Article Contents
Article Contents

# Modeling student engagement using optimal control theory

• * Corresponding author: Debra Lewis

Dedicated to Professor Anthony Bloch on the occasion of his 65th birthday.

• Student engagement in learning a prescribed body of knowledge can be modeled using optimal control theory, with a scalar state variable representing mastery, or self-perceived mastery, of the material and control representing the instantaneous cognitive effort devoted to the learning task. The relevant costs include emotional and external penalties for incomplete mastery, reduced availability of cognitive resources for other activities, and psychological stresses related to engagement with the learning task. Application of Pontryagin's maximum principle to some simple models of engagement yields solutions of the synthesis problem mimicking familiar behaviors including avoidance, procrastination, and increasing commitment in response to increasing mastery.

Mathematics Subject Classification: Primary: 49N90, 91E40, 93B50.

 Citation:

• Figure 1.  Left: graphs of $u_c(\mu, \cdot )$ for sample values of $\mu$; the lowest curve is the graph of $u_c(1, \cdot )$. Right: contour plot of $u_{\rm opt}$, with contour values $\frac j {10}, j = 0, \ldots, 10$. Black: u = 0; white: u = 1

Figure 2.  Left: $K_{\frac \pi {10}}(r)$ for $r$ near $\tan \frac \pi {10}$, right: $\upsilon_\circ(r)$ associated to $\xi_\circ(u) = \sqrt{1 - u^2}$

Figure 3.  Graphs of $X_{h_0}^\pm$ for $\mathcal{S} = [0, 1]$, $\gamma(m) = (1 - m)^2$, $\mu(m) = {1 \over 2} \left ( 1 - \frac m 2 \right )^2$, constant functions $\psi_{\rm pe} = {1 \over 2}$ and $\phi = \frac \pi {12}$, and representative values of $h_0$ for which $\breve C_{h_0}$ has a strict global minimum ${c_{h_0}}$. Solid curves: $X_{h_0}^+$; dashed curves: $X_{h_0}^-$

Figure 4.  Representative vertically aligned graphs of $K_\phi(r)$ and $\breve C_{h_0}(m)$ for constant $\phi$ and three values of $h_0$. Solid: projections of the trajectories with initial data $(m_0, r_0)$; dotted: points on the graphs outside those projections. Upper row: $r_0 < r_\phi$ and $m_0 > m_c$; initially, $m$ decreases and $r$ increases. Lower row: $r_0 > r_\phi$ and $m_0 < m_c$; initially, $m$ increases and $r$ decreases. Left: $m'$ changes sign as $r$ passes through $r_\phi$; center: $m$ and $r$ asymptotically approach the equilibrium $(m_c, r_\phi)$; right: $r'$ changes sign as $m$ passes through $m_c$

•  [1] M. C. Anderson, R. A. Bjork and E. L. Bjork, Remembering can cause forgetting: Retrieval dynamics in long-term memory, Journal of Experimental Psychology. Learning, Memory, and Cognition, 20 (1994), 1063-1087.  doi: 10.1037/0278-7393.20.5.1063. [2] L. Averell and A. Heathcote, The form of the forgetting curve and the fate of memories, Journal of Mathematical Psychology, 55 (2011), 25-35.  doi: 10.1016/j.jmp.2010.08.009. [3] J. Baillieul and J. Willems, Mathematical Control Theory, Springer, 1999. doi: 10.1007/978-1-4612-1416-8. [4] A. Bandura,  Self-Efficacy in Changing Societies, Cambridge University Press, 1995. [5] A. M. Bloch, Nonholonomic Mechanics and Control, Springer, 2003. doi: 10.1007/b97376. [6] H. Chang and S. L. Beilock, The math anxiety-math performance link and its relation to individual and environmental factors: a review of current behavioral and psychophysiological research, Current Opinion in Behavioral Sciences, 10 (2016), 33-38.  doi: 10.1016/j.cobeha.2016.04.011. [7] K. Choe, J. Jenifer, C. Rozek, M. Berman and S. Beilock, Calculated avoidance: Math anxiety predicts math avoidance in effort-based decision-making, Science Advances, 5 (2019).  doi: 10.1126/sciadv.aay1062. [8] R. Dorfman, An economic interpretation of optimal control theory, The American Economic Review, 59 (1969), 817-831. [9] H. Ebbinghaus, Über das Gedächtnis, Dunker, 1885. [10] E. O. Finkenbinder, The curve of forgetting, The American Journal of Psychology, 24 (1913), 8-32.  doi: 10.2307/1413271. [11] T. L. Griffiths, F. Lieder and N. D. Goodman, Rational use of cognitive resources: Levels of analysis between the computational and the algorithmic, Topics in Cognitive Science, 11 (2015), 217-229.  doi: 10.1111/tops.12142. [12] K. Hall, E. Fawcett, K. Hourihan and J. Fawcett, Emotional memories are (usually) harder to forget: A meta-analysis of the item-method directed forgetting literature, Psychonomic Bulletin & Review, 28 (2021), 1313-1326.  doi: 10.3758/s13423-021-01914-z. [13] D. Kirk, Optimal Control Theory: An Introduction, Dover Publications, Inc., 2004. [14] S. Koessler, H. Engler, C. Riether and J. Kissler, No retrieval-induced forgetting under stress, Psychological Science, 20 (2009), 1356-1363.  doi: 10.1111/j.1467-9280.2009.02450.x. [15] D. Lewis, A soothing invisible hand: Moderation potentials in optimal control, in Geometry, Mechanics, and Dynamics: The Legacy of Jerry Marsden, Springer, 2015,257–284. doi: 10.1007/978-1-4939-2441-7_12. [16] G. R. Loftus, Evaluating forgetting curves, Journal of Experimental Psychology, 11 (1985), 397-406.  doi: 10.1037/0278-7393.11.2.397. [17] I. Lyons and S. Beilock, When math hurts: Math anxiety predicts pain network activation in anticipation of doing math, PLOS ONE, 7. doi: 10.1371/journal.pone.0048076. [18] E. Maloney, M. Schaeffer and S. Beilock, Mathematics anxiety and stereotype threat: Shared mechanisms, negative consequences, and promising interventions, Research in Mathematics Education, 15 (2013), 115-128.  doi: 10.1080/14794802.2013.797744. [19] A. Mattarella-Micke, J. Mateo, M. Kozak, K. Foster and S. Beilock, Choke or thrive? The relation between salivary cortisol and math performance depends on individual differences in working memory and math anxiety, Emotion, 11 (2011), 1000-1005.  doi: 10.1037/a0023224. [20] H. Melville, Bartleby, the scrivener: A story of Wall Street, Putnam's Magazine, November and December (1853), 446–457 and 609–615. doi: 10.1093/oseo/instance.00209193. [21] R. Montgomery, Optimal control of deformable bodies and its relation to gauge theory, Math. Sci. Res. Inst. Publ., 22 (1991), 403-438.  doi: 10.1007/978-1-4613-9725-0_15. [22] J. Murre and J. Dros, Replication and analysis of Ebbinghaus' forgetting curve, PLOS ONE, 10 (2015).  doi: 10.1371/journal.pone.0120644. [23] L. Pontryagin, V. Boltyanskii, R. Gamkrelidze and E. Mishchenko, The Mathematical Theory of Optimal Processes, Interscience Publishers, 1962. [24] R. Rydell, A. McConnell and S. Beilock, Multiple social identities and stereotype threat: Imbalance, accessibility, and working memory, Journal of Personality and Social Psychology, 96 (2009), 949-966.  doi: 10.1037/a0014846. [25] A. Shenhav, S. Musslick, F. Lieder, W. Kool, T. Griffiths, J. Cohen and M. Botvinick, Toward a rational and mechanistic account of mental effort, Annual Review of Neuroscience, 40 (2017), 99-124.  doi: 10.1146/annurev-neuro-072116-031526. [26] F. Sirois and T. Pychyl, Procrastination and the priority of short-term mood regulation: Consequences for future self, Social and Personality Psychology Compass, 7 (2013), 115-127.  doi: 10.1111/spc3.12011. [27] N. Slamecka, On comparing rates of forgetting: Comment on Loftus (1985), Journal of Experimental Psychology. Learning, Memory, and Cognition, 11 (1985), 812-816.  doi: 10.1037/0278-7393.11.1-4.812. [28] E. Sontag, Integrability of certain distributions associated with actions on manifolds and applications to control problems, in Nonlinear Controlability and Optimal Control, Marcel Dekker, Inc., 133 (1990), 81–131.

Figures(4)