The evaluation of multilingual pupils is more complex than the evaluation of monolingual students. Previous research (Ewijk, 2011) shows that ethnic minority students perform poorer in school when they are taught by teachers belonging to the ethnic majority. The reason for this remains unclear, but Ewijk (2011) does suggest that teachers report lower expectations and unfavourable attitudes that both likely affect their behaviour towards minority students, potentially inducing them to perform below their ability level. Effects of having ethnic majority teachers on minority students’ grades hence seem more likely to be indirect than direct. Note that high expectations may encourage high results, while low teacher expectations discourage pupils (ibid.).
Similarly, Hajer (2003: 26) described the so-called Pygmalion effect:

“I often hear teachers say ‘they cannot work in groups, they are too non-independent’. But if you come up with a well-prepared task and you give it a try, it turns out not to be that bad. Anyone who thinks that a pupil who does not talk much is not capable of much, will ask less from that pupil. This way, they get fewer opportunities to learn through interaction and the prophecy fulfils itself. If instead you understand that language is something that continues to evolve and that stammering is useful, you can break through that idea. If you radiate high expectations on pupils it also affects their well-being and level of motivation, making them want to participate more actively” (Rosenthal & Jacobson, 1968).

Furthermore, both teachers and board members should keep in mind that most of the standard evaluations are based on the target language that the pupil does not yet fully master. This is still the case if the evaluation regards numeracy skills: Le Pichon & Kambel (2016) present a study which focuses on the connection between language skills and numeracy and shows that the results of a maths test depend on the language in which it is taken. In other words, while a teacher may think that a pupil got a low-grade due to the fact that he or she is not good at mathematics, the pupil may simply not have developed sufficient language skills (yet). Therefore, pupils should be tested for math skills in the school language and in their home languages before a diagnosis can be made.

Moreover, Le Pichon & Kambel (2016) recommended to adapt the test culturally as it may contain elements from the welcoming culture that are not known to the newly arrived pupils. Indeed, there is a chance that the evaluations are not only linguistically but also culturally biased. Malda et al. (2008) show that a cultural bias in tests keeps those tests from measuring a pupil’s full potential. For example, when trying to remember a series of numbers a pupil is not only dependent on memory, but also on the number of syllables. In other words, language can have an impact on the outcome of a cognitive test. This is something that needs to be kept in mind when using Western-oriented cognitive tests for pupils who have a non-Western background.