On Memory for Everyday Symbols
Bibliographic record
Abstract
This thesis investigated the memorability of common graphic symbols (e.g., !@#$%) and logos. In an initial set of 4 experiments, participants were presented during study with symbols or words (e.g., $ or ‘dollar’). In Experiment 1, memory performance assessed using free recall demonstrated markedly better memory for symbols relative to their word counterparts, manipulated within-subject. Experiment 2 systematically varied whether symbols and words were presented at encoding and during a subsequent recognition test, manipulated between-subjects. Results conceptually replicated the findings of the first experiment, showing superior memory for symbols even when the retrieval test and study design were changed. Furthermore, by analyzing group data based on which stimuli (words or symbols) were used in the encoding and retrieval phases of the experiment, symbol superiority in memory was determined to be driven by encoding-based mechanisms. An alternative explanation holds that symbols may benefit memory as a result of their smaller overall set size compared to words. Experiment 3 addressed this potential issue by restricting the to-be-remembered set of words to a single category (common kitchen produce) whose set size was like that of the symbols that I used. Once again, symbols were better remembered than the words. This experiment showed not only that set size was not likely to be driving the previously seen memory benefit for symbols, but also that representing abstract concepts with symbols successfully reversed the concreteness effect in memory: Symbols were better remembered even when compared to highly concrete nouns. A fourth experiment directly tested a dual coding account by comparing memory for symbols, pictures, and words. There, symbols and pictures were both better remembered than words, and memory for symbols and pictures did not differ. Symbols not only were remembered just as well as images—as I predicted based on dual coding theory—but they also entirely eliminated the concreteness advantage in memory for pictures as well: Memory for symbols representing abstract concepts was equivalent to that for pictures depicting concrete objects. In Experiment 5A and 5B, I compared memory for professional sports teams presented in three encoding conditions: team names only, team logos without team names, and team logos with integrated team names. Across two experiments, while memory was often best for logos relative to team names, familiarity moderated this relation. When assessing memory for team names, the magnitude of the benefit for the logos-only condition depended on whether participants knew what the logos represented. In the sixth and final experiment, 337 naïve participants rated the set of symbols used in Experiments 1-4 on their meaning-based familiarity with each symbol and on their frequency of encountering it. Machine learning estimations of inherent stimulus memorability were provided by the ResMem residual neural network. These computer-derived memorability estimates correlated with memory for symbols, but familiarity and frequency ratings did not. Hierarchical linear regression revealed that inherent memorability estimates explained significant portions of variance for symbol memory, over and above effects of familiarity and frequency. This dissertation is the first to present evidence that, like pictures, graphic symbols and logos are better remembered than words, in line with dual coding theory and with distinctiveness accounts. Symbols offer a visual referent for abstract concepts that are otherwise unlikely to be spontaneously imaged. Symbols also provide visual stimuli that are often both physically and conceptually unique. It is the visual nature of symbols that confers the impressive memory performance benefits.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; both teacher heads agree on what is shown here.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".