This work should be seen as a pilot study of the effect that the common-sense based statement that one understands and learns best in their best language has on pupils in Swedish schools.

A number of students have been given a mathematical test in the language they use for their normal school activities, and that is not their mother tongue. Some of them (approximately 50%) were led to believe this particular test to be language fair. Their average score on the test has then been compared to the average score on the same test of the remaining 50% students who considered the test a normal one. The difference in performance between the two subgroups has been interpreted with the help of the concept of stereotype threat.

This pilot study showed a trace of the hypothesised best language stereotype threat in a specific group of students and will hopefully serve as a guide for a larger work that could prove (or falsify) the existence of the best language stereo-type threat with statistical certainty, extend its range of applicability to a wider group of students and establish its size as compared to other related factors. A final caveat: This study is focussed on (and relevant only for those) students that perform their school activities in a language different from their mother tongue but that are fully operational in the teaching language.