Stimulus Set: Auditory and Visual Primes

Visual primes in this study were selected based on aspects of daily lives that are more relatable to the average Singaporean, and revolve around themes such as neighbourhood icons, food, and clothing. Another search strategy revolved around cultural festivals, celebrations or practices that individuals would have experienced during their lives in Singapore (e.g., Chinese New Year, Deepavali, National Day, etc.). These primes are screened by three independent individuals from the respective ethnic groups (i.e., Chinese, Malay, Indian) to ensure that the images do not contain any offensive content, and that they do not conflate cultural and religious imageries as far as possible. All visual primes have been sourced online. A majority are from photo depositories such as Flickr, Unsplash, and Canva, although a minority of the images have been sourced from search engines such as Google and Wikipedia, as well as websites of media and news articles. Nonetheless, all images have been checked to ensure that they do not violate any copyright regulations. All images are presented in greyscale and cropped so that faces are either not presented or obvious in the images. Please contact the researchers for a full list of their sources, and the raw (unprocessed) files. Auditory primes are selected based on a series of criteria and practical considerations. In line with the presentation of visual primes (i.e., 2 seconds), auditory primes must not exceed 2 seconds (±20%, giving allowance for the length of the words). Similar to the visual primes, they must not contain offensive material and should therefore be of neutral valence. To further standardise the auditory primes in terms of language structure, the following criteria were also introduced: (1) each auditory prime must contain in its sentence or clause a subject, object, and verb (i.e., “I (subject) eat (verb) chicken (object)”; (2) each auditory prime must be in active voice and not passive voice (i.e., “I eat chicken” (active) and not “The chicken was eaten by me” (passive)); (3) each auditory prime must be understood as a statement and not a question to minimise irrelevant cognitive associations should the participants mentally respond to the spoken question. Auditory primes were first selected from published language corpuses online. These corpuses are free to access and feature, at least in part, conversational spoken language. Specifically, Singlish is sourced from the National Speech Corpus, Mandarin from the open-source Mandarin Chinese Conversational Speech Corpus and Mandarin Chinese Speech Corpus for TTS - Daily Use Sentence by Beijing Magic Data Technology Co., Malay from the Malay Multilingual Spoken Corpus by Tokyo University of Foreign Studies and the open-source Malay Conversational Speech Corpus by Beijing Magic Data Technology Co., and Tamil from Microsoft Speech Corpus (Indian languages) by Microsoft Research Open Data. Please contact the researchers for a full list of the sourced audio clips.