Article Figures & Data
Figures
- Figure 1.
Experimental design consisting of four phases: first target-detection, object–scene association, second target-detection, and retrieval. The first and second target-detection phases were completed in the fMRI scanner. In the first target-detection phase, participants were shown a series of objects followed by a series of scenes while completing a target-detection task. In the object–scene association phase, objects were superimposed on scenes and participants were instructed to imagine using the object in the place depicted. During the second target-detection phase, objects were presented one-by-one again (in pseudorandomized order). Finally, during retrieval, objects (either old or new) were superimposed on scenes (either old or new) and participants were asked, first, whether the object was old or new, and then, whether the object–scene pair was old or new.
- Figure 2.
Illustration of item- and category-level distinctiveness metrics. Item-level distinctiveness was measured as the difference between within-item similarity (i.e., the across-voxel correlation of the same object from the first and second target-detection phases; shown in blue) and between-item similarity (i.e., the similarity of one object from first target-detection to all other objects from second target-detection; shown in green). Category-level distinctiveness was calculated as the difference between within-category similarity (i.e., the similarity between an object and all other objects from first target-detection; displayed in purple) and between-category similarity (i.e., the similarity between objects and scenes; displayed in yellow).
- Figure 3.
Proportion of correct responses with respect to item memory (left) and pair memory (right). No age differences were found for item memory between younger (YA; orange) and older (OA; blue) adults, but younger adults had a higher proportion of correct pair-memory responses than older adults. Box plots represent the interquartile range (first and third quantile with a median center line), with dots representing individual participants.
- Figure 4.
Regions demonstrating neural distinctiveness at the item (left column) and category (right column) level in younger (YA; top row) and older (OA; bottom row) adults. Peak loci of each cluster are reported in the Extended Data (Figure 4-1).
- Figure 5.
Regions demonstrating age differences in neural distinctiveness. At the item level, younger adults (YA; orange) had higher neural distinctiveness than older adults (OA; blue) in occipital cortices (top row), but lower distinctiveness in parietal cortices (middle row). At the category level, YA also showed higher neural distinctiveness than OA in occipital cortices (bottom row). Half-violin plots illustrate the sample density of distinctiveness within each region. Peak loci of each cluster are reported in the Extended Data (Figure 5-1).
- Figure 6.
Relationship between neural distinctiveness and memory identified by PLSC. Correlation coefficients (left) of the significant latent relationship for item and pair memory in younger (YA; orange) and older (OA; blue) adults. Bootstrap ratios (middle) revealing age-related neural dedifferentiation at both the item and category level as the stable components with the robustness cutoff indicated by the dashed line. Higher neural distinctiveness was associated with better memory performance across individuals. Within-group least-squares lines are shown along with 95% confidence intervals (in gray). Solid lines indicate the within-group regression and dashed lines show the corresponding slopes projected along the whole x-axis to compare across groups.
- Figure 7.
Regions demonstrating greater category-level distinctiveness for subsequently remembered objects than forgotten objects in younger adults (YA; top), older adults (OA; middle), and age differences in this memory effect (bottom). Peak loci of each cluster are reported in the Extended Data (Figure 7-1).
- Figure 8.
Neural distinctiveness was associated across representational levels both across individuals (left) and within individuals (right). Younger adults (YA) are depicted in orange and older adults (OA) are depicted in blue. Across-group averages are shown in black. Dots represent individual participants in the scatter plot. For visualization, item-level distinctiveness was binned into quintiles according to category-level distinctiveness in each participant. Item-level distinctiveness was then averaged within each age group for each quintile. Error bars indicate ±1 standard error.
Extended Data
Figure 4-1
Clusters identified by searchlight similarity analyses revealing high item- and category-level distinctiveness in younger and older adults. Download Figure 4-1, DOCX file.
Figure 5-1
Clusters revealed by searchlight similarity analyses demonstrating age differences in item- and category-level distinctiveness. Download Figure 5-1, DOCX file.
Figure 7-1
Clusters identified by searchlight similarity analyses revealing greater category-level distinctiveness for subsequently remembered objects than forgotten objects and age differences therein. Download Figure 7-1, DOCX file.
In this issue
