Skip to main content
. 2020 May 29;9:e55335. doi: 10.7554/eLife.55335

Figure 5. Individual differences in pattern completion assays.

(a–c) Effects of age on hippocampal activity (associative hit – correct rejection) and category-level reinstatement strength (mean logits) in VTC and ANG during associative hits. (d–f) Independent of age, individual differences in hippocampal activity and category-level reinstatement strength in VTC and ANG during associative hits explain significant variance in exemplar-specific recall. (g–i) Independent of age, individual differences in hippocampal activity and VTC category-level reinstatement strength also explain significant variance in standardized delayed recall performance; the relation with ANG category-level reinstatement did not reach significance. Scatterplots reflect raw values for each measure. See Figure 5—figure supplement 1 for partial plots controlling for nuisance variables. Each point represents an individual participant. Plots also show linear model predictions (black line) and 95% confidence intervals (shaded area). VTC = ventral temporal cortex; ANG = angular gyrus.

Figure 5—source data 1. Behavioural and neural measures used in individual differences analyses, depicted in Figure 5a–i and Figure 5—figure supplements 1–5.

Figure 5.

Figure 5—figure supplement 1. Partial plots of individual differences in pattern completion assays.

Figure 5—figure supplement 1.

(a–c) Effects of age on hippocampal activity (associative hits – correct rejections) and category-level reinstatement strength (mean logits) in VTC and ANG during associative hits. (d–f) Independent of age, individual differences in hippocampal activity and category-level reinstatement strength in VTC and ANG during associative hits explain significant variance in exemplar-specific recall. (g–i) Independent of age, individual differences in hippocampal activity and VTC category-level reinstatement strength also explain significant variance in standardized delayed recall performance; the relation with ANG category-level reinstatement did not reach significance. Scatterplots reflect partial plots controlling for nuisance variables. Each point represents an individual participant. Plots also show linear model predictions (black line) and 95% confidence intervals (shaded area). VTC = ventral temporal cortex; ANG = angular gyrus.
Figure 5—figure supplement 2. Event-Level reinstatement strength (within-event – within-category ERS) during associative hits declines with age in (a) VTC, but not (b) ANG.

Figure 5—figure supplement 2.

Each point on the scatterplot represents an individual subject. Plots also show the linear model predictions (black line) and 95% confidence interval (shaded area). VTC = ventral temporal cortex; ANG = angular gyrus.
Figure 5—figure supplement 3. Relationship between Encoding Strength and Category-Level Reinstatement Strength in (a) VTC and (b) ANG.

Figure 5—figure supplement 3.

In both regions, encoding strength (mean logits across leave-one-run-out-n-fold cross validation) is a significant predictor of category-level reinstatement strength (mean logits) during associative hits (VTC: β = 0.45, p < 10−5; ANG: β = 0.62, p < 10−11). Each point on the scatterplot represents an individual subject. Plots also show the linear model predictions (black line) and 95% confidence interval (shaded area). VTC = ventral temporal cortex; ANG = angular gyrus.
Figure 5—figure supplement 4. Individual differences in pattern completion assays predict associative d’.

Figure 5—figure supplement 4.

Independent of age, individual differences in associative d’ are explained by (a) hippocampal activity (associative hits – correct rejections; β = 0.34, p < 0.001) and category-level reinstatement strength (logits) in (b) VTC (β = 0.38, p < 0.001) and (c) ANG (β = 0.39, p < 0.001) during associative hits (see Supplementary file 1 for model parameters). Each point on the scatterplot represents an individual subject. Plots also show the linear model predictions (black line) and 95% confidence interval (shaded area). VTC = ventral temporal cortex; ANG = angular gyrus.
Figure 5—figure supplement 5. Pattern Completion Metrics Covary across Participants.

Figure 5—figure supplement 5.

(a) Hippocampal activity during associative hits (associative hits – correct rejections) is related to category-level reinstatement strength (logits) in VTC (β = 0.19, p < 0.05) and (b) ANG (β = 0.21, p < 0.01) during associative hits. (c) Category-level reinstatement strength (logits) in VTC and ANG during associative hits are related (β = 0.44, p < 10−7). Each point on the scatterplot represents an individual subject. Plots also show the linear model predictions (black line) and 95% confidence interval (shaded area). VTC = ventral temporal cortex; ANG = angular gyrus.