Imagery - Implicit Perceptual (8)

Effects of Imagery on Perceptual Implicit Memory Tests

Reference: McDermott & Roediger, 1994 (not directly on quiz 2)0

Implicit Tests

  • Types of Implicit Tests:

    • Conceptual (Meaning-Driven)

    • Perceptual (Data-Driven)

  • Dissociations from Explicit Tests:

    • Example 1: Picture superiority effect in recall

    • Example 2: Word fragment completion shows priming from words only, and no priming from pictures*

  • Priming:

    • Defined as based on the appropriateness of processing from study to test.

    • Note: If the test is based on perceptual information, then priming will not occur unless that perceptual information was processed at study.

    • Priming is not strongly influenced by tasks participants do at study*

    • It turns out this is mostly true for perceptual implicit tests.

Cross-Modal Priming

  • Studying pictures produces some priming on word fragment completion.

  • Auditory presentation of words can prime visual tests.

  • Hypothesis: Participants may create their own “data,” leading to derived “percept” that causes priming.

Imagery

  • Evidence indicates that imagery functions similarly to perception:

    • Mental Rotation Time: The time it takes for individuals to mentally rotate an image shows a relationship with physical rotation.

    • Resolution of Images: The clarity or detail of mental images can mirror physical perception.

    • ERP and PET Data: Studies show similar brain areas are active during imagery tasks and perception tasks.

    • Patient Data: Individuals with visual and imagery deficits provide further evidence of imagery's relation to perceptual functions.

Questions Raised

  • Can imagery be utilized to prime implicit memory for perceptual information that is not actually perceived?

  • Is it a necessary condition of priming that the physical stimulus must be encountered?

Mixed Imagery Effects on Implicit Tests

  • Jacoby & Witherspoon (1982):

    • Finding: Hearing words and spelling them produced priming of (0.21) on a word identification test compared to (0.30) from reading the words.

    • Hypothesis: Participants may have imagined the printed word while spelling it.

  • Jacoby & Dallas (1981):

    • Finding: Hearing high-frequency words did not produce priming on a word identification task (small effect for low frequency).

  • Donnelly (1988):

    • Findings: Hearing & spelling (0.22) vs. just hearing (0.16) on fragment completion compared to (0.27) for visual.

    • Follow-up showed no additional priming when instructed to imagine printed words versus just hearing them.

  • Roediger & Blaxton (1987):

    • Finding: Imagining printed words led to similar priming as seeing printed words (0.21 vs. 0.24), greater than hearing only (0.15).

Experiment 1: Imagery & Picture Fragment Identification

  • Design:

    • Three study conditions (between-subjects)

    • Read words and rate pleasantness: word condition

    • Read words, form an image of referents, and rate pleasantness: word-image picture condition

    • See pictures and rate pleasantness: picture condition

  • Implicit Picture Fragment Completion Task:

    • Half of fragments were studied items, and half were unstudied items.

    • Fragments were presented briefly (100ms or 200ms) and then masked.

Experiment 2: Replication

  • Conditions:

    • 2A: Word-image picture condition.

    • 2B: Word-image picture condition, rating vividness of images instead of pleasantness to emphasize imagery instructions and possibly increase priming.

Experiment 3: Geons

  • Recoverable vs. Unrecoverable Fragments:

    • Unrecoverable fragments are much harder to identify due to incomplete geon recovery.

    • More time assists with identifying recoverable fragments but not unrecoverable ones.

    • Generating a “percept” with imagery may help in priming the recoverable parts of geons.

    • It is posited that imagery should not aid identification of unrecoverable fragments.

Experiment 3 Variables (2 x 2 x 2)

  • Study Conditions:

    • Word vs. Word-image picture

  • Test Items:

    • Studied vs. Non-studied

  • Fragment Types:

    • Recoverable vs. Non-recoverable fragments

Experiment 4: Design

  • Fully Crossed Design:

    • Examines match of type of imagery with type of implicit test.

    • Tests whether imagery instructions yield greater effort at processing.

  • Conditions:

    • Study words or pictures

    • Rate for pleasantness or perform opposite-format imagery plus vividness

    • Complete word fragment or picture fragment test

    • Test items included both studied and non-studied fragments.

Experiment 4 Results

  • Key Findings:

    • 1) Imaging pictures from words produced priming on the picture fragment identification task.

    • Resulted in greater priming than deep processing of words (without imagery).

    • Resulted in less priming than viewing actual pictures.

    • 2) Imagery of pictures assisted identification of recoverable fragments only.

    • 3) Imagery must match format to show transfer:

    • Imaging words primed word fragment completion; imaging pictures primed picture fragment identification.

    • Imagery did not universally enhance priming beyond actually seeing the relevant stimuli.

Discussion

  • Main Conclusions:

  • Perceptual priming does not necessarily require exposure to the physical stimulus; self-created percepts can result in priming.

  • Evidence suggests that imagery is functionally similar to perception.

  • Perceptual priming aligns with Transfer Appropriate Processing (TAP) theory.

  • The assertion from the study: "Implicit memory tests are not badly contaminated by intentional recollection" (p. 1387)

    • Expectation: With contamination, we might anticipate greater cross-format “priming.”

  • Other studies have demonstrated implicit/explicit dissociations concerning Picture Superiority Effect (PSE).