Expediting arithmetic automaticity: Do inefficient computation methods induce spontaneous testing effects?

Before achieving arithmetic automaticity (direct answer recall), learners self-compute answers or use calculators/tables. Mechanisms supporting arithmetic automaticity are of interest because automaticity predicts math achievement and employment outcomes. Future recall is enhanced when, during study, learners attempt to recall the to-be-learned information (“testing effect”). However, learners may not attempt retrieval when another strategy is available (computation, re-reading). We expected learners to more frequently attempt recall when the alternative strategy was inefficient. Participants learned novel arithmetic problems, and could recall answers (if able) or use a computation method, which varied by condition: (i) self-computation; (ii) calculator-computation; and (iii) using a calculator with scrambled keys. Final recall ability was highest after the condition with the most inefficient non-retrieval method (scrambled calculator). We assumed recall attempts preceded computation, but also considered parallel execution. As evidence against parallel retrieval, answer recall rarely interrupted computation (mental/calculator) after it began. Thus, learners should make pre-emptive recall attempts.

• View at Publisher
• View at Scopus