<p>The subsequent memory effect (SME) refers to neural patterns (e.g., in EEG or fMRI) at encoding that predict later memory performance. In N400-based SMEs, for example, items later remembered elicit less negative N400 amplitudes at encoding compared to items later forgotten. These effects have traditionally been interpreted as reflecting idiosyncratic neural states during encoding—in the case of the N400, states related to semantic activation—that influence episodic encoding success. However, recent work on memorability, a stable, item-level property indicating the population-level likelihood that an image will be remembered, has shown that high (compared to low) memorability images elicit less negative N400 amplitudes, suggesting that memorability is linked to more targeted semantic mapping. This raises the question of whether encoding-related effects are more tied to intrinsic stimulus properties or in-the-moment encoding variability. The present study examined both factors in tandem: ERPs were recorded while participants viewed images varying in memorability and were later classified by recognition outcome (hit vs. miss). Analyses revealed that N400 amplitudes were significantly predicted by memorability scores even when controlling for subsequent memory performance. Memorability also predicted Late Positive Complex SMEs. These findings suggest that neural activity traditionally associated with later memory success may capture item-level properties rather than transient encoding states. Consequently, memorability appears to be a key driver of differences in memory performance, challenging interpretations of SMEs as purely state-dependent and highlighting the importance of considering intrinsic stimulus characteristics when evaluating effects correlated with memory success.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

ERP effects at encoding: Image memorability or recognition success?

  • Will Deng,
  • Kara D. Federmeier

摘要

The subsequent memory effect (SME) refers to neural patterns (e.g., in EEG or fMRI) at encoding that predict later memory performance. In N400-based SMEs, for example, items later remembered elicit less negative N400 amplitudes at encoding compared to items later forgotten. These effects have traditionally been interpreted as reflecting idiosyncratic neural states during encoding—in the case of the N400, states related to semantic activation—that influence episodic encoding success. However, recent work on memorability, a stable, item-level property indicating the population-level likelihood that an image will be remembered, has shown that high (compared to low) memorability images elicit less negative N400 amplitudes, suggesting that memorability is linked to more targeted semantic mapping. This raises the question of whether encoding-related effects are more tied to intrinsic stimulus properties or in-the-moment encoding variability. The present study examined both factors in tandem: ERPs were recorded while participants viewed images varying in memorability and were later classified by recognition outcome (hit vs. miss). Analyses revealed that N400 amplitudes were significantly predicted by memorability scores even when controlling for subsequent memory performance. Memorability also predicted Late Positive Complex SMEs. These findings suggest that neural activity traditionally associated with later memory success may capture item-level properties rather than transient encoding states. Consequently, memorability appears to be a key driver of differences in memory performance, challenging interpretations of SMEs as purely state-dependent and highlighting the importance of considering intrinsic stimulus characteristics when evaluating effects correlated with memory success.