The three experiments reported in this article examined the process by which • frequency of occurrence information is registered in memory. Based on the hypothesis that this information is encoded automatically, performance on a frequency discrimination task was predicted to be insensitive to a variety of manipulations expected to influence free recall, a task requiring considerable effortful processing and therefore used as a counterpoint to the frequency task in two of the experiments. These expectations were confirmed. Frequency performance did not increase with practice, was unaffected by appropriateness of practice, was not influenced by accuracy of test expectations, and was not hindered by competing demands. Also, no stable individual differences were obtained in discriminating relative frequency, and strategy effects were small. These results conform to criteria of automatic processing. They are also in striking contrast to the free-recall findings. Free-recall performance increased with practice, was affected by appropriateness of practice and accuracy of test expectations, and was hindered by competing demands. Also, stable individual differences were obtained in free recall.

In three experiments we explored the relation between normal variation in depressed mood and memory in college students. Subjects read and subsequently recalled stories whose protagonists experienced good, bad, and neutral events. Contrary to predictions arising independently from capacity theory and from schema theory, the recall of depressed and nondepressed subjects did not differ in either overall level or in affective content. The results are not easily handled by a conceptualization of depression, tied to schema theory, which proposes that negative cognitions are important for the initiation and maintenance of depression. The general usefulness of induction procedures in research on the depressive syndrome is discussed.

The reported experiment tested the suggestion that encoding of temporal order is automatic. Specifically, two of Hasher and Zacks's (1979) automaticity criteria were examined: (1) that the amount and appropriateness of practice received would not affect acquisition of temporal information, and (2) that reliable individual differences would not be found on a test of memory for temporal order. Contrary to expectations, neither of these criteria was confirmed: Retention of temporal order increased with practice at three (or four) successive lists. And, reliable individual differences were indicated by the findings that subjects' relative performance levels remained stable across lists, and that groups with higher average academic ability outperformed those with lower ability. Similar results were obtained for a free-recall task (in which case they were expected

In the five experiments reported here we attempted to demonstrate an effect on item memorability of the amount of effort expended during the encoding process. The encoding task in two experiments was anagram solving; here, solution difficulty was varied. In the third experiment, subjects were required to judge whether a word fit meaningfully into a sentence frame, arid the ease of making this decision was manipulated. The final two experiments involved picture naming under time pressure; pictures were displayed either with no labels (easy condition) or, as in the picture-word version of the Stroop task, with superimposed interfering labels (hard condition). In none of the experiments did our manipulations of difficulty/ effortfumess of encoding influence item retention. These findings raise questions about the robustness of the effort phenomenon.The relation between amount of cognitive effort expended during encoding and the subsequent memorability of inputs has received considerable attention of late. A frequently reported finding is that the more effort required by an orienting or cover task, the greater is the incidental retention of the items encountered during the cover task (e.g., Jacoby, Craik,

Four experiments demonstrated that adults can reliably remember frequency of occurrence information about items they have been exposed to under truly incidental memory conditions. Subjects neither knew that the ultimate test task would concern item frequency nor that they had any reason to remember the items. This was accomplished by presenting items under the guise of one of three cover tasks: anagram solving, sentence completion, and picture naming in a Stroop-like task. In addition, one experiment found that subjects who were prewarned for either a nonspecific memory test or a frequency test were no better able to judge frequency than were subjects operating under truly incidental conditions.