The participants listened to the 86 synthesized /mVf/ vowel stimuli over good-quality headphones in individual sessions. The participant saw eleven response buttons on a computer screen. The eleven response vowels were exemplified on the buttons by a single keyword each. The keywords were supposed to be well-known to the students: feel /i/, fill /ɪ/, sale /e/, fell /ε/, shall /æ/, null /ʌ/, doll /ɑ/, call /ɔ/, whole /o/, full /ʊ/ and fool /u/. The participants were told to imagine what the vowel in each keyword sounds like, and then to decide which of the eleven response vowels came closest to the vowel sound in the /mVf/ token they had just heard. They indicated the vowel of their choice by clicking on the corresponding response button. The next stimulus was made audible 1000 ms after the response click. Stimulus presentation and data collection were controlled by a computer script written for the ExperimentMFC module in the Praat software (Boersma, Weenink, 2019). For details on the procedure see Van Heuven et al. (2020) and Afshar (2022: 61–62).