CoChat
Sign Up
Cited by 137Intelligent Tutoring Systems and Adaptive LearningEEG and Brain-Computer InterfacesNeuroscience, Education and Cognitive Function

Learn Piano with BACh: An Adaptive Learning Interface that Adjusts Task Difficulty Based on Brain State

Open access
Yuksel, Beste F., Oleson, Kurt B., Harrison, Lane, Peck, Evan M., Afergan, Daniel, Chang, Remco, Jacob, Robert JK·2016·Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems
DOI 10.1145/2858036.2858388
Open this paper in CoChat — free accountOpen PDF

Abstract

We present Brain Automated Chorales (BACh), an adaptive brain-computer system that dynamically increases the levels of difficulty in a musical learning task based on pianists' cognitive workload measured by functional near-infrared spectroscopy. As users' cognitive workload fell below a certain threshold, suggesting that they had mastered the material and could handle more cognitive information, BACh automatically increased the difficulty of the learning task. We found that learners played with significantly increased accuracy and speed in the brain-based adaptive task compared to our control condition. Participant feedback indicated that they felt they learned better with BACh and they liked the timings of the level changes. The underlying premise of BACh can be applied to learning situations where a task can be broken down into increasing levels of difficulty.

Integrity signals

Registry

Crossref · 10.1145/2858036.2858388

Cites

56 of 56 references verified0 citation errors · 0 not checkable

Statistics

8 tests recomputed · 0 with the reported and recomputed p on opposite sides of .05

Claims the paper makes

0 of 48 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    The brain data effectively classified high versus low cognitive workload while participants played easy and hard pieces.

    “We first verify that the brain data demonstrated effective classification of high and low cognitive workload while users played easy and hard pieces on the piano.”
    Quote in the paperSubstantial claim
  • Findingreasoning

    All participants segmented the pieces while learning in the normal condition.

    “As a result, all participants segmented the pieces while they learnt during the normal condition.”
    • Also stated in: § Learning Task Real Time Classification · plain
      “As a consequence, all learners segmented the pieces while they learnt during the normal condition.”
    Quote in the paperNot a substantial claim
  • Findingown data

    Participants played the pieces faster in the BACh adaptive condition than in the control condition.

    “We found that learners played with significantly increased accuracy and speed in the brain-based adaptive task compared to our control condition.”
    • Also stated in: § INTRODUCTION · plain
      “By analyzing performance data, we found that participants played the pieces faster and with increased accuracy when using our BACh system compared to how they would normally learn a piece of music.”
    • Also stated in: § Performance Data · hedged
      “Results indicated a pattern for two traits: 1) Increased Accuracy, and 2) Faster speed with BACh.”
    Quote in the paperSubstantial claim
  • Findingown data

    Participants played with significantly greater accuracy in the BACh adaptive condition than in the control condition.

    “We found that learners played with significantly increased accuracy and speed in the brain-based adaptive task compared to our control condition.”
    • Also stated in: § INTRODUCTION · plain
      “By analyzing performance data, we found that participants played the pieces faster and with increased accuracy when using our BACh system compared to how they would normally learn a piece of music.”
    • Also stated in: § Performance Data · hedged
      “Results indicated a pattern for two traits: 1) Increased Accuracy, and 2) Faster speed with BACh.”
    Quote in the paperSubstantial claim
  • Findingown data

    The mean change in oxygenated hemoglobin was significantly higher when participants played hard pieces compared to easy pieces (t(29) = -2.42, p = .01).

    “This revealed a significant difference between conditions when participants played an easy piece (µ = -0.068,σ = 0.077) versus a hard piece (µ = -0.00005,σ = 0.124) on the piano (t(29) = -2.42,p = .01).”
    • Also stated in: § RESULTS fNIRS Data · plain
      “The mean change in oxy-Hb was significantly higher in participants when they played an hard piece than an easy piece (p = .01).”
    Quote in the paperSubstantial claim
  • Classified as probably not a substantial claim. 20
  • May combine two claims. 24

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    Eleven of 16 participants rated themselves as beginners; of the five intermediates, three no longer played piano; median piano-playing experience was 3 years.

    “While eleven out of 16 participants rated themselves as beginners, out of the remaining five who rated themselves as intermediate, 3 of them no longer played piano.The median time our participants had played piano was 3 years.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    Most participants began learning with their right hand only, while one participant started with the left hand.

    “Most participants started learning with their right hand only, one participant who was very good with his left hand started with his left hand first.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Findingown data

    When cognitive workload measurement began before the full level of difficulty, BACh incorrectly determined readiness to advance earlier because it read workload for an easier level.

    “We found that if we started measuring their cognitive workload earlier before the full level of difficulty, BACh would quite correctly think that they were ready to move on earlier as it was reading their cognitive workload for an easier difficulty level.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Participants spent less time on easier levels and longer on levels of increasing difficulty.

    “While participants spent less time on easier levels and longer lengths of time on levels with increasing difficulty, we did find individual differences in time spent within levels.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Participants' interview feedback about the timing of difficulty changes was generally positive, with no standalone or overtly negative comments about the timings.

    “their feedback was generally positive:”
    • Also stated in: § Interview Data · plain
      “Overall, there were no standalone or overtly negative comments about the timings of BACh.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Participants' subjective ratings showed a noticeable preference for BACh over the normal condition in felt mastery of the piece, felt correctness of playing, and felt ease of learning.

    “Figure 6 shows that participants showed noticeable preferences for BACh in how well they felt they mastered the piece, how correctly they felt they played, and how easy they felt the piece was to learn.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Good teachers often guide their students into the zone of proximal development.

    “Good teachers, whether they realize it or not, often guide their students into the zone of proximal development.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [51] 51. Lev Vygotsky. 1978. Interaction between learning and development. Mind and Society (1978), 79-91. 51. Christopher D. Wickens, Shaun. Hutchins, Thomas. Carolan, and John. Cumming. 2013. Effectiveness of Part-Task Training and Increasing-Difficulty Training Strategies: A Meta-Analysis Approach. Human Factors: The Journal of the Human Factors and Ergonomics Society (2013). DOI: http://dx.doi.org/10.1177/0018720812451994

      Not checked in the free run.

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    Cognitive load theory holds that working memory has a finite capacity, so instructional designers should avoid overloading students.

    “It is based on idea that there is only a finite capacity for working memory available [39,46] so designers of learning material should avoid overloading students.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [39] 39. Grega Repovs and Alan Baddeley. 2006. The multi-component model of working memory: Explorations in experimental cognitive psychology. Neuroscience 139, 1 (2006), 5-21. source

      Not checked in the free run.

    • cites [46] 46. John Sweller. 1999. Instructional Design in Technical Areas. ACER Press.

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Computer-based education has shown that students tend to learn more, take around 30% less time, and have increased enjoyment of their subject.

    “CBE has shown that students tend to learn more, take around 30% less time and have increased enjoyment of their subject [27].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [27] 27. James A. Kulik. 1994. Technology Assessment in Education and Training. Eva Baker and Harold F. O'Neil Jr, Chapter Meta-analytic Studies of findings on computer-based instruction., 9-34.

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Secondary tasks can interfere considerably with the primary task, especially if the primary task is complex and requires much of the learner's limited cognitive capacity.

    “An important disadvantage to this methodology is that the secondary task can interfere considerably with the primary task, especially if the primary task is complex and requires much of the learner's limited cognitive capacity [34,7].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [34] 34. Fred Paas, Juhani E. Tuovinen, Huib Tabbers, and Pascal W.M. Van Gerven. 2003. Cognitive load measurement as a means to advance cognitive load theory. Educational psychologist 38, 1 (2003), 63-71.

      Not checked in the free run.

    • cites [7] 7. Roland Brünken, Jan L Plass, and Detlev Leutner. 2003. Direct Measurement of Cognitive Load in Multimedia Learning. Educational Psychologist 38, 1 (2003), 53-61. DOI: http://dx.doi.org/10.1207/S15326985EP3801_7 source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Interpretationnone given

    The underlying premise of BACh can be applied to learning situations where a task can be broken down into increasing levels of difficulty.

    “The underlying premise of BACh can be applied to learning situations where a task can be broken down into increasing levels of difficulty.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    fNIRS has been demonstrated by HCI researchers to measure user cognitive workload in adaptive systems and is resilient to subject movement such as playing piano.

    “Brain sensing using functional near-infrared spectroscopy (fNIRS) has recently been demonstrated by HCI researchers to measure user cognitive workload in adaptive systems [4,44,1] and is resilient to subject movement as might be seen while playing a piano [43].”
    Quote in the paperSubstantial claim0 of 4 sources checked
    • cites [4] 4. Hasan Ayaz, Patricia A Shewokis, Scott Bunce, Kurtulus Izzetoglu, Ben Willems, and Banu Onaral. 2012. Optical brain monitoring for operator training and mental workload assessment. Neuroimage 59, 1 (2012), 36-47. source

      Not checked in the free run.

    • cites [44] 44. Erin Treacy Solovey, Paul Schermerhorn, Matthias Scheutz, Angelo Sassaroli, Sergio Fantini, and Robert J K Jacob. 2012. Brainput : Enhancing Interactive Systems with Streaming fNIRS Brain Input. Proceedings of the SIGCHI conference on Human Factors in Computing Systems (2012). source

      Not checked in the free run.

    • cites [1] 1. Daniel Afergan, Evan M. Peck, Erin T. Solovey, Andrew Jenkins, Samuel W. Hincks, Eli T. Brown, Remco Chang, and Robert J.K. Jacob. 2014a. Dynamic difficulty using brain metrics of workload. Proceedings of the SIGCHI conference on Human Factors in Computing Systems (2014), 3797-3806. source

      Not checked in the free run.

    • cites [43] 43. Erin Treacy Solovey, Audrey Girouard, Krysta Chauncey, Leanne M Hirshfield, Angelo Sassaroli, Feng Zheng, Sergio Fantini, and Robert J K Jacob. 2009. Using fNIRS Brain Sensing in Realistic HCI Settings : Experiments and Guidelines. Proc. UIST 2009 (2009). source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    Physiological measures provide continuous data, which is especially important for adaptive tasks that react to learners' needs in real-time.

    “This misses the advantage that physiological measures provide which is the continuous availability of data which is especially important for adaptive tasks that can react to learners' needs in real-time.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning

    The segmentation behavior in the normal condition can be viewed as akin to self-judgment of when to increase task complexity while learning.

    “This can be viewed as akin to self-judgment of when to increase task complexity while learning.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Part-task training is only helpful with sequential material, not concurrent tasks.

    “Meta-analyses have found parttask training is only helpful with sequential material, such as playing sequential notes on a piano (e.g.[3]), rather than concurrent tasks, such as playing two hands on a piano at the same time [52,53].”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [3] 3. Daniel W. Ash and Dennis H. Holding. 1990. Backward versus forward chaining in the acquisition of a keyboard skill. Human Factors 32 (1990), 139-146. source

      Not checked in the free run.

    • cites [52] 52. Dennis C. Wightman and Gavan Lintern. 1985. Part-Task Training for Tracking and manual control. Human Factors 27, 3 (1985), 267-283. source

      Not checked in the free run.

    • cites [53] 53. Bin Xie and Gavriel Salvendy. 2000. Prediction of mental workload in single and multiple task environments. International journal of cognitive ergonomics 4 (2000), 213-242. source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    Adaptive learning tasks are important, especially when using increasing-difficulty strategies.

    “These findings highlight the importance of adaptive learning tasks, especially when using increasing-difficulty strategies.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    For increasing-difficulty training to be adaptive, there must be some measurement of when the user is ready to move to the next level.

    “In order for increasing-difficulty training to be adaptive, there has to be some measurement of when the user is ready to move on to the next level of difficulty.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Interpretationnone given

    Removing some notes meant that some minor counterpoint rules are no longer met, but the underlying structure and form of both pieces remain intact.

    “We note that by removing some of the notes, some minor counterpoint rules are no longer met.However, the underlying structure and form of both pieces is still intact.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    The self-reporting questionnaire has issues with sensitivity, reliability, and validity in measuring cognitive workload.

    “De Jong [10] and Moreno [31] raise several issues with the scale's sensitivity, reliability and validity in measuring cognitive workload.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [10] 10. Ton de Jong. 2010. Cognitive load theory, educational research, and instructional design: Some food for thought. Instructional Science 38 (2010), 105-134. DOI: http://dx.doi.org/10.1007/s11251-009-9110-0

      Not checked in the free run.

    • cites [31] 31. Roxana Moreno. 2010. Cognitive load theory: more food for thought. Instructional Science 38, November 2009 (2010), 135-141. DOI: http://dx.doi.org/10.1007/s11251-009-9122-9 source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    Brain sensing provides a more direct measurement of physiological changes during higher cognitive workload and can be measured continuously for use as a real-time determinant in adaptive learning.

    “Brain sensing provides a more direct measurement of the physiological changes occurring in the brain during higher cognitive workload and can be measured continuously which can be used as a real-time determinant in adaptive learning studies.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    Meta-analyses show that students using intelligent tutoring systems and computer-based education tend to learn faster, more accurately, and with increased enjoyment.

    “There have been promising results; meta-analyses revealed that students tend to learn faster and more accurately, and have increased enjoyment of their learning material [27].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [27] 27. James A. Kulik. 1994. Technology Assessment in Education and Training. Eva Baker and Harold F. O'Neil Jr, Chapter Meta-analytic Studies of findings on computer-based instruction., 9-34.

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation

    Task-evoked pupillary responses have been associated with changes in cognitive workload.

    “Task-evoked pupillary responses (TEPRs) have also been associated with changes in cognitive workload [49,20].”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [49] 49. Pascal W M Van Gerven, Fred Paas, Jeroen J G Van Merriënboer, and Henk G. Schmidt. 2004. Memory load and the cognitive pupillary response in aging. Psychophysiology 41 (2004), 167-174. DOI:http: //dx.doi.org/10.1111/j.1469-8986.2003.00148.x

      Not checked in the free run.

    • cites [20] 20. Shamsi T Iqbal, Xianjun Sam Zheng, and Brian P Bailey. 2004. Task-evoked pupillary response to mental workload in human-computer interaction. Extended abstracts of the 2004 conference on Human factors and computing systems CHI 04 (2004), 1477. DOI: http://dx.doi.org/10.1145/985921.986094 source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    The fNIRS signal is resilient to respiration, heartbeat, eye movement, minor head motion, and mouse and keyboard clicks.

    “The fNIRS signal has been found to be resilient to respiration, heartbeat, eye movement, minor head motion, and mouse and keyboard clicks [43].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [43] 43. Erin Treacy Solovey, Audrey Girouard, Krysta Chauncey, Leanne M Hirshfield, Angelo Sassaroli, Feng Zheng, Sergio Fantini, and Robert J K Jacob. 2009. Using fNIRS Brain Sensing in Realistic HCI Settings : Experiments and Guidelines. Proc. UIST 2009 (2009). source

      Not checked in the free run.

  • May combine two claims.

    Study design and conductown data

    Early studies showed that a fixed percentage threshold for increasing difficulty did not work, because learners' brain measurements were highly individual.

    “Early studies showed us very quickly that this was not going to work.Each learner was very individual in their brain measurements, and what worked very well for some, would not even trigger a fixed percentage threshold for others.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductreasoning

    BACh was designed not to interfere when learners' cognitive workload was high, and its adaptations never decreased difficulty during periods of high cognitive workload.

    “We purposefully designed BACh not to interfere when learners' cognitive workload was high in order to avoid disrupting a period of possible learning.”
    • Also stated in: § Modeling and Adapting to Cognitive Workload · plain
      “This is the reason why the adaptations never decreased the difficulty level during periods of high cognitive workload, as this could have proven both disruptive and frustrating to the learners.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    In the normal condition, learners were told to learn in any way they wished, and all of them segmented the pieces as they learned.

    “Learners were told that they could learn the normal condition in any way they wished as they normally would.As a consequence, all learners segmented the pieces while they learnt during the normal condition.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    BACh was able to correctly respond to learner cognitive workload so as to improve playing speed and accuracy.

    “By examining the performance data and participant feedback, it seems that BACh was able to correctly respond to learner cognitive workload to improve playing speed and accuracy.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    BACh helps piano players play with significantly higher accuracy and speed than the normal condition.

    “Taken together, these results suggest that BACh helps piano players to play with significantly higher accuracy and speed than the normal condition.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    By using brain sensing, BACh acquires an objective view of the learner's cognitive state and adjusts the learning task to guide the learner into the zone of proximal development.

    “By using brain sensing, BACh is able to acquire an objective view into the learner's cognitive state and adjust the learning task to best guide the learner into the zone of proximal development.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

    Interpretationnone given

    The underlying premise of BACh can be applied to any learning situation involving a complex task that can be broken into steps of increasing difficulty.

    “the underlying premise of BACh can be applied to any learning situation for a complex task that can be broken down into steps of increasing difficulty.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Harrison et al. found that oxygenation levels measured by fNIRS increased along with self-reports of increased cognitive workload as the number of aircraft under operator control increased.

    “In an air traffic control simulation, they found oxygenation levels measured by fNIRS increased along with self-reports of increased cognitive workload as the number of aircrafts under operator control increased [19].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [19] 19. Joshua Harrison, Kurtulus Izzetoglu, Hasan Ayaz, Ben Willems, Sehchang Hah, Ulf Ahlstrom, Hyun Woo, Patricia Shewokis, Scott C. Bunce, and Banu Onaral. 2014. Cognitive workload and learning assessment during the implementation of a next-generation air traffic control technology using functional near-infrared spectroscopy. IEEE Transactions on Human-Machine Systems 44, 4 (2014), 429-440. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    The order of the two conditions alternated between participants.

    “The order of the conditions alternated between participants.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Fixed scheduling of increases in task difficulty is not helpful.

    “While meta-analyses have shown that fixed scheduling when increasing task difficulty is not helpful [52,53]”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [52] 52. Dennis C. Wightman and Gavan Lintern. 1985. Part-Task Training for Tracking and manual control. Human Factors 27, 3 (1985), 267-283. source

      Not checked in the free run.

    • cites [53] 53. Bin Xie and Gavriel Salvendy. 2000. Prediction of mental workload in single and multiple task environments. International journal of cognitive ergonomics 4 (2000), 213-242. source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data

    BACh is targeted at and useful for beginner piano players.

    “We, therefore, simply suggest that our system is targeted at and useful for beginner piano players.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductown data

    An expert musician who had taught piano rated participants' learning compared to groundtruth piano rolls.

    “An expert musician who had taught piano in the past rated participants' learning compared to the groundtruth piano rolls.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Premisecitation

    Instructional techniques that benefit beginners can have reverse effects on more experienced learners (the expertise reversal effect).

    “Learning and CLT literature has explored a phenomenon called the expertise reversal effect whereby instructional techniques that are beneficial to beginners can have the reverse effects on more experienced learners [22,23].”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [22] 22. Slava Kalyuga. 2003. The expertise reversal effect. Educational psychologist 38, 1 (2003), 23-31.

      Not checked in the free run.

    • cites [23] 23. Slava Kalyuga. 2007. Expertise reversal effect and its implications for learner-tailored instruction. Educational Psychology Review 19, 4 (2007), 509-539. source

      Not checked in the free run.

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductown data

    Shapiro-Wilk tests verified that the data was non-parametric, followed by Wilcoxon Signed-rank tests on all dependent variables.

    “We carried out Shapiro-Wilk tests that verifed that the data was non-parametric, followed by Wilcoxon Signed-rank tests on all dependent variables (Table 1).”
    Quote in the paperSubstantial claim
37 of 37

  • The paper relies on this without support.

    When a learner's cognitive workload falls below a certain threshold, they have mastered the material and can handle more cognitive information.

    “As users' cognitive workload fell below a certain threshold, suggesting that they had mastered the material and could handle more cognitive information, BACh automatically increased the difficulty of the learning task.”
    “BACh increased the difficulty level by adding a full line of music when users' cognitive workload fell below a threshold, indicating that they could handle more information.”

  • The paper relies on this without support.

    The normal condition, in which participants learned as they normally would, is a valid control for how they would learn without BACh.

    “One chorale was presented in normal form and participants were instructed to learn the piece the way they normally would.”
    “In the normal condition, learners were told they could learn the piece the way they normally would.”

  • The paper relies on this without support.

    The expert musician's ratings accurately reflect participants' learning.

    “An expert musician who had taught piano in the past rated participants' learning compared to the groundtruth piano rolls.”

  • The paper relies on this without support.

    The measurement periods determined by pilot studies are appropriate for the main study.

    “For each level of difficulty, we measured each learner's cognitive workload for a period of time determined by pilot studies.”
    “We ran many pilot studies to create an algorithm that worked for beginners.”

  • The paper relies on this without support.

    Piano players are typically classically taught to play pieces first with their right hand, then their left hand, and finally both together.

    “Piano players are typically classically taught to play pieces first with their right hand, then their left hand, and finally both together.”

  • The paper relies on this without support.

    A learner's cognitive workload, as measured by fNIRS brain sensing, reliably indicates whether that learner is able to handle more information and a higher level of difficulty.

    “The goal of BACh was to change the level of difficulty at a time when the participants' cognitive workload was low enough to handle more information and a higher level of difficulty.”

  • The paper relies on this without support.

    The performance differences observed between the BACh and normal conditions are attributable to BACh's cognitive-workload-based adaptation rather than to other differences between the two conditions.

    “We showed that BACh significantly increased speed and accuracy compared to a control condition where participants learned pieces the way they normally would.”
7 of 7

Statistics that don’t add up

Each reported p-value recomputed from its test statistic and degrees of freedom.

  • p-value fits an unstated one-tailed test: t(29) = -2.42,p = .01

    Minor

    The paper reports p = .01 here; recomputing from the test statistic and its degrees of freedom gives p = .022 (two-tailed). The reported value is what a one-tailed test gives, but the paper does not say near this result that a one-tailed test was used.

    Reported p = .01 · recomputed p = .022

1 of 1

Statistics recomputed

Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.

  • Doesn’t add upt(29) = -2.42reported p = .01 · recomputed p = .022

    “t(29) = -2.42,p = .01”

  • Not confirmedz = 2.5337reported p = .009 · recomputed p = .011

    “(Z = 2.5337, p = .009”

  • Not confirmedz = -2.719reported p = .005 · recomputed p = .007

    “(Z = -2.719,p = .005”

  • Consistentz = -1.9689reported p = .05 · recomputed p = .049

    “(Z = -1.9689,p = .05”

  • Consistentz = 2.3151reported p = .019 · recomputed p = .021

    “(Z = 2.3151, p = .019”

    The paper states an exact rank test

  • Consistentz = 2.4401reported p = .015 · recomputed p = .015

    “(Z = 2.4401, p = .015”

  • Consistentz = 3.0351reported p = .004 · recomputed p = .002

    “(Z = 3.0351, p = .004”

    The paper states an exact rank test

  • Consistentz = 2.482reported p = .01 · recomputed p = .013

    “(Z = 2.482, p = .01”

8 of 8

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Self-check results appear here once an independent check has confirmed them.

Cites

Flagged first, then by citations

Research works printed but not linked: 9

  1. Ellen R. Bialo and Jay Sivin-Kachala. 1996. The effectiveness of technology in schools: A summary of recent research. (1996).
  2. 10.5555/2787789 · DOI
  3. 10.1007/s11251-009--9110-0 · DOI
  4. Hugh Garavan, Dan Kelley, Allyson Rosen, Steve M. Rao, and Elliot A. Stein. 2000. Practice-related functional activation changes in a working memory task. Microscopy research and technique 51, 1 (2000), 54-63.
  5. 10.1007/s10648-010--9150--7 · DOI
  6. James A. Kulik. 1994. Technology Assessment in Education and Training. Eva Baker and Harold F. O'Neil Jr, Chapter Meta-analytic Studies of findings on computer-based instruction., 9-34.
  7. 10.5555/265013 · DOI
  8. 10.1007/s10648-007--9053--4 · DOI
  9. 10.1111/j.1469--8986.2003.00148.x · DOI

Cited by

Sorted by citations

Related

Authors
Yuksel, B. F., Oleson, K. B., Harrison, L., Peck, E. M., Afergan, D., Chang, R., & Jacob, R. J.
Affiliations
  • Tufts University (US)
  • Worcester Polytechnic Institute (US)
  • Bucknell University (US)
ORCID
Venue
Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems
Publisher
ACM
Published
2016-05-07
Pages
5372-5384
DOI
10.1145/2858036.2858388
OpenAlex
W2402939184
Cited by
137
References
56
Citations by year
2026: 14 · 2025: 10 · 2024: 10 · 2023: 12 · 2022: 10 · 2021: 21 · 2020: 16 · 2019: 21 · 2018: 14 · 2017: 6 · 2016: 3
FWCI
7.90
Type
conference paper
Language
en
Open access
Licence
www.acm.org/publications/policies/copyright_policy#Background (vor, 365 d)
Locations
CHI Conference on Human Factors in Computing Systems (publishedVersion)
Funders
  • National Science Foundation (IIS-1218170 IIS-1065154)
  • Google Inc.
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Intelligent Tutoring Systems and Adaptive Learning
  • Life Sciences › Neuroscience › Cognitive Neuroscience › EEG and Brain-Computer Interfaces
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Neuroscience, Education and Cognitive Function
Keywords
piano learning, adaptive learning, brain-computer interface, mental workload, functional near-infrared spectroscopy, music learning, user feedback

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-09-29.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-09-29.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review