Unit 2: Sensory Memory - Comprehensive Academic Notes

Introduction to Sensory Memory

• Definition: Sensory memory is a specific memory term applied to the extremely brief storage of information within a particular sensory modality.
• General Capability: Humans are capable of perceiving and remembering visual stimuli that are extremely brief. The efficiency and level of adaptation remain constant regardless of whether the objects perceived are:
  • Persistent/Permanent: Objects that remain in the field of vision.
  • Transient/Fleeting: Objects that appear and disappear rapidly.
• The Continuity of Perception: Sensory memory allows the perceptual system to store visual information long enough to bridge the gap between static images. For example, when a sequence of static images is presented rapidly with blank intervals in between, we perceive a continuous moving image. This is because the system integrates each static image with the next slightly different one.
• Specific Subtypes (Exam Priority):
  • Iconic Memory: The brief storage of visual information. It contains patterns of light waiting to be analyzed rather than letters or digits.
  • Echoic Memory: The brief storage of auditory information.
• Primary Function: These stores preserve input beyond the physical existence of the generating stimulus, essentially "extending the life" of the stimuli (Ruiz-Vargas, 1980).

Historical Origins and the Apprehension Amplitude

• Sir William Hamilton (19th Century): Hamilton repeatedly found that when a handful of balls are thrown onto a flat surface, a person cannot easily see more than $6$ or $7$ at a glance.
• W. S. Jevons: Jevons conducted experiments by estimating the number of black seeds falling into a tray.
  • Method: He would "estimate without the slightest hesitation" the number of seeds immediately after they stopped moving.
  • Findings: The average estimate was correct until it reached approximately $8$ or $9$ seeds. Beyond this amount, the error rate increased gradually.
• James Mckeen Cattell (1883):
  • Tool: Utilized a specialized machine called a Tachistoscope.
  • Technique (Total Report): Subjects were presented with sets of letters or words and asked to report every item they could remember.
  • Results:
    • When stimuli contained up to $4$ items, subjects made zero mistakes.
    • As items increased, the correct average identification settled between $4$ and $5$ items.
  • Definition: The amount of information correctly identified and remembered after a brief visual exposition is defined as the Apprehension Amplitude.

The Iconic Memory and George Sperling's Experiments

• Sperling (1960) Research Problems: Sperling identified two fundamental issues with the "apprehension amplitude" findings:
  1. Observers see more information than they can actually verbalize immediately after a brief presentation.
  2. The observer's report includes only the recalled information, not the totality of what was actually perceived.
• Total Report Task Findings:
  • Method: Exposing subjects to letter arrays for times ranging from $15\,ms$ to $500\,ms$.
  • Data: Regardless of the exposure duration ($15\,ms$, $50\,ms$, $150\,ms$, $300\,ms$, or $500\,ms$), the results were unchanged.
  • Observation: Subjects were consistently unable to remember more than $4$ or $5$ letters ($4.5$ on average).
  • Conclusion: There was no perception problem; the limitation was not attributable to the visual system itself but to a memory limit.
• Partial Report Task (The Breakthrough):
  • Hypothesis: Subjects have more information available than they can provide before the memory trace fades.
  • Method: Subjects were shown letter arrays (typically $3$ rows). Exposure was always $50\,ms$.
  • Cues: Immediately after the presentation, a tone was played to indicate which row to report:
    • High pitch ($2,500\,Hz$): Top row.
    • Medium pitch ($650\,Hz$): Middle row.
    • Low pitch ($250\,Hz$): Lower row.
  • Delay Conditions: The interval between the disappearance of letters and the cue tone varied ($0\,ms$, $150\,ms$, $300\,ms$, $500\,ms$, and $1,000\,ms$).
  • Results:
    • At $0\,ms$ delay: Subjects remembered an average of $9.1$ letters (approx. $76\%$, twice the capacity of the total report).
    • At $500\,ms$ delay: Subjects still averaged $9.1$ letters.
    • At $1,000\,ms$ delay: Performance dropped to the level of the total report (approx. $4.5$ letters).
• Key Conclusions:
  1. Information is available for longer than the physical stimulus via sensory stores.
  2. Visual information (iconic mode) decays as a function of signal delay, lasting approximately $500\,ms$.
  3. The superiority of the partial report over the total report highlights the existence of two systems:
     • Iconic Memory: High capacity, very short-lived.
     • Short-Term Memory (STM): Limited capacity, significantly longer duration.

Visual Persistence vs. Informative Persistence

• Visual Sensory Memory Unitarity: It is not a unitary system; it includes two distinct phenomena.
• Visible Persistence: This refers to the visual stimuli remaining visible for some time after physical disappearance. It includes the persistence of the characteristics of the stimuli (e.g., "I see something yellow").
• Informative Persistence: This allows for the extraction of information from a stimulus after it is physically gone. It functions as if the stimulus were still present (e.g., extracting the detail of a ball in a scene).

Masking Processes in Sensory Memory

• Masking Defined: A process where the perception or storage of a stimulus is influenced by events occurring immediately before (forward masking) or after (backward masking) the stimulus.
• Brightness Masking (Sperling, 1963):
  • The brighter the light during the interval after stimulus presentation, the poorer the memory performance.
  • Light occupies or disrupts the memory trace.
• Pattern (Visual) Masking:
  • Occurs when targets are followed by a mask containing broadly similar features.
  • Binocular Stage Influence: Pattern masking works even if the target is shown to one eye and the mask to the other. This indicates the mask affects a late stage of processing where information from both eyes is combined into a single percept.
• Averbach and Sperling (1961):
  • Immediate presentation allowed for $14$ to $18$ items to be available.
  • Delay dramatically decreased availability, specifically when fields before and after the stimulus were illuminated.

Measuring the Duration of Iconic Memory

• General Estimates: Provisional duration is between $250\,ms$ and $500\,ms$.
• Demographic Variations:
  • Young Subjects (Average $20$ years): Approx. $340\,ms$.
  • Elderly (Average $81$ years): Approx. $300\,ms$.
  • Elderly with Mild Neurocognitive Impairment (Average $84$ years): Approx. $70\,ms$.
• Measurement Techniques (Vargas, 2010):
  • Retroactive Masking: A circle cue is placed around a letter.
    • $0\,ms$ interval: Subject sees the letter inside the circle (single image).
    • $0-300\,ms$ interval: Subject reports an empty circle; the letter recognition process was interrupted before completion.
    • $>300\,ms$ interval: No interference; recognition was already completed and transferred to STM.
  • Synchrony Judges: Subjects judge if an auditory event coincides with the beginning of a visual event. The interval at which they appear simultaneous defines the icon duration.
  • Phenomenal Continuity (Haber and Standing, 1969): Alternating a black circle with a white field. If the white field lasts $250-300\,ms$, the subject perceives the circle as continuously present.
• Content Characteristics: Sperling (1960) found that when asked to report only numbers or only letters from a mixed array, subjects performed no better than the total report limit ($4.5$ items).
  • Conclusion: Iconic memory is uncategorized (pre-perceptual/physical characteristics).

Echoic Memory: The Auditory Storage System

• Necessity: The brain must maintain early segments of auditory stimuli until enough signal is received to understand meaning.
• Moray, Bates & Barnett (1965) - "The Man with Four Ears":
  • Method: Four messages sent simultaneously from four positions (left, right, front, back).
  • Results: Confirmed the superiority of the partial report in auditory mode. Using a light cue to signal the relevant channel resulted in more letters recalled than the total report condition.
• Massaro (1972) - Echoic Masking:
  • Method: Presented vowels for $20\,ms$ followed by a masking tone at intervals of $0-500\,ms$.
  • Findings:
    • Short intervals ($0-35\,ms$): Identification was poor ($50\%$ error rate).
    • At $250\,ms$: Identification reached an asymptote (almost perfect).
  • Conclusion: Echoic memory duration is approximately a quarter of a second ($250\,ms$).
• Crowder (1971) - The Suffix Effect:
  • A list of $9$ digits was presented.
  • If followed by a tone (no masking), recall was high.
  • If followed by the digit "zero" (suffix/masking), recall was significantly worse.
• Two Levels of Auditory Representation:
  1. Auditory Persistence: Peripheral, crude, pre-categorical. Duration: $250-300\,ms$.
  2. Post-categorical Auditory Representation: Perceptual and abstract. It is generated after the stimulus disappears and can maintain information for over $10$ seconds to allow for full recognition.