Discussion

Because individual departments are often charged with teaching multiple languages, the survey did not constrain respondents to describing the process for any individual language and, in fact, most respondents checked several boxes in response to the question of for which language the survey is valid. This multiplicity of languages coupled with the format of the survey which allowed for answers of "Other" for almost every question lead to a situation in which it became difficult to determine what weight to give any single answer. In other words, if a person had indicated that the survey covered French, Italian, and German, and that reading, writing, grammar, and speaking were assessed for placement purposes, it could be the case German placement was based primarily on reading, while French and Italian were based on the other skills. Though some respondents included marginal notes to the effect that a particular answer only covered a particular language, most of the surveys did not include such notation. Since there was no principled way disambiguate the responses, it was impossible to get detailed language specific information (e.g. Japanese placement most often includes reading whereas French placement most often includes grammar). For this reason, no differentiation between placement procedures and languages can be made.