Investigating strength and frequency effects in recognition memory using type-2 signal detection theory

Area of Science:

• Cognitive Psychology
• Neuroscience
• Memory Research

Background:

• Recognition memory effects, such as frequency and strength, are debated between criterion-shift and distribution-shift accounts.
• Traditional Type-1 signal detection indices often fail to distinguish between these competing models.
• Metacognitive monitoring, assessed via Type-2 signal detection receiver operating characteristic (ROC) analysis, offers a potential solution.

Purpose of the Study:

• To investigate criterion- versus distribution-shift accounts of frequency and strength effects in recognition memory.
• To determine if Type-2 ROC analysis can discriminate between memory models that are ambiguous using Type-1 indices alone.
• To validate assumptions of Type-2 analysis by manipulating response bias.

Main Methods:

• Utilized Type-2 signal detection receiver operating characteristic (ROC) analysis to measure metacognitive monitoring.
• Conducted experiments manipulating word frequency (Experiment 1) and item strength via repetition (Experiment 2).
• Manipulated Type-1 response bias by varying response categories (Experiment 3) to confirm analytical assumptions.

Main Results:

• Experiment 1 showed a frequency-based mirror effect (higher hit rate, lower false alarm rate for low-frequency words).
• Experiment 2 demonstrated an increased hit rate but no change in false alarm rate with item strength manipulation.
• Type-2 ROC analysis successfully discriminated between criterion- and distribution-shift models, unlike Type-1 indices.

Conclusions:

• Type-2 signal detection analysis is a valuable tool for investigating recognition memory.
• This method provides insights into memory processes that are not discernible with traditional Type-1 analyses.
• Type-2 analyses, when combined with Type-1 methods, enhance the understanding of memory models.