Why Empirical Forgetting Curves Deviate from Actual Forgetting Rates: A Distribution Model of Forgetting.

Area of Science:

• Cognitive Psychology
• Neuroscience
• Educational Psychology

Background:

• Forgetting research traditionally models recall decay using power or exponential functions.
• Empirical forgetting curves are often assumed to directly reflect individual memory forgetting rates.

Purpose of the Study:

• To propose and validate a memory distribution model explaining the divergence between forgetting curves and item forgetting rates.
• To investigate the influence of memory strength distribution on recall performance over time.

Main Methods:

• Development of a theoretical distribution model of memory strengths relative to a recall threshold.
• Analysis of existing evidence and a new experiment involving participants learning word pairs to a criterion.

Main Results:

• The distribution model explains why forgetting curves can appear linear or concave, especially at high percentages of correct recall.
• Empirical data supported the model's predictions regarding the relationship between memory strength distribution and forgetting curves.

Conclusions:

• Forgetting curves are influenced by the distribution of memory strengths, not solely by individual memory decay rates.
• Educational strategies should consider memory strength distribution for long-term retention, as memories just above threshold are not durable.