Academic journal article Memory & Cognition

False Recognition across Meaning, Language, and Stimulus Format: Conceptual Relatedness and the Feeling of Familiarity

Academic journal article Memory & Cognition

False Recognition across Meaning, Language, and Stimulus Format: Conceptual Relatedness and the Feeling of Familiarity

Article excerpt

Four experiments examined contributions of conceptual relatedness and feelings of familiarity to false recognition. Participants first studied lists of unrelated items (e.g., table, lock) followed by a recognition test with three types of items: (1) studied items (e.g., table), (2) semantically related lures (e.g., key), and (3) unrelated lures (e.g., cup). Participants falsely recognized more related than unrelated lures when the stimuli were words (Experiment 1A) and pictures (Experiment 1B), when the studied items and related lures differed in language (Experiment 2), and when they differed in perceptual format (Experiment 3). In Experiment 4, an attribution manipulation, designed to make feelings of familiarity nondiagnostic for memory judgments, eliminated the false-recognition effect obtained in Experiment 3. Overall, the study suggests that conceptual relatedness produces false recognition even in the absence of shared perceptual surface features between study and test items, and it does so by generating feelings of familiarity.

Memory illusions capture the attention of psychologists and the general public. A frequently studied type of memory illusion occurs when a person "recognizes" a test item that was never presented at study. The present research explores three questions about false recognition: First, does false recognition occur when a critical test item is only conceptually related to the study item, even if the test item is presented in a different physical format or in a different language than the study item? second, does false recognition occur even when a critical item is only weakly prepared by a single conceptually related study item? Third, is false recognition of conceptually related items supported by a distinct feeling of familiarity, and can it be eliminated by making that feeling nondiagnostic for recognition judgment? Answers to these questions will help us better to understand the scope and mechanisms of memory illusions.

Perceptual and Conceptual Contributions to False Recognition

To explain false recognition, previous research often focused on perceptual fluency, or the ease of processing a stimulus's physical identity (e.g., Jacoby & Dallas, 1981; Johnston, Hawley, & Elliot, 1991; Lindsay & Kelley, 1996; Rajaram, 1993; Verfaellie & Treadwell, 1993). Several studies have found that false recognition can be enhanced with simple perceptual manipulations, such as subliminally priming the test word with an identical word (e.g., Jacoby & Whitehouse, 1989); increasing clarity of the test word (e.g., Whittlesea, Jacoby, & Girard, 1990); or having participants perform a perceptual task, such as fragment completion, on the test item (e.g., Luo, 1993). Presumably, these manipulations enhance false recognition because greater perceptual fluency of the nonstudied test item is misattributed to its presentation during study (Jacoby, Kelley, & Dywan, 1989). Accordingly, false-recognition rates are reduced when participants have an alternative explanation for the enhanced fluency, as when, for example, the test word is preceded by an easily visible prime (Jacoby & Whitehouse, 1989).

False recognition is also susceptible to conceptual influences. Whittlesea (1993) found that test words (e.g., boat) were falsely recognized at a higher rate when preceded by semantically predictive sentence stems (e.g., The stormy seas tossed the . . .) than when preceded by nonpredictive stems (e.g., She saved her money and bought a . . .). Presumably, the predictive stem unobtrusively enhances conceptual fluency of the test word, which is then misattributed to the study episode (Whittlesea, 1993). Conceptual influences on false recognition have also been explored using a converging associates paradigm, also known as the Deese, Roediger, and McDermott (DRM) paradigm. In this paradigm, participants study a list of semantic associates (e.g., table, sit, legs, seats, etc. …

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

Oops!

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.