Damas-López, J. & García-Orza, J.
Faculty of Psychology. University of Málaga
In the last decade there has been a great controversy regarding whether two-digit numbers are processed compositionally (i.e., each digit pair being processed separately as a decade digit and a unit digit), holistically (i.e., each digit pair being processed as one number) or both. As the evidence accumulated, the focus of the debate has changed from the initial question to explore the conditions in which the holistic or the decomposed processing occurs. It has been suggested that holistic processing is favoured when, in comparison tasks, a given number has to be compared to a standard presented serially some seconds before. As well, it has been observed that it is easier to find compositional effects when, together with between-decade trials (e.g., 23 56), within-decade trials (e.g., 23 27) are included in the experiments. Here, two experiments using a number matching task (NMT) explored whether two-digit numbers are processed holistically or in a compositional fashion. In the NMT participants are required to decide whether one of the two numbers initially provided (cues) is presented some ms later or not (probe). Probes which have some arithmetic relationship to the cues (e.g., cues: 2 3, probe: 6) are rejected slower than those probes unrelated to its cues (e.g. cues: 2 3, probe: 7) -interference effect-, which is considered as an evidence of the automatic activation of that relationship. In our experiments, participants were presented with two-digit cues and probes which had an arithmetic progression relationship with common difference 1 or -1, only detectable once the numbers were decomposed (Experiment 1: cues: 56 7, probe: 89; Experiment 2: cues: 45 67, probe: 89). Results showed longer response times in these conditions compared to unrelated conditions. Data support a componential processing even when the numbers involved in the task are presented serially.