Furthermore, corpus research has also revealed that miscollocations
in learner English continue to persist even after many years of learning English (Levitzky-Aviad & Laufer, 2013; Nesselhauf, 2005), with negative transfer or interference from learners' L1 being one of the main factors contributing to the erroneous use of L2 phrases.
A number of studies have focused on the verb+noun miscollocations in the writing of university students of English (e.g.
Although there were some clear cases of miscollocations, especially those involving restricted collocations with a delexical verb, acceptability seems to be a matter of degree, since there are combinations that are not frequent and sound non-native, but we cannot categorically state that they are unacceptable.
The problems with signalling nouns reported in Flowerdew's study help to explain some miscollocations involving the verb achieve, as in the examples below:
The delexical verbs most frequently involved in miscollocations in the learner corpus are make and do.
Specifically, they might not be able to differentiate collocations from non-native-like word combinations, that is, miscollocations
or false collocations (Barnbrook, Mason, & Krishnamurthy, 2013; Wray & Perkins, 2000).