Unit 2 Essay Questions Flashcards
Discuss the 3-network model of attention, focusing on the function of each of the networks (Alerting, Orienting, and Control). Explain why a deficit in each (or multiple) of these would negatively impact learning. Choose one of these elements of attention and discuss the neural correlates (make sure to give an example and name a brain region).
The 3-model network of attention is represented in by the attention networks test (ANT) and includes the Alerting, Orienting, and Executive Control networks, each with different functions and neural correlatives. The alerting network plays a role in the top-down and bottom-up pathways of attention and is active in the “cueing” phase of ANT when there is a center cue or double cue, quite literally alerting the participant that something is about to happen but without spatial direction.
The orienting network plays a role in the overt and covert attention processes and is also active in the cueing phase of ANT, but it provides directional information about where something is going to happen, orienting the participant. These two networks typically work together for spatial cueing, with alerting providing the “something is happening” info and the orienting providing the directional info.
The executive control network relates to endogenous vs. exogneous attention and is active in the target analysis phase of ANT, helping us determine where to focus when there is a conflict (determing direction of middle arrow).
Each of these conceptions of attention can be individually analyzed via fMRI during the ANT. For example, fMRI shows increased activity in the PFC and anterior cingulate cortex (ACC) during this processing.
Explain the components of Baddeley’s model of working memory (right). Based on what we have learned, predict whether WM involves one network of brain regions or several. Make sure to give an example and name at least one brain region in your answer.
In Baddeley’s model of working memory, the central executive system (PFC) oversees all of working memory. Specifically, it oversees the fluid systems and the crystallized systems. Fluid systems are the WM components and are unstable (not consolidated), and can be described as follows: WM held in the visuospatial sketchpad is visual WM, phonological loop is auditory WM, and episodic buffer is the part of WM that tries to put sounds and visual input into context, binding these two types of WM together. Sufficient rehearsal of these fluid systems results in memory consolidation, which is represented by crystallized systems, which are stable. These systems represent longer term memory and are stabilized versions of the fluid components, depicted in the form of visual semantics, episodic LTM, and language, and info is interchangeable between them.
Working memory can be represented as a distributed frontal parietal network that encompasses a range of brain regions rather than a single, localized region. Specifically, WM relies on the DLPFC region as well as parts of the parietal lobe (e.g., PPC). This is seen via fMRI analysis, which reveals an increase in activity within these regions as N-back (test of WM) difficulty increases. This same paper also discusses the idea that frontal lobe lesion patients perform poorly for 2+ back tests, further implicating the frontal lobe in WM.
We discussed a number of findings that initially seemed to show that attention and working memory can be trained. However, these are not commonly included as targets for imporvement in current classroom practice. Explain why not, building on the notions of “near” and “far” transfer discussed in class. What do we know now about transfer of attention and WM training?
Initial studies did in fact show that WM interventions are effective at improving WM as a whole. This was demonstrated in 3 studies, the most promising perhaps implicating 42 children in the bottom percentile of WM abilities. When adaptive WM training (increasing difficulty) was applied amonst these students, a significant increase in WM performance was seen, resembling full remediation. However, each of these studies, including the last, is not sufficient at proving that WM can be trained because they do not implicate “far” transfer, or transfer to different domains. That is, these WM interventions helped individuals get better at the games/skills provided in the intervention period, but not necessarily get better with real-world WM tasks. This is also shown via the Lumosity study. Near transfer, but training on Lumosity shows zero improvement on tasks that are different from the apps, i.e., zero far transfer. To provide further concrete evidence, studies also show that this is not just true for WM, and that application of an active control in WM, chess, and music training shows no effect on chess/music/WM abilities, again indicating the lack of far transfer that these studies so desire.
We now know that there is a weight of evidence against WM training and other “far” transfer approaches, coinciding with an influx of evidence in favor of specific-study strategies for LTM.
Patient H.M. was the subject of major neuropsychological investigations on long term memory. What could he do, what couldn’t he do? Explain how modern science understands patient H.M.’s ability to exhibit implicit learning or memory even though he was incapable of consolidating new experiences in long-term declarative memories. Make sure to give an example and mention a brain region.
Patient H.M. was a man who had entire medial temporal lobe, including hippocampus, removed, resulting in anterograde amnesia (old memory intact, can’t form new memories) but no impairments to perception and reasoning. Specifically, his declarative memory was lost, but nondeclarative/implicit memory (procedural memory) and STM/WM were intact. For example, H.M. could learn new procedural implicit tasks, which was proven in that his performance on “star tracing” study in a mirror improved just as it did in control subjects; however, he could never remember having ever done the task before due to impaired declarative memory.
This is explained in modern science by the fact that STM/WM involve a frontal-parietal network, especially the DLPFC, which was not removed in H.M. However, the hippocampus and nearby regions in the medial temporal lobe are crucial for the consolidation of facts and events into LTM, and their removal meant H.M. could no longer form new memories in this regard. Implicit memory does not rely on these regions, hence it remained intact in H.M.
Explain how studying helps move information from short to long term memory. Keeping this process in mind, describe two effective study strategies and one effective teaching strategy. Be sure to provide evidence from the readings/lecture for these strategies, and explain how you could use the things we’ve covered in lecture about the spacing and testing effects to improve exam grades.
Studying moves information from short to long term via both rehearsal as well as a model for levels of processing. A study of LTM showed that increasing rehearsal time had no change on the primacy effect in the free recall task, but that regulating rehearsal led to a decrease/elimination of the primacy effect. This indicates that rehearsal is part of the equation but is not sufficient on its own. An older study proposed an alternative model of memory implicating levels of processing in long term memory consolidation. In this study, it was shown that questions demanding deeper processing (such as the meaning of the word rather than sound or print) led to significantly heightened memory capabilities. Moreover, effortful action (such as in rehearsal/retrieval) has shown to increase PFC activity, correlating with an increase in memory as well. Given this, effective study and teaching strategies can be formed.
First, an effective study strategy is to space out your studies by the length of time you desire to remember the information. The spacing effect is the most highly replicable psychological science study, with one study from an art history class dividing students into massed vs. spaced studying groups. In each test block, students on the spaced protocol scored much higher than those on the massed protocol. Another study showed that individuals who studied with a 7 day test delay scored similarly with a 4 week test delay and vastly outperformed a 0 day test delay.
Another effective study strategy involves the testing effect. A study took two prose passages and divided them into 30 units for scoring purposes, then assigned six groups of 30 people to either a study/study group (e.g., study —> break —> study) or a study/test group (e.g., study —> break —> test). It was shown that while a 5 min delay yielded the best scores from the study/study group, the study/test group significant;y outperformd them after both 2 days and one week, inidicating the impact of the testing effect on long term memory.
While the spacing effect is the most promising body of work in regards to improved teaching practices, interventions such as “Everyday Math” make me seek to veer away from this direction; however, given the substantial evidence depicted above for the testing effect, I believe it can be effectively applied within the realm of education. Specifically, I think it should be implemented via both formative and summative assessment.
We learned that a number of studies support the testing effect, including some that showed the neural mechanisms involved in long-term memory retrieval. Describe one example from the lecture or reading that provides both behavioral and neural evidence of the testing effect (making sure to name a brain region and mention methodology). Also, explain what happens in the brain when you study and take a test.
The testing effect has proven to enhance learning and memory consolidation from both a behavioral and neural perspective. A study by Roediger & Karpicke presented participants with two prose passages, dividing each for scoring purposes. Groups were formed on the basis of study/study and study/test (e.g., study –> 5 min. break –> test), and tests were performed with intervals of 5 min., 2 days, and 1 week. With a 5 min. retention interval, it was observed that the study/study group (no testing) scored highest on the test; however, at both the 2 day and 1 week intervals, the study/test group vastly outperformed the study/study group, unveiling a potent role of testing in the consolidation of information.
As for the neural basis of the testing effect, Nelson et al. subjected participants to four phases, an initial study phase, then a test phase, then a final study phase, then a final test phase. Each of the study phases involved fMRI scans to analyze the testing/retrieval effect, and it was shown that there was decreased PFC after the initial study session, showing the role of the PFC in effortful input of information; moreover, there was increased IPL activity after the initial study session, implicating the IPL in “feeling familiar.”
As this applies to studying and taking a test, assuming you were well prepared, the PFC is active during studying new terms and is less active during studying known terms and is less active during exams. Conversely, during studying the IPL is less active with new terms, for the IPL represents feelings of familiarity, but during a test, activity is heightened. According to van den Broek et al., the medial temporal lobe is also active during testing, while the DLPFC and VLPFC are more active during studying.
We discussed the parieto-frontal integration theory (P-FIT) of intelligence, and noted that the same parieto-frontal networks that are implicated in attention and working memory are implicated in attention and working memory are also critical in the P-FIT model. Based on this, would you consider intelligence to be domain general or domain specific? In your answer, describe an example into each of these three abilities showing how this ability works and which brain regions were involved.
P-FIT is a theory of intelligence that proposes that the neural basis of intelligence can be represented by an interplay of several brain regions distributed within the frontal parietal network in the brain. This same network is also critical for attention and working memory, thereby suggesting that intelligence is a domain general function that involves the integration of a multitude of brain areas. As described by Duncan in a 2013 paper, this network is “multi-demand,” and comes up again and again as critical for domain general abilities.
Studies involving fMRI analysis have shown that increasing difficulty on the N-back test of working memory is correlated with increased activity in the DLPFC as well as parts of the parietal lobe (e.g., PPC). Moreover, studies of attention utilize the attention networks test (ANT) to provide evidence that each of the three attention networks have a neural basis correlated with this same frontal parietal network:
Alerting: active during the cueing phase, uses brain arousal to indicate a state of readiness, that “something is about to happen,” involves the DLPFC, TPJ, Insula, Thalamus, and Locus coeruleus
Orienting: can be overt or covert and refers to the directional allocation of attention. From alerting, we know something is about to happen, and orienting networks help us direct our attention to where this thing is about to happen. Involves FEF, SPL, Cerebellum, and superior colliculus
Executive control: used to resolve the conflict that is target analysis in ANT (deciphering direction of middle arrow). Involves PFC and ACC.
We know that self-reported preferred learning styles exist, but there is currently a lack of empirical evidence for the “meshing hypothesis.” Define this hypothesis, then using the Pashler et al. study in your readings, describe what would count as acceptable evidence in support of the meshing hypothesis. Then, discuss the evidence that people do have preferred ways of encoding information (e.g., visual/verbal preferences), and convert input to their preference. Be sure to include both neuroimaging results (name a brain region) and the educational implications of the “conversion hypothesis.”
The meshing hypothesis is the idea that instruction via an individual’s preferred learning modality correlates with the largest correlation in actual learning ability and consequent performance. According to the Pashler et al. study, a specific criteria must be met in order to argue for this hypothesis. This criteria includes: 2 distinct learning modality groups (e.g., visual and auditory), division of individuals between these 2 modality groups, and a result that reveals a clear correlation between individuals’ preferred method of learning and performance. That is, preferred visual learners assigned to study via visual materials must significantly outperform visual learners assigned to study from auditory materials. Only one of hundreds of studies have argued in favor of this hypothesis, but given the need for strict study design (proper crossover), we shouldn’t be so quick to over-interpret these negative results.
This begs the question of how people with preferred learning styles are still proven to (in general) perform the best via visual learning. While the meshing hypothesis may not be true, the conversion hypothesis likely explains why. The conversion hypothesis involves cognitive styles, and suggests that individuals are of either verbal cognitive style or visual cognitive style. Verbal cognitive style involves generating verbal labels even when the stimuli is visual, and visual cognitive style involves generating visual imagery even when the stimuli is verbal. A 2009 study by Kramer et al. investigated this hypothesis, presenting participants with a shape or set of words, then asking them to choose which of the following shapes/sets of words was most similar. It was found that when “verbalizers” were presented with the shape images, there was increased activity in the left language areas (left SMG), indicating the process of converting visual stimuli into verbal labels and confirming the conversion hypothesis. Moreover, “visualizers” expressed increased activity in the right fusiform area when presented with verbal stimuli, again indicating conversion.
This hypothesis was also studied via rTMS, in which TMS was used to inhibit the left SMG as well as the vertex (control) in visualizers following preliminary MRI scans. It was shown that the more verbal an individual claimed to be, the more activity in left supramarginal gyrus (SMG) went down during a task
with TMS compared to control.
The educational implications of the cognitive hypothesis are that learning styles are a complete myth, and that any learning style is acceptable because individuals of the respective cognitive styles will simply internally convert information into their preferred learning method.
