Commentary 1: Neural Imaging to Track Mental States while using and Intelligent Tutoring System
Researchers have been on the frontline to try and determine what goes on in a student’s brain as they try to solve problems, in this case mathematical problems. The article offers insight on the capacity of utilizing neural imaging to aid student modeling in intelligent tutoring systems. The intelligent tutoring system is a computerized application that notes the activities of the brain as one engages in problem solving activities following data collected from the actions that the student opts to choose. The study conducted was significant as a continuation of research carried out on intelligent tutoring systems since it proved effective in improving student mathematical problem solving. Functional Magnetic Resonance Imaging (FMRI) data were collected while students interacted with an intelligent tutoring system that taught an algebra isomorph and the resulting hemodynamic measures of the brain were used to interpret one’s mental state during problem solving using a hidden Markov algorithm.
The psychological aspect of the study involves a cognitive model that predicted the distribution of solution derived from different degrees of problem complexities. As established, when we thing, different parts of our brain become active as we try to find the best alternative for any given situation. In addition, a linear discriminant analysis used the collected data from FMRI to predict whether or not the students engaged in problem solving. The results of the study that registered a high percent accuracy illustrated the importance of putting together the bottom-up information obtained from FMRI with the top-down information obtained from the cognitive model which can be achieved via two tutor based approaches. The model-tracing algorithm must be restricted on the basis of pilot data and then be used later in a desired situation. The model tracing algorithm must provide an actionable diagnosis in real time as opposed to waiting until all data is collected so as to deliver a diagnosis because models are designed for interpretations at the exact moment when the problem solving activity is taking place. Knowledge tracing can be obtained at a later period because it utilizes the diagnosis of current student problem solving to select later problems. After all, once we learn something new, the knowledge gained is usually for application on a later date either to a similar or different situation.
The study can be used to further advance problem solving capabilities since it indicated that was possible to merge cognitive model and brain imaging to provide a fairly accurate diagnosis of a student’s mental state in episodes that last up to ten minutes. More advanced student models could be used to track specific student difficulties that could lead to more definite state distributions. Furthermore, improved brain imaging interpretation would result to greater signal separation hence accurate diagnosis. However, studies need to be carried out further as to why brain imaging is different in several aspects of several students even as they try to tackle the same problem. Other researchers suggest that several factors influence brain activity at the time of imaging such as nervousness or other psychological issues; however all these limitations need to be integrated into a careful manner to come up with an accurate diagnosis. The results can be applied to other situations in everyday life where people are forced to think and not only to solve mathematical problems.
Commentary Two: Improving Student’s Long-term Knowledge Retention through Personalized Review.
In psychology, research studies indicate that knowledge retention in humans is imperfect. In most cases, people remember what they studied recently, and if enough revision is not done, they tend to forget the already stored information. The article, therefore, reports on a study conducted using a systematic and personalized review approach in combination with statistical techniques for deducing individual differences with memory based on psychological theory. The methodology was applied in a middle school foreign language course that lasted a semester using retrieval practice software. The results from cumulative exams administered at the end of the semester were compared to time –matched review strategies. Results indicated that there was a 17% increase in course retention over current course content and a 10% improvement over standardized exams strategy for spaced study usually based on exams administered at the end of the semester.
Newly acquired information is susceptible to being forgotten despite the nature of material or skills being taught and the student’s background or age. This assumption was proved in a memory study conducted on students from kindergarten and those in college, or those studying in remote or urban areas. Investigations reveal that forgetting is influenced by the temporal distribution of study that characterizes institutions of learning. It is noted that although spaced practice of what is taught and learnt in class is of significance in many tasks beyond mere memorization and indicates potential in improving educational end results, most academic systems have a reward structure that does not provide an incentive to methodologically revise what was previously studied. The strategy of teaching involves facilitators introducing course material in sections and evaluations are done after the completion of each section. This streategy, therefore, implies that students will concentrate on what was recently taught to get good grades in a practice of memorization to deliver what is required but not to necessarily keep it for future use. This methodology although optimal for short term learning goals, it is costly for long term goals of maintaining knowledge and skills.
The approaches used in the study involved classroom-based educational content that compared massed with spaced presentation of material, and laboratory based investigation whereby a student is selected for material to study on the basis of past study history and performance in what was referred to as adaptive scheduling. Although previous studies on human memory indicated that the same can be improved via timely reviews, the present study demonstrate that incorporating personalized review software in classrooms yield significant improvements in long term educational end results. The study was an improvement of previous studies in that there was a defined time frame, covered entire course content and introduced material in a stumbled manner in line with other course activities. The aspect of experimental manipulation was limited by uncontrolled exposure from other classroom instructions such as homework and textbook information. The research reported on the article was significant in that the issue of forgetting educational content can be minimized by engaging a digital tool that is practical and time efficient at all levels.
A question to ponder on, however, is that do the findings of the study promote long term memory of information and skills acquired in school for application in further endeavors such as employment? The answer is no, just as the study indicated forgetting as a limitation. Further research should, however, be carried out to link why certain skills once studied it is not easy to forget such as swimming and driving yet the same does not apply to information.
