Sequential dependencies in recognition memory are decision based.

Decision perseveration is consistently observed in recognition tests, such that judgements tend to repeat (e.g., "old" responses tend to follow "old" responses) across trials. This effect has been found across a range of testing styles, including old/new judgements, judgements of frequency, and confidence, and has been interpreted as reflecting the transfer of mnemonic information between trials. However, an alternative explanation that response repetition is rather the product of motor action perseveration has not yet been fully evaluated. Despite the range of response styles used across studies, repeat decisions have consistently been confounded with repeat motor responses. Across three experiments, the present study divorces decision repetition from motor priming, to determine whether decision perseveration maintains. Experiments 1 and 2 found that when participants switch hands between trials, decisions are still more likely to repeat than switch. Similarly, Experiment 3 found no difference in the influence of Previous Decision when mouse paths were able to repeat between trials compared with when they could not. In addition, all experiments show a speed advantage for repeating decisions that cannot be attributed to motor priming. We conclude that decision carryover during recognition tests is ultimately a decision-based effect. The results are discussed in terms of mnemonic models of information transfer.