Academic journal article Memory & Cognition

Orthographic Neighborhood Size Effects in Recognition Memory

Academic journal article Memory & Cognition

Orthographic Neighborhood Size Effects in Recognition Memory

Article excerpt

This study argues for the importance of physical word features in recognition memory by investigating the influence of orthographic distinctiveness. Experiment 1 demonstrated a mirror effect in a yes/no recognition test by manipulating orthographic neighborhood size. Words with small neighborhoods showed more hits and fewer false alarms than did words with larger neighborhoods. Experiment 2 replicated the neighborhood size mirror effect using null pairs in a forced choice recognition test. Experiment 3 required remember/know judgments in a yes/no recognition task. Experiment 4 used the same yes/no test as did Experiment 1, adding a study task that drew attention away from orthographic information in the study list. The mirror pattern disappeared with the addition of the study task.

A common assumption in memory theory is that words are represented in memory in terms of three different types of features-orthographic, phonemic, and semantic-all of which combine to form a memory representation. Whereas semantic features have long been thought to be dominant in memory, nonsemantic features are thought to have only a transient effect. This assumption is implicit in both stage models (see, e.g., Atkinson & Shiffrin, 1968; Waugh & Norman, 1965) and levels of processing models (Craik & Lockhart, 1972), which emphasize the importance of semantic over nonsemantic features in creating accessible, durable memory traces.

However, semantic features may not be the only important information influencing memory. Diana, Peterson, and Reder (2004) examined the role of nonsemantic features in recognition by manipulating the frequency of fonts used in presentation of the study list. In the subsequent recognition test, low-frequency fonts showed higher hit rates as well as lower false alarm rates even though subjects had been instructed to decide only on the basis of the word, regardless of font. This is evidence that physical features do influence recollection of target words.

The orthographic distinctiveness effect, the finding that orthographically distinct words facilitate recollection, is another example of physical word features' influencing memory. As early as 1972, researchers were reporting evidence that orthographic distinctiveness increased both recognition and recall (Hirshman & Jackson, 1997; Hunt & Elliott, 1980; Zechmeister, 1972). Zechmeister hypothesized that an implicit response to a "strange" or "unusual" word might be elicited by orthographic information. This response could then be stored as part of the memory for the study phase, providing an additional retrieval cue.

Recently it has been claimed that orthographic distinctiveness may be a contributing factor in the word frequency effect, which is the finding in recognition tests that words of low normative frequency (low-frequency, or LF, words) show both a higher hit rate and a lower false alarm rate than do words of high normative frequency (high-frequency, or HF, words) when presented in a mixed list (Glanzer & Adams, 1985, 1990). Malmberg, Steyvers, Stephens, and Shiffrin (2002) investigated the hypothesis that LF words are recognized better than HF words because they contain more distinct orthographic features-hi this instance, the frequency of individual letters within words on the study list. Malmberg et al. found that words made up of primarily LF letters were recognized better than words made up of HF letters. More importantly, they found a mirror effect as a function of letter frequency: Words made up primarily of LF letters showed both higher hit rates and lower false alarm rates than did words made up primarily of HF letters.

Although early studies relied on a poll of their subject pool to report which words "looked unusual" (Hunt & Elliott, 1980; Zechmeister, 1972), more recent studies have attempted to operationally define orthographic distinctiveness hi terms of orthographic neighborhood size. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.