Depth of processing pictures of faces and recognition memory.

These studies ask whether S remembers a picture better the greater the depth of processing he allots to it. Depth of processing pictures of faces was varied according to judgments of sex (superficial) or judgments of likableness or honesty of the person pictured. Performance on a later recognition memory test was high for pictures judged for likableness or honesty, and low for pictures judged for sex. This ordering held astrue for intentional learners as for incidental learners. A final experiment showed that face recognition memory was not materially affected by a context manipulation: an old test picture was remembered at a level determined by its original depth of processing and independently of how it was tested—either alone, along side an old picture it had been studied with, or with a new picture. In a recent paper, Craik and Lockhart (1972) argued that depth of processing of stimulus material is a direct determinant of how well that material will be remem-bered. The underlying assumption in their approach is that a stimulus is processed through a series of stages with different kinds of information being extracted from or triggered off by the stimulus at suc-cessive stages. Sensory features of the stimulus are presumably extracted first, whereas associative information (such as the name or meaning of a grapheme) be-comes available later. In support of their depth of processing hypothesis, Craik and Lockhart review studies showing higher incidental learning for words which 5s had processed for meaning than for items proc-essed for physical attributes. For instance, Hyde and Jenkins (1969) oriented 5s to answer different questions with respect to a word, either counting the number of letters in it or the number of es, or rating it for pleasantness. Those 5s who did the pleasantness judgments recalled the words later much better than did the other 5s. Similarly, Johnston and Jenkins (1971) showed that 5s required to think