The purpose of the experiment was to determine which Learning Method was the most effective for short- and long-term memory formation. Method efficacy was tested via a Short-Term Memory (STM), and Long-Term Memory (LTM), glyph recall test. There were four groups, each using a different Learning Method which varied in two factors: written repetition (10X/none) and movie viewing (before/after STM test).
The Learning Method was used to introduce the symbols, and there was a short maths test before the STM test. Symbol recall was assessed again in the LTM test, three weeks later. As the LTM test was taken after the STM test, all students had watched the movies.
Analysis suggested that Learning Method 1 produced higher scores; students who wrote the symbol 10X and viewed the movie prior to taking the STM test had significantly greater glyph recall compared to those who used the other methods (p < 0.05).
While Learning Method 1 appeared to be the most effective, it is possible that the results were affected by experimental design flaws; notably, the non-standardised test conditions.
The degree of symbol retention demonstrated on the LTM test may be related to memory consolidation, which is aided by hippocampal ripple oscillations.
The Learning Methods for each group were as follows (for the method code definitions and full method, see the Appendix):
Group 1 (Learning Method 1): WB-1X_MB_10X
Group 2 (Learning Method 2): WB-1X_MA_10X
Group 3 (Learning Method 3): WB-1X_MB
Group 4 (Learning Method 4): WB-1X_MA
The independent variable is the Learning Method and the dependent variable is the number of symbols recalled correctly on the memory tests (STM and LTM). Method 1 was the most involved (writing 10X, and watching the movie beforehand). The alternative hypotheses are:
Ha: if Short-Term glyph recall is related to the Learning Method (Method 1, 2, 3, 4), then students in Group 1, who used Method 1 (WB-1X_MB_10X), will recall a greater number of symbols correctly on a Short-Term Memory test.
Ha: if Long-Term glyph recall is related to the Learning Method used for the Short-Term Memory test (Method 1, 2, 3, 4), then students in Group 1, who used Method 1, will recall a greater number of symbols correctly on a Long-Term Memory test.
For the STM data, Levene’s test established that there was an effect of variance (p < 0.05). As the variance is significant, a Welch one-way ANOVA and Games-Howell post-hoc test must be used to test the hypothesis. The Welch one-way ANOVA established that STM test scores differed as a function of Learning Method used [F(3, 124) = 13.1230, p < α = 0.05]. A Games-Howell post-hoc test revealed that STM test scores of students who were in Groups 1 (16.3125 ± 1.9582, p <0.05) and 2 (15.9688 ± 1.7271, p <0.05) were significantly higher than those who were in Groups 3 (13.8750 ± 2.5368) and 4 (13.4844 ± 2.5128). There was no significant difference between the STM scores of students in Groups 1 and 2 (p = 0.8790) and those in Groups 3 and 4 (p = 0.9260).
For the LTM data, Levene’s test established that there was no effect of variance (p > 0.05). As the variance is not significant, a One-way ANOVA and Tukey post-hoc test can be used to test the hypothesis. The One-way ANOVA established that LTM test scores differed as a function of Learning Method used [F(3, 124) = 66.0280, p < 0.05]. A Tukey post-hoc test revealed that LTM test scores of students who were in Group 1 (12.7500 ± 2.8905, p <0.05) were significantly higher than those who were in Groups 2 (7.6250 ± 2.6397), 3 (6.8281 ± 2.6537), and 4 (3.6875 ± 2.2388). There was no significant difference between the LTM scores of students in Groups 2 and 3 (p = 0.6160).
The results support the alternative hypotheses that if glyph recall is related to the Learning Method used, then students in Group 1 (who used Method 1) will recall a greater number of symbols correctly on both a STM and LTM test. As the test scores for students who used Method 1 were significantly higher in both the STM and LTM tests, at a significance level of 0.05, the alternative hypothesis is favoured. By adopting the alternative hypotheses there is a possibility of Type 1 error in both cases.
The hippocampus contributes critically to memory formation, organisation, and storage Memory consolidation, a process that transforms newly acquired information into long-term memory, also depends on the hippocampus. Through consolidation, labile newly formed memory traces are progressively strengthened into long-term memories and become more resistant to interference. However, it is suggested that they remain susceptible to updating and modification
The hippocampus generates high-frequency ripple oscillations in local-field potentials (LFPs), observed most prominently in the hippocampal CA1 pyramidal layer Ripples participate in strengthening and reorganising memory traces, possibly by mediating information transfer to neocortical areas Memory traces are represented by assemblies of principal neurons that are activated during ripple-associated network states
There is evidence suggesting that memory consolidation is enhanced during sleep and resting (“off-line”) states Sleep is a state which optimises the consolidation of newly acquired information in memory, depending on the specific conditions of learning and the timing of sleep It induces long-lasting cellular and network modifications responsible for memory stabilisation
A proposed neural mechanism for sleep-dependent memory consolidation, is reactivation of awake experience (neuronal replay) in the hippocampus which is associated with sharp wave-ripple (SPW-R) events that occur primarily during off-line states SPW-Rs are “aperiodic, recurrent instances of large deflections (sharp waves) in the hippocampal LFP”, and they are associated with synchronous fast-field oscillations (ripples)
During SPW-R events, hippocampal cell firing closely follows the pattern that took place during the initial experience Theta (4-8 Hz) oscillations and ripples (~200 Hz) occurring during sharp waves may mediate encoding and consolidation, respectively. Pyramidal neurons replay previous waking activity in a temporally compressed manner, thus reactivated firing patterns occur within shorter time windows propitious for synaptic plasticity within the hippocampal network and in downstream neocortical structures. Slow-wave sleep (SWS) supports system consolidation and rapid eye movement (REM) sleep supports synaptic consolidation through specific patterns of neuromodulatory activity and electric field potential oscillations. During SWS, there is a diminution in cholinergic activity and the ripples stimulate the redistribution and transfer of hippocampus-dependent memories to the neocortex The thalamocortical spindles generated by the thalamus arrive at the neocortex at the same time as the hippocampal memory information, due to the slow oscillations which facilitate the transfer, and this synchronisation is thought to be vital to the long-term storage of memories within neocortical networks
During REM sleep, at high cholinergic and theta activity, local increases in plasticity-related immediate-early gene activity may promote synaptic consolidation of memories in the cortex
Incoming signals move through the hippocampus via a ‘trisynaptic loop’ consisting of synapses between principal cells in the dentate gyrus (DG), CA3 and CA1 Hofer et al investigated the cellular and network properties of SPW-Rs with simultaneous laminar multielectrode and intracellular recordings in a rat hippocampal slice model. Spontaneous SPW-Rs were generated in the DG, CA3, and CA1 regions During the memory encoding phase, the hippocampus binds neocortical representations to local memory traces. Then, during the off-line periods, the traces are concurrently reactivated in the hippocampus and cortex to potentiate the corticocortical connections underlying stored representations
Studies show that disruption of ripples during post-learning SWS impairs memory consolidation and learning In an experiment conducted by Ego-Stengel and Wilson rats were trained daily in two identical tasks, each followed by a one hour rest period. Following one of the tasks, neuronal activity associated with ripple events was disrupted, without changing the sleep-wake structure, via selective stimulation of hippocampal afferents. It was found that the rats learned the control task significantly faster than the task followed by the stimulation, which suggests that interfering with hippocampal processing during sleep led to decreased learning Similarly, Nokia et al. found that disrupting hippocampal ripples using electrical stimulation either during training in awake animals, or during sleep after training, had a negative impact on learning
A study by Wang et al indicated that the median raphe region (MnR) is important for regulating hippocampal ripple activity and memory consolidation. A fear conditioning procedure was used to determine this relationship, via interruption of ripple activity. Simultaneous in vivo recording in the MnR and hippocampus of mice showed that, when a group of MnR neurons was active, ripples were absent; ripple activity was related to the activity of MnR neurons. Additionally, MnR may regulate memory consolidation via its projections to thalamocortical regions, which facilitate interactions between the hippocampus, thalamus and cortical regions during SWS.
Under the assumption that there was little deliberate reactivation of memory traces for the symbols in the three weeks following the STM test, the degree of retention of the symbols on the LTM test may be related to memory consolidation via hippocampal ripples occurring during sleep and rest.
There were a number of limitations which may have affected the validity of the results. The major limitation was the non-standardised test conditions. Since different groups took the test at different times of the day, and different individuals have performance peaks at different clock times randomisation of subjects is important.
This also leads to the possibility that students in an earlier group may have informed others of the symbols or experimental procedures. Knowledge of the tests could have influenced the students’ concentration, with those who knew being more likely to apply themselves to the Learning Method. While all students would be likely to undergo some degree of memory consolidation following the STM test, those who were aware of the experimental design would likely have greater retention. If all students took the test at the same time under standardised conditions, it would be less likely for this issue to affect the results.
Additionally, the small sample size, and the fact that the participants were all students of one course (Neuroscience), also means that the external validity of the experiment, and thus the generalisability to groups other than the experimental group, cannot be established. There was also only one group per Learning Method, so it is unknown if the results are repeatable.