Is 26 + 26 smaller than 24 + 28? Estimating the approximate magnitude of repeated versus different numbers

It has recently been suggested that regardless of the dimension at hand (i.e., numerosity, length, time), similar operational mechanisms are involved in the comparison process based on approximate magnitude representation. One piece of evidence for this hypothesis lies in the presence of similar behavioral effects for any comparison (i.e., the distance effect). In the case of length comparison, the comparison process can be biased by summation toward either an underestimation or an overestimation: The sum of equal-size stimuli is underestimated, whereas the sum of different-size stimuli is overestimated. Relying on the hypothesis that similar operational mechanisms underlie the comparison process of any magnitude, we aim at extending these findings to another magnitude dimension. A number comparison task with digit numbers was used in the two experiments reported presently. The objective was to investigate whether summation also biases magnitude representation of numerical and symbolic information. The results provided evidence that the summation bias can also apply to numerical magnitude comparison, since the sum of repeated numbers (26 + 26) was underestimated whereas the sum of different numbers (24 + 28) was overestimated. We propose that these effects could be accounted for by a heuristic linking cognitive effort and magnitude estimation.