10: Levels of Processing Theory and Experiments

Levels of Processing Theory

  • Origin of the Theory

    • Developed by Craik in 2002.

    • Influenced by Treisman’s theory of selective attention, which evolved into the feature integration theory.

    • Concept: Features build up into larger components, culminating in objects.

    • Key principle: Deeper analysis necessitates more attention.

    • Proposition: Memory may exhibit the same pattern where depth of analysis predicts memory longevity.

    • Collaborated with Lockhart on a theoretical piece (Craik & Lockhart, 1972) which has received significant citations:

    • Current citation counts:

      • Scopus: 5,666; 6,208; 6,531; 6,866

      • Google Scholar: 15,804; 16,949; 17,760; 18,543

  • Concept of Levels of Processing (LOP)

    • Asserts that remembering is not simply retrieving stored information, likened to finding items in a warehouse.

    • Proposes that there are not discrete structural memory stores; rather, there are different memory processes categorized as primary and secondary.

    • Claims that remembering reflects the quality of processing during encoding:

    • The deeper the analysis during encoding, the higher the likelihood of subsequent remembrance.

    • Processing ranges from shallow (surface features) to deeper conceptual analysis.

    • Updated understanding indicates that a fixed order of depth may be unlikely.

    • Shallow vs. deep processing is characterized as a qualitative difference rather than a quantitative one.

  • Evidence Supporting LOP: Craik & Tulving (1975)

    • Conducted 10 experiments aimed at empirically testing LOP principles.

    • Utilized semantic orienting questions to assess memory recall:

    • Items that are congruent (e.g., Is a soprano a singer?) are remembered better than incongruent items (e.g., Is mustard concave?).

    • Discussed challenges regarding organization at encoding and retrieval, including item-specific vs. relational organization.

    • Noted issues with the circularity in defining “deep” processing.

    • Highlighted concerns about the processing time involved in memory tasks.

General Method in LOP Experiments

  • Study Methodology

    • Participants were misled to believe the experiment focused on perception and reaction speed.

    • Words were presented for 200 milliseconds using a tachistoscope.

    • Participants answered orienting questions regarding each word, with reaction times recorded:

    • Types of Questions:

      • Structural (e.g., Is it in capitals?)

      • Phonemic (e.g., Does it rhyme with __?)

      • Category (e.g., Is it a ___?)

      • Sentence (e.g., Does it fit in: “He met a ___ in the street”?)

    • Conducted a surprise retention test encompassing free recall, cued recall, and recognition tasks.

  • Experiment Breakdown

    • Experiment 1 (n = 20):

    • Explored five levels of processing:

      1. Is there a word present?

      2. Is it in capitals?

      3. Does it rhyme with ___?

      4. Is it in the category ____?

      5. Does it fit into the sentence __?

    • Included an old/new recognition test with printed words, where participants circled familiar ones.

Findings from Experiments

  • Experiment 2 (n = 24):

    • Recognitions plotted against initial decision time revealed a linear relationship.

    • Slope for affirmative (Yes) responses was steeper compared to negative (No) responses, suggesting processing time may be more significant than mere level of processing.

  • Experiments 3 & 4 (n = 20):

    • Explored three levels of processing involving case, rhyme, and sentence structures.

    • Experiment 3 involved incidental encoding while Experiment 4 focused on intentional encoding.

    • Half of the words in each condition were presented twice to enhance encoding, with a consistent LOP and response congruency condition but varying queries for rhyme and sentence types.

Comparison of Incidental vs. Intentional Encoding

  • Discussion of Results from Experiments 1-4:

    • Provided support for the LOP principle: deeper processing enhances retention, applicable to both incidental and intentional encoding.

    • Noted inconsistencies such as:

    • Decision times for Yes and No responses were similar, yet memory performance varied significantly, particularly at deeper LOPs.

    • Suggested that decision time may not serve as a suitable indicator of memory success.

    • Proposed a distinction between quantitative processing time and qualitative elaboration degree.

Further Experiments and Analysis

  • Experiment 5 (n = 24):

    • Investigated the role of processing time:

      • Tested with a slow non-semantic question to determine if a word fits a consonant-vowel pattern.

      • Example: CCVVC → brain (yes);

      • Used a semantic sentence question and conducted an old/new recognition test.

  • Experiment 6 (n = 16):

    • Addressed why congruent (Yes) items are more easily remembered than incongruent (No) items:

    • Both require semantic processing albeit differ in the extent of integration or elaboration.

    • Example: Four-footed animal? BEAR vs. CLOUD.

    • Congruent targets integrate better with cues, while elaborate features aid recall more accurately.

    • Sample statistics: Recall for Yes = 0.36, for No = 0.39.

  • Experiment 7 (n = 20):

    • Utilized only sentence questions with varying encoding complexity.

    • Sentences like “He dropped the _” or “The old man hobbled across the room…” employed complexity in eliciting responses. Watch is ‘yes’ to both

    • Finished with free recall (No Congruence Recall) and followed with sentence cued recall.

Refining Depth and Elaboration Understanding

  • Depth vs. Elaboration:

    • While depth (from structure to phoneme to semantics) is crucial, the degree of elaboration within the level also plays a significant role.

    • Explored spreading or quantity of encoded features.

    • Suggested that the effects of differential encoding could be minimized.

    • Noted the variation between frequent vs. infrequent trial types impacting memorability.

Classroom Demonstration and Monetary Incentives

  • Experiment 9: Classroom Demonstration:

    • Implemented a relaxed control over procedures for intentional encoding.

    • Words were projected in a group setting (n = 12 participants).

    • Responses recorded on paper with no reaction time measures employed during testing.

  • Experiment 10:

    • Examined the effect of offering monetary incentives ($0.01, $0.03, or $0.06) for the recognition of shallowly processed words within the context of intentional encoding.

    • Conditions manipulated across three groups of participants (n = 3 groups of 12).

Broad Conclusions and Implications

  • Discussion Insights:

    • The crux of memory retention is not solely about the intention to learn or the effort exerted; rather, it is the qualitative nature of tasks and operations performed on items that significantly influences retention.

    • Emphasized how depth is not as rigid as initially thought, underscoring the importance of elaboration.

    • Critiqued the role of processing time as a predictor for retention.

  • Concept of Consolidation:

    • Noted that while deep processing is both necessary and beneficial, it is not independently sufficient for later episodic memory recall.

    • Raised questions regarding the phenomenon where individuals with amnesia can execute processing tasks without demonstrating retention.

    • Also queried why divided attention during encoding diminishes retention while allowing task execution to remain intact.

    • Concluded that consolidation involves a set of neurological processes that operate outside conscious awareness and cognitive control.

Advances in Memory Research

  • 30 Years Later… Hierarchical Model:

    • Discusses a context-free approach to understanding memory.

    • Categories of memory identified as:

    • "Semantic memory" for global concepts.

    • "Episodic memory" for context-specific details, such as names.