2.2.1. Behavioral number word-quantity mapping task

Children completed a number word-quantity mapping computer task just before the fMRI scan to encourage active task processing during the passive scan. In this task, children heard a number word while simultaneously seeing a nest of 1–4 eggs on a computer screen. Participants were told that “Dani the Dino likes to show off how many dinosaur eggs she has. She will say a number and show you some eggs, but she doesn’t always say the right number.” Children were instructed to say “yes” if the number word matched the number of eggs on the screen, and “no” if the number word did not match the number of eggs on the screen. An experimenter recorded their responses. For the congruent trials, the number word they heard matched the number of eggs on the screen, and for the incongruent trials, the number word did not match the number of eggs on the screen. On incongruent trials, the number of eggs differed from the number word by a numerical distance of one or two. The numerical values 1–4 occurred with equal frequency and the numerical distance between the visual and spoken numerical value was 1 or 2 with equal frequency (see Appendix for table of trial list). The stimuli remained on the screen until the child made a response. The experimenter initiated each trial by pressing the space bar when the child was judged to be attentive and ready to proceed. Children completed 64 trials (32 incongruent and 32 congruent trials). Data from the behavioral number word-quantity mapping task was available from 72 of the 80 child participants. Twenty-seven children completed an earlier version of the number-word mapping task that required a key press response. Those children were instructed to press “s” if the number word matched the number of eggs on the screen or “l” if the number word did not match the quantity on the screen. A Welch’s t-test indicated that children who provided a verbal response were marginally more accurate (M = 95%) than children who responded with a button press (M = 93%), t(58.50) = −1.80, p = .08. Note that the response mode for the behavioral mapping task was switched to make the task easier for younger children (<4 years).

Audio stimuli were spoken by a female experimenter using a SONY ICD PX333 Digital Voice Recorder. Each number word was uttered in a sentence. Noise reduction was performed over the whole recording. The best recorded number word was then extracted from the recorded sentence using Audacity (https://www.audacityteam.org). The audio was cut at onset and offset of speech (using zero crossing). Two hundred and fifty milliseconds of silence were added to the beginning of each file. The files were then normalized so that peak amplitude for all files was the same. And finally, the files were converted from WAV to MP3. The same audio and visual stimuli were shown across incongruent and congruent trials. The number word-quantity mapping task was presented with Eprime 2.0 software (Psychological Software Tools, Pittsburg, PA) and administered using a Lenovo 15″ touchscreen laptop.