Abstract
The present study aimed to replicate Slamecka and Graf’s (1978) findings on the phenomenon of the generation effect in a sample of undergraduate psychology students at Texas A&M University. Twenty-two undergraduate participants were randomly assigned to either a read or generate condition, in which they were presented with rule based trials (antonym, synonym, same category, associate, and rhyme) in randomized blocks. Participants assigned to the read condition were required to read the second word within a given word pair, while those in the generate condition were instructed to generate a related word. Both groups then completed recognition tests that consisted of three-word lists pertaining to the response words they provided. It was predicted that participants in the generate condition would display better recall of the second word during the recognition tests in comparison to participants in the read condition. Results revealed that memory performance was significantly greater for participants in the generate condition than for those in the read condition, regardless of rule type. These findings provide evidence for the generation effect amongst a new sample.
Evaluating the Generation Effect and how it Aids Learning and Memory
The phenomenon known as the generation effect refers to the finding that words generated through the active participation of a subject are remembered much more easily than words that are simply read. Since its introduction through the research of Slamecka and Graf (1978), other memory researchers have conducted their own variations of this experiment in an attempt to refine and further understand the concept. Research examining this effect uses a variety of material types, generation tasks, and memory tests. The purpose of the current research was to replicate the results of Slamecka and Graf’s (1978) past findings and extend them to a new population.
Past literature has found that the generation effect has a profound impact on learning, particularly when combined with mnemonic techniques (Craik & Lockhart, 1972; Rogers, Kuiper, & Kirker, 1977). Mnemonic technique variations have been used to examine the process of learning in students and are associated with the levels of processing theory, another explanation for the generation effect, which states that how long something is remembered is dependent on how it has been processed (Craik & Lockhart, 1972). Thus, memory is more efficient for words that are generated due to the deeper processing that takes place, in comparison to when words are supplied and read. Further research done by Rogers et al. (1977) found that self-referential processing, also known as the self-reference effect, relates to mnemonic generation in aiding memory. They proposed that words self-generated by the learner have personal meaning and importance, which explained the high likelihood of successful retention.
Research has also examined whether the generation effect is consistent among child samples. In a study conducted by McFarland, Duncan, and Bruno (1983), the generation effect was examined in the development of young children. Results from retention and recall tests indicated that the initial basis of the effect was attributed to “encoding orientation,” the processing of information in the brain for memory, of semantic or phonetic categories and the type of task assigned to the child. In relation to this finding, research done by Sophian and Hagen (1978) discussed that the development of memory performance is dependent on the changes of mnemonic strategies used in retrieval as the child ages. Furthermore, Zaslofsky, Scholin, Burns, and Varma (2016) evaluated the generation effect in children using a mathematical approach. Results from their study indicated that responses generated by the students for multiplication combinations played an important role for rehearsal intervention success.
Additionally, physiological research further explored the implications of the generation effect. Past literature suggests that the effect itself is the product of the cognitive process that takes place in the brain known as top-down processing. In their research of neural processes that prompt the generation effect, Rosner, Elman, and Shimamura (2013) assessed activations in the brain through fMRI data. Activation took place broadly throughout regions in the brain, but more significant generation activity was observed in the prefrontal cortex. This finding aligns with the previous research of Raye, Johnson, Mitchell, Reeder, and Greene (2008) who proposed that activity in the prefrontal cortex relates to the generation effect due to the cognitive processing of refreshing and updating that involve the use of generated information. In support of the notion that the generation effect is prompted by top-down processes, Beg, Snider, Foley, and Goddard (1989) evaluated the concept of item distinctiveness. They proposed that the process of active generation distinctively encodes items specifically to aid in memory later. Further research on the role of distinctiveness was done by Kinoshita (1989) whose findings suggest that personal generation according to a nonsemantic rule enhances a distinct encoding process.
The past literature has found the generation effect to be consistent amongst its profound effects on learning and its impact on the development of children, with further support through physiological findings. In the current research, we sought to replicate the foundational research of Slamecka and Graf (1978) amongst present-day college students. A mixed subjects design was used where task type (read vs. generate) was assigned randomly between conditions and rule type (antonym, synonym, same category, association, and rhyme) within conditions. It was hypothesized that students assigned to the generate condition would display better memory performance than those who were assigned to the read condition.
Discussion
The aim of the present study was to replicate the past findings of Slamecka and Graf’s (1978) research on the generation effect, which is the finding that words generated through the active participation of an individual are remembered much more easily than words that are simply read. In the current study, the analysis of college-aged students revealed a significant main effect of task type between the participants placed in the generate condition and participants in the read condition. These results supported our hypothesis that participants assigned to the generate condition would display better memory performance if they had generated the word than if they were assigned to the read condition and had only read the word. These results indicated that on the memory performance task the generate condition had significantly greater recall rates compared to those of the read condition.
The results of the current study were consistent with previous research done on the generation effect and successfully replicated the findings introduced by Slamecka and Graf’s (1978) study. Our results also support the findings of past literature and show that people learn better through generating in memory rather than through passive processes such as reading. The credibility of the generation effect as shown through our replication study demonstrates the consistency of not only the generation effect itself but also further extensions of the findings, including research pertaining to children. A study done by McFarland et. al (1983) demonstrated how the early development of young children is impacted by the generation effect. Through their active participation, these children did better on assignments and tests as a result of better recall. Similarly, a study was conducted by Rogers et al. (1977) on young adults in an introductory psychology course. Results of the study revealed that participants experienced more success in academic performance through the use of various mnemonic techniques. Both of these studies show the benefit of the generation effect on learning and allow for a better understanding of the processing that takes place in the brain, regardless of age differences. Additionally, physiological research furthered study over the implications of the generation effect. Rosner et al. (2013) observed the generation effect and its activation of brain areas through fMRI data on patients. They found that when an individual was to self-generate for memory, areas of the brain such as the prefrontal cortex, displayed much more activity in comparison to when there was no self-participation.
Although the results of our study replicated past findings, there were potential limitations. Our convenience sample consisted of a small sample size of twenty-two participants that were all undergraduate psychology majors enrolled in a psychology course. The participants of the study also displayed an uneven ratio of female to male students that could limit the generalizability of the findings. Another potential limitation of the current study was the use of English word pairings, which compromised the generalizability to other languages and cultures. Suggestions for future research would be to test the generation effect amongst different cultures and bilingual people. Further improvements include incorporating a more diverse population of interest through sampling people of varying SES backgrounds and different education levels.
The generation effect has been shown to be a significant factor in memory. Through past research, it has been found to be beneficial for learning in various age groups and has been observed through physiological studies. Additionally, the findings from our current study support the generation effect and have important implications for the understanding of how processing impacts learning and memory. These findings can be helpful for educational settings in teaching students the most efficient ways to learn and remember.