Utilizing retrieval practice has demonstrated superior and reliable learning gains compared to re-studying, a finding called the testing effect. Previous research has investigated cognitive factors, including working memory, fluid intelligence, and episodic memory, as potential individual differences in the testing effect, but less is known about personal attributes. In the present study, self-efficacy and enjoyment of thinking are explored as potential individual differences in the testing effect. Individuals with greater self-efficacy and enjoyment of thinking might benefit more from retrieval practice, as these students should apply greater effort and persistence when learning complex information, resulting in better performance in a test requiring higher order thinking. Participants will learn geology concepts via videos, and then either answer a question or read a study point for each concept. After the learning phase, a final test will be administered, consisting of application questions of the material. Established and validated scales of self-efficacy and enjoyment of thinking will be administered and multiple regressions will be used to examine the relationship between these personal attributes and final test performance. This study has implications for the mnemonic benefits of testing and potential individual differences in learning in a classroom setting.

The concept of transfer-appropriate processing posits that performance on a memory task will be optimized when the cognitive processes engaged during encoding match those required for retrieval. Although transfer-appropriate processing is potentially useful for explaining the mnemonic benefits of retrieval practice, relatively few studies have empirically tested this theoretical idea. In four experiments, we manipulated the type of processing (involving item-specific and relational processing) induced by both the retrieval practice task and the subsequent criterial test in order to investigate how a match in processing influences criterial test performance. The results of the four experiments generally support the idea that transfer-appropriate processing is important to observing the effects of retrieval practice. However, the predicted pattern of criterial test performance was not observed in full until Experiment 4, which showed better performance for matched processing. Additional analyses suggested that an underlying issue in Experiments 1-3 was the difficulty of creating retrieval practice tasks and criterial tests that predominantly induced only one type of processing.

Retrieval practice, or the act of bringing information to mind after its initial presentation, has been widely acknowledged as a robust cognitive strategy for learning new information. Previous research has explored various factors (e.g., the use of feedback, type of material) in a retrieval practice paradigm, providing evidence that there is an overall benefit of retrieval practice compared to re-study in the final test, a finding known as the testing effect. Individual differences in cognitive abilities and personal attributes amongst learners have been shown to influence the magnitude of the testing effect, but there is limited research, resulting in mixed findings. The mixed findings could be due to methodological or conceptual issues within the current published studies. Methodological issues in current research include dichotomizing the individual differences measure, administering multiple individual differences measures for each construct, and not reporting descriptive statistics and reliability values. Conceptual issues are present in previous work, as simple learning materials (e.g., word pairs) might not elicit differences in personal attributes that require greater engagement and persistence during learning. More research is warranted to determine to what extent individual differences in cognitive ability and personal attributes influence the benefits of retrieval practice, specifically in complex learning material and in classroom settings.

Individual differences in cognition are present for younger and older adults, as vocabulary and general knowledge increase with age while working memory and processing speed decline with age. Baseline abilities, including working memory, vocabulary and self-report factors of self-efficacy, memory errors and need for cognition, predict cognitive training performance. In our current study, we investigated how baseline abilities from computerized tasks and self-report surveys predict gain in cognitive training. Three hundred and twenty four younger and older adults were randomly assigned to train on either a working memory or general knowledge task for at least 10 sessions. There was a significant interaction between age and treatment in hierarchical regression analyses, as younger adults gained more in the working memory training and older adults gained more in the knowledge based training. Vocabulary baseline ability predicted training gain, but there were no significant associations with working memory baseline. Self-report factors did not predict training gain, but late learning slope was positively correlated with training gain in younger adults.

This study investigates whether people become underconfident in their memory abilities when assessing their metamemory, or awareness of their memory capabilities. Previous research has investigated the Underconfidence-With-Practice (UWP) effect in the context of various manipulations of metamemory tasks, where participants become underconfident in their memory abilities with task experience. Importantly, the UWP effect demonstrates this disconnect between participant’s metamemory and performance. We examined the UWP effect in a novel, free recall metamemory task. The novel metamemory task showed participants one word at a time, in which they placed a judgment of learning (JOL) on how likely they thought they were to remember the word. After being shown 30 words, the participant freely recalled as many words as they could remember. In the first experiment, we induced a betting paradigm with feedback. We compared the JOLs and accuracy for each block of words and could not replicate the UWP effect in our novel paradigm. In the second experiment, JOLs were labeled as confidence ratings and no feedback was given to participants to replicate previous research. The UWP effect was still not present, as participants never became underconfident throughout the blocks of words. Since our manipulations did not yield the UWP effect, the UWP effect might not generalize to all metamemory paradigms.