Abstract: As the population of a society ages and evolves, the cognitive capacity and life-long learning ability of older adults is extremely significant. The current study aims to investigate whether interpolated testing can effectively improve the abilities of older adults to learn and retain new information. In other words, this study targets to verify whether the forward testing effect (FTE) can be applied to older adults comparing to the younger participants in previous research. Previous research on younger adults and older children (rather than younger elementary school children) indicates that the FTE in learning of single items is primarily due to the release from proactive interference (PI). Research on younger elementary school children indicates that they cannot reduce PI; thus, they do not have FTE in their early years as they are just learning. However, this finding cannot explain why people have the FTE during complex learning, which is also attributed to the reduction of mind-wandering (MW). Other research demonstrates that learning engagement plays a role to make a great difference between PI and MW amongst older and younger adults. Will older adults (aged over 60) experience FTE when learning simplified things? How about complex materials?
The current study conducted three experiments to research the above questions progressively. All participants (older adults) were randomly divided into two groups using a computer. The tested group was then given interim tests, and the other (control) group went without any test until the last list or segment. Finally, both groups took a memory test on the things they have encountered through the learning process. The size of these groups was determined according to the effect size in previous studies. The number of participants in each experiment was 30, 32 and 49. In experiment 1, both groups were given five lists of words. The test group was given an interim test after studying each of the five lists, whereas the control group was required to solve the math problem after studying each list, where they only took an interim test on list 5. In experiment 2, the participants had five lists of common supermarket items to learn. The test group was also given an interim test after studying each of the five lists. However, the control group only had to restudy lists 1-4, and then, they were given an interim test after studying the fifth list. In experiment 3, the participants were instructed to study a four-segment lecture video where the instructions and experimental procedures were the same as those in experiment 2. All the participants in the three experiments mentioned above were over 60 years of age.
The result of experiment 1 shows that the test group (M = 5.87, SD = 1.21) recalled approximately twice as many correct words as the math group (M = 3.20, SD = 2.01) in the list 5 interim test (difference = 2.67 words, 95% CI = [1.02, 4.32], t(28) = 3.31, p = 0.003, Cohen’s d = 1.21). The result indicates that interpolated tests facilitate older adults’ learning of new single items. For the test group, the repeated measures ANOVA, with lists 2-5 as the within-subjects variable, showed that PI linearly increased across lists (F(1, 14) = 8.15, p = 0.01, η_p² = 0.37). This finding indicates that interim tests could not completely prevent the build-up of PI across lists for older adults. Notably, the math group (M = 3.40, SD = 1.59) experienced approximately three times as much PI as the test group (M = 1.20, SD = 1.21) in the list 5 interim test (difference = 2.20 words, 95% CI = [1.14, 3.20], t(28) = 4.26, p < 0.001, Cohen’s d = 1.56). The result indicates that interim tests prevent the build-up of PI for older adults. In Experiment 2, the test group (M = 5.13, SD = 1.41) recalled approximately twice as many correct words as the restudy group (M = 2.38, SD = 1.63) in the list 5 interim test (difference = 2.76 words, 95% CI = [1.64, 3.88], t(29) = 5.03, p < 0.001, Cohen’s d = 1.80). The repeated measures ANOVA showed that, for the test group, PI linearly increased across lists 2-5 interim tests (F(1, 14) = 20.70, p < 0.001, η_p² = 0.60). The restudy group (M = 3.06, SD = 2.11) committed approximately three times as much PI as the test group (M = 0.93, SD = 0.96) in the list 5 interim test (difference = 2.13, 95% CI = [0.91, 3.35], t(29) = 3.57, p = 0.001, Cohen’s d = 1.30). The result of Experiment 3 shows that the test group (M = 6.32; SD = 3.11) recalled approximately twice as many correct items as the restudy group (M = 3.50; SD = 2.17) in the Segment 4’s interim test (difference = 2.82 items, 95% CI = [1.28, 4.36], t(47) = 3.68, p = 0.001, Cohen’s d = 1.05). This finding reveals that interpolated testing enhances the learning ability of older adults for new complex materials.
The results show that interpolated testing effectively enhances older adults’ learning and retrieval of new information. The test groups scored higher than the math and restudy groups in learning of single items and complex materials. The results support an integrated theory of interference reduction and learning engagement, which can comprehensively explain the FTE difference amongst younger children, older children, younger adults and older adults. Therefore, we can use FTE as a effective method to modify the learning and memorising condition of older adults.

Key words: interpolated testing, forward testing effect, older adults, learning engagement, interference reduction