work with the providers we know to be reliable because of their historical track record and the endorsement of the ACTFL and other such organizations, which have the ability to research the practices of said vendors.