Position coding in two-digit arabic numbers.

Digit position coding in two-digit Arabic numbers was examined in two masked priming experiments. In Experiment 1, participants had to decide whether the presented stimulus was a two-digit Arabic number (e.g., 67) or not (e.g., G7). Target stimuli could be preceded by a prime which (i) shared one digit in the initial position (e.g., 13-18), (ii) shared one digit but in a different position (83-18), and (iii) was a transposed number (81-18). Two unrelated control conditions, equalized in terms of the distance between primes and targets with the experimental conditions, were also included (e.g., 79-18). Results showed a priming effect only when prime and target shared digits in the same position. Experiment 2 employed a masked priming same-different matching task - a task that has been successfully employed in the literature on letter position coding. Results showed faster response times when prime and target shared digits - including the transposed-digit condition - relative to the control conditions. Thus, the identity of each digit in the early stages of visual processing is not associated with a specific position in two-digit Arabic numbers. We examine the implication of these findings for models of Arabic number processing.