Multi-digit number processing beyond the two-digit number range: A combination of sequential and parallel processes

Investigations of multi-digit number processing typically focus on two-digit numbers. Here, we aim to investigate the generality of results from two-digit numbers for four- and six-digit numbers. Previous studies on two-digit numbers mostly suggested a parallel processing of tens and units. In contrast, the few studies examining the processing of larger numbers suggest sequential processing of the individual constituting digits. In this study, we combined the methodological approaches of studies implying either parallel or sequential processing. Participants completed a number magnitude comparison task on two-, four-, and six-digit numbers including unit-decade compatible and incompatible differing digit pairs (e.g., 32_47, 3 < 4 and 2 < 7 vs. 37_52, 3 < 5 but 7 > 2, respectively) at all possible digit positions. Response latencies and fixation behavior indicated that sequential and parallel decomposition is not exclusive in multi-digit number processing. Instead, our results clearly suggested that sequential and parallel processing strategies seem to be combined when processing multi-digit numbers beyond the two-digit number range. To account for the results, we propose a chunking hypothesis claiming that multi-digit numbers are separated into chunks of shorter digit strings. While the different chunks are processed sequentially digits within these chunks are processed in parallel.

► We examine if digits of multi-digit numbers are processed sequentially or in parallel.
► Magnitude comparison task on 2-, 4-, and 6-digit numbers.
► Inclusion of compatible (e.g., 42_57) and incompatible (e.g., 47_52) digit pairs.
► RTs and fixations revealed aspects of both sequential and parallel processing.
► We propose a chunking hypothesis to combine sequential and parallel processing.