Prior to a recognition test, subjects studied one set of words in a medium level of processing (LOP) task and another set of words in either a shallow or deep LOP task. Medium items received more remember judgments (and fewer know judgments) when mixed with shallow than with deep items (Experiment 1)—even when a basis was required for each remember judgment (Experiment 4). These effects were due
... [Show full abstract] to the test-list context: judgments for medium items were equivalent for the two groups when only the medium items were presented at test (Experiment 2). The relative weighting subjects assigned to particular kinds of recollected information as the basis of their remember judgments was affected by list context (Experiment 4), but their ability to remember list source was not (Experiment 3). The test-list context appears to have influenced subjects’ functional definitions of remembering and knowing rather than the contents of their recollections.