Post (4): Optimizing K-12 Educational Production with EdTech

Recap from previous posts:

EdTech can improve educational production in K-12 schools with limited resources, but the current state of EdTech design and deployment does not harness its full potential. EdTech should integrate all entities and processes in K-12 education to promote EdTech use among students and to develop their cognitive and noncognitive skills. Accordingly, I proposed an integrated framework of educational production in my last post.

Links to previous posts

https://www.dhirubhai.net/pulse/how-can-edtech-improve-educational-production-k-12-schools-anuj-kumar

https://www.dhirubhai.net/pulse/post-2-edtech-deployment-k-12-schools-aligned-maximize-anuj-kumar ?

https://www.dhirubhai.net/pulse/post-3-integrated-framework-educational-production-anuj-kumar

?Optimizing Educational Production –

In this article, I discuss how EdTech applications can utilize the integrated framework (Figure 1) to improve educational production.

To improve educational production, one needs to improve the output of underlying education-generating processes. In the following, I propose a generic framework for enhancing the output of an educational process and classify past interventions into this framework.

I will argue that these interventions are often complementary, and combining them on an EdTech platform can substantially increase their combined output. Therefore, EdTech applications should integrate these interventions to optimize educational production.

Framework for optimizing educational process output –

I categorize various entities participating in an educational process in Figure 1 into focal and supporting entities. The focal entities directly participate in the educational process, while supporting entities indirectly do so. For example, students (focal entity) participate directly in the content consumption process, but peers/family/teachers (supporting entities) support them students during the process.

The output of an educational process can be increased by improving its input and efficiency:

1.???? Improving process inputs – Higher inputs into the process should lead to higher output, all else equal. Accordingly, improving the quantity and quality of focal entities’ inputs in the educational process can increase its output. For example, students can learn more if they exert more effort in consuming educational content. The inputs from focal entities in an educational process can be enhanced by (1) direct interventions to focal entities and (2) interventions to the supporting entities to help increase the focal entities’ inputs.?

2.???? Improving process efficiency – An efficient process can generate higher output for a given input from a focal entity. I classify various ways to increase process efficiency into three categories.?

·?????? Reducing the cost of focal entity effort

·?????? Improving process effectiveness using ICT.

·?????? Improving process output with information sharing and feedback from connected processes.

Following the above classification, I will discuss how to improve the output of the content consumption process in the K-12 educational system. Students are the focal entity in this process. The inputs to the process are students’ learning effort (i.e., consuming educational content), and the output is their learning.

I will organize past research interventions to improve student learning in this classification scheme and then suggest opportunities for new EdTech-enabled interventions.

1.0??????????? Improving process inputs ?–There are two reasons why students underinvest in education - lack of student input in content consumption. First, many students may not put adequate effort into learning from educational content because they are unmotivated – effort gaps (or lower input quantity). The other reason is that even if students put in the effort, they cannot learn because they lack the requisite ability or skill to self-learn – ability/skill gaps (or lower input quality).

I classify interventions to address students’ effort and ability gaps into two categories – (1) student-level interventions and (2) interventions to other supporting entities.? ?

1.1?????? Student-level interventions–

1.1.1??? Addressing effort gaps – Past research has identified various reasons why students underinvest in education and suggest interventions to address them. Below, I discuss these underlying reasons and how EdTech-enabled interventions can address them.

A.??????? Information Friction – Students (especially underprivileged students) may underinvest in education because they are not fully aware of the value of education. Informing them about the value of education could improve their educational attainment – such as informing high school students in the Dominican Republic about the differential wages with and without reduced high school dropout rates (Jensen 2010).

Sometimes, students realize the value of education but have cognitive limitations in understanding how to attain quality education. Interventions that explain to students how to select, apply, and get admitted to good schools could relieve their cognitive strain. For example, students from low-income families applied to more selective colleges when given an information package explaining these colleges’ admissions and financial assistance (Hoxby and Turner 2013).

Students from low-income households also don’t invest in education because they lack aspirations (they don’t believe they can succeed given their background). Showing role model videos can effectively raise their aspirations and, thus, educational performance (Riley 2022).

EdTech applications are especially suited for providing personalized information to targeted students at the appropriate time at scale to maximize its effect.???

B.??????? Limited attention – Students, especially young ones, have limited attention problems. They may forget to complete the learning activities or may get easily distracted while doing them (DellaVigna 2009). Past research has shown that EdTech-enabled personalized reminders to students to attend learning activities can address their limited attention.

C.??????? Behavioral biases – Even though students know that education is valuable for their future well-being, they underinvest in education because of behavioral biases.

C.1????? Self-control/Present bias – Students may find the present cost of learning more salient to education’s vague and uncertain future benefits – present bias/hyperbolic discounting of future benefits/ self-control problem. Due to this reason, students choose current consumption (e.g., entertainment and leisure) and procrastinate the costly investment of studying (Lavecchia and Oreopoulos 2016). Past interventions have attempted to mitigate this problem by making future benefits of education more salient (goal-setting interventions in Morisano et al. 2010), offering immediate extrinsic (monetary and nonmonetary) rewards to offset immediate costs, and raising students’ intrinsic motivation to study (such as raising their self-confidence and self-esteem) to improve their learning effort.

Informational nudges are an effective way to mitigate students’ psychological biases. Nudges encourage desired outcomes by affecting seemingly trivial changes in the cost of choices or how choices are presented without forbidding or significantly changing their costs (Thaler and Sunstein 2009). For example, showing students the optimal option as the default or at a higher rank in the menu of choices could result in students making optimal choices (Marx and Turner 2017). Loss framing of options can also motivate students to undertake desirable behavior – e.g., students receiving loss framing of incentives for learning efforts in US schools made higher effort (Levitt et al. 2016).

Setting up commitment devices for students has also shown promise in mitigating their self-control problems. For example, setting up interim deadlines for assignments (exams) may be a commitment device for students to do things sooner rather than later (O’Donoghue and Rabin 1999). Similarly, self-set goals by students also act as a commitment device. Students are motivated to avoid the psychological costs (due to loss aversion) of not achieving the goals.? ?

C.2????? Inertia – Students often have inertia in changing their routines. They are predisposed to making choices relying on familiar knowledge and discounting the new information. This problem is especially salient when students’ educational environment changes (such as going to higher grades). Due to inertia, often underprivileged students don’t acquire information on opportunities and make decisions based on readily available information from their current environment. Nudges, reminders, and guidance to students have shown promise in addressing student inertia. ???

C.3????? Social identity – Students’ educational choices are determined by how they perceive themselves (self-image) and how others perceive them (social image). Students may make suboptimal educational choices under pressure to conform to social norms in their social comparison group. Students may put in effort in school based on not only the individual benefits (such as grades) but also the social benefits (such as whether their effort and outcomes are consistent with their social group). If the norm in the social group is to procrastinate and indulge in leisure, students adhere to these norms. Manipulating students’ peer groups, such as paring vulnerable students (those with less self-control) with better peers, could prevent them from adhering to negative social norms. After manipulating students’ peer groups, they may be motivated to study by showing the average performance of their social group – social comparison nudges – (Azmat and Iriberri 2010).

D.??????? EdTech-enabled games/activities – Students may naturally exert more effort if they find learning fun and engaging or have an incentive (monetary or otherwise) to perform well. EdTech-enabled gamification in education could achieve these objectives. For example, students may exert more effort in learning to win digital class tournaments.

E.??????? Psychosocial interventions – Students’ behavioral biases stem from their lack of noncognitive skills. For example, a lack of conscientiousness leads to self-control issues, or a lack of attention control leads to forgetfulness. Previous interventions compensate for students’ effort gaps due to their lack of noncognitive skills but do not develop them. For example, informational nudges (reminders) mitigate the effect of self-control (limited attention) issues, but they do not make students more conscientious (attentive). Recent research has examined psychosocial interventions to build students’ noncognitive skills. For instance, growth mindset interventions teach students that they can develop skills over time with effort (noncognitive skill of grit and work ethic) (Paunesku et al. 2015), showing videos highlighting the role of effort makes students grittier (Alan et al. 2019), and goal-setting interventions make students ambitious and achievement striving (Morisano et al. 2010). EdTech, with multimedia features, interactivity, and technology-enabled personalization, is especially suited for psychosocial interventions for students in early grades.

1.1.2 ?? Addressing ability gaps – Often, students have a below-grade level understanding of content, especially true for underprivileged and developing countries students. These students may not learn from the content even if they put in the effort. EdTech-based interventions could identify students’ knowledge gaps by analyzing their homework performance and offering them remedial content to fill their knowledge gaps (Muralidharan et al. 2019, Kumar and Mehra 2018).

1.2?????? Other entity-level interventions – Other entities in the educational production system can motivate and support students to learn. I classify past interventions into interventions on the existing entities in K-12 education and the opportunity to integrate new entities to support students on EdTech platforms.?

1.2.1??? Existing entities – Existing entities (such as parents, peers, and teachers) can provide guidance, pressure, and support to promote students’ investment in education.

A.??????? Parents – A large body of research has examined how to promote parental engagement in students learning. These studies examine why parents underinvest in students’ education (such as limited attention, cognitive strain, and delayed gratification) and how to address these underlying reasons with appropriate interventions (York et al. 2019, Mayer et al. 2015 & 2019). EdTech-enabled notifications offering parents personalized information (such as students’ knowledge gaps) along with appropriate resources (such as educational content in students’ knowledge gaps or who can help students) can be highly effective in improving students’ learning (Mayer et al. 2019).?

B.??????? Teachers – Teachers can also motivate and support students to learn. If students complete homework on digital platforms, EdTech applications can identify and convey individual students’ and aggregate class knowledge gaps to teachers in real-time. Informational nudges, reminders, and extrinsic/ intrinsic motivation can be deployed to motivate teachers to address students’ effort and ability gaps individually or collectively.

C.??????? Peers – Peers are another valuable resource in the K-12 educational systems. Although ample evidence exists for peer effects on students’ learning (Epple and Romano 2011), it is underutilized in the EdTech applications. Given the right incentives, peers can motivate students to engage with educational content and provide the required support to help students learn from it. Harnessing the power of peers is especially relevant in developing countries where other resources, such as teachers and family support, are limited. In such scenarios, the EdTech platform can identify students’ knowledge gaps and connect them to their peers who understand those topics well. With an appropriate incentive design (such as the social recognition of helpful peers), peers may be motivated to help students with learning needs. ??????

1.2.2??? Integrating new entities – ICT removes the location and time constraint for entities to participate in educational processes. Thus, EdTech platforms could allow untapped entities in the K-12 educational ecosystem to motivate and help students learn educational content. For example, EdTech platforms can integrate professionals/retired personnel as mentors in the K-12 educational production system. These mentors can interact with students periodically to inform them about the value of education, raise their aspirations, motivate them to study, and provide the required support in cognitive and noncognitive skill development. Thus, the EdTech platform could allow the integration of unused/underutilized resources in the educational ecosystem, an option especially relevant in resource-constrained schools and less educated parents in developing countries.

2.0?????? Improving process efficiency? – Improving process efficiency means generating higher output with less input. EdTech can do it in the following ways.

2.1?????? Reducing effort costs: EdTech can reduce students’ costs to consume content in many ways. (1) Search and recommendation algorithms on the EdTech platform can reduce student costs (effort) in finding relevant educational content. (2) The multimedia and interactivity of ICT can make learning more enjoyable and thus reduce students’ learning costs. (3) ICT removes the colocation and synchronicity constraint for learning. So, it reduces students’ learning costs from the educational content by allowing them to consume it at the pace, place, and time of their choice (especially for students with location and time constraints). ?????????? ??

2.2?????? Enhancing process output: EdTech offers functionalities to generate higher process output for a given input. ICT can improve the efficiency of students’ learning on EdTech platforms

(1) Adaptivity - EdTech applications serving content per students’ learning needs allow students to efficiently expend effort in their knowledge gaps and, thus, have a higher learning from a given effort. Recent advances in AI and machine learning techniques allow more granular modeling of students’ cognitive processes and, accordingly, serve them personalized content. Examples of such CAL software tested in the research studies are ASSISTments and the Cognitive Tutor Program of Carnegie Learning in the US and Mindspark software in India. ?

(2) Multimedia and interactivity – EdTech applications can significantly improve learning by offering multimedia and interactive educational content. Students may better internalize theoretical concepts with their visual display and the ability to interact with such display. An example of such software tested in research studies is SimCalc in the US (Roschelle et al. 2010).??? ?

2.3?????? Process integration and feedback: EdTech platforms integrating various education-generating processes can allow information sharing to improve the overall efficiency of educational production. For example, EdTech applications utilizing advanced machine learning and AI techniques can identify students’ knowledge gaps during their content consumption process and share them with teachers. Teachers can accordingly adjust class instructions to revisit existing content and add more content on topics where most students perform poorly. Such integration of the content consumption process by students and teachers’ content creation and delivery processes can further improve students’ overall learning.

In the above exposition, I proposed a generic classification scheme for systematically assimilating past research interventions to improve student learning in the content consumption process. Similarly, past interventions to improve the content creation and delivery processes can be organized in this classification scheme.

Combining interventions on an EdTech platform –

While the above-discussed interventions have individually shown promise in addressing a part of the educational production system, they do not harness the full potential of EdTech. I proposed the above classification scheme as a roadmap for combining these interventions into an integrated application. These interventions, when combined, can complement each other and have a multiplier effect on educational production.

I illustrate this complementarity with an example of blended learning on an EdTech platform. Suppose teachers provide face-to-face class instructions on different topics, and students can also learn from watching videos on those topics and completing homework on the EdTech platform. I can think of implementing the following applications on such an EdTech platform.?

-??????? Identify students’ content consumption behavior (pattern of watching lecture videos/ doing homework) and offer them informational nudges, reminders, and social comparison nudges to improve their learning efforts.

-??????? Identify students’ knowledge gaps and offer personalized educational content to improve their learning efficiency.

-??????? Alert parents about students’ knowledge gaps and provide them with relevant content (lecture videos) to motivate and help students learn those concepts.

-??????? Manipulate students’ peer groups to maximize knowledge diffusion among them – such as pair up a student weak on a topic with peers who know that topic well and create incentives for knowledge sharing among them.

-??????? Share information on students’ knowledge gaps/performance with teachers so that they can adjust the content of their lecture and its delivery to optimize educational production.

-??????? Design and organize periodic educational games/activities to improve noncognitive skill formation in students, such as improve grit /perseverance, self-confidence/self-esteem.

Offering these interventions together can have a multiplier effect on their outcomes. For example,

A.??? Offering (1) personalized content to students in their knowledge gaps and (2) information and relevant content on students’ knowledge gaps to parents can improve the efficacy of both interventions. On the one hand, knowing what students need and how to help them improves parents’ ability to motivate and help students learn those concepts. On the other hand, with parental motivation and support, students may exert effort and, thus, learn from the personalized content.

B.??? Offering (1) interventions to build students’ noncognitive skills (such as self-esteem/self-confidence) and (2) adaptive learning to students can complement each other. The students can benefit more from consuming personalized content if they have higher self-esteem. ?

In my next post, I will discuss how I have combined different interventions to develop an integrated EdTech platform (EPInc App) to maximize peer-driven knowledge diffusion in classrooms. ?

?

References –

Alan, S., Boneva, T., & Ertac, S. (2019). Ever failed, try again, succeed better: Results from a randomized educational intervention on grit.?The Quarterly Journal of Economics,?134(3), 1121-1162.

Azmat, G., & Iriberri, N. (2010). The importance of relative performance feedback information: Evidence from a natural experiment using high school students.?Journal of Public Economics,?94(7-8), 435-452.

DellaVigna, S. (2009). Psychology and economics: Evidence from the field.?Journal of Economic literature,?47(2), 315-372.

Epple, D., & Romano, R. E. (2011). Peer effects in education: A survey of the theory and evidence. In?Handbook of social economics?(Vol. 1, pp. 1053-1163). North-Holland

Hoxby, C., & Turner, S. (2013). Expanding college opportunities for high-achieving, low income students.?Stanford Institute for Economic Policy Research Discussion Paper,?12(014), 7.

Jensen, R. (2010). The (perceived) returns to education and the demand for schooling.?The Quarterly Journal of Economics,?125(2), 515-548.

Kumar, Anuj and Mehra, Amit, Personalized Education at Scale: Evidence from a Randomized Field Experiment in India (August 10, 2018). Available at SSRN:?https://ssrn.com/abstract=2756059?or?https://dx.doi.org/10.2139/ssrn.2756059

Lavecchia, A. M., Liu, H., & Oreopoulos, P. (2016). Behavioral economics of education: Progress and possibilities. In?Handbook of the Economics of Education?(Vol. 5, pp. 1-74). Elsevier.

Levitt, S. D., List, J. A., Neckermann, S., & Sadoff, S. (2016). The behavioralist goes to school: Leveraging behavioral economics to improve educational performance.?American Economic Journal: Economic Policy,?8(4), 183-219.

Marx, B. M., & Turner, L. J. (2019). Student loan nudges: Experimental evidence on borrowing and educational attainment.?American Economic Journal: Economic Policy,?11(2), 108-141.

Mayer, S. E., Kalil, A., Oreopoulos, P., & Gallegos, S. (2015). Using behavioral insights to increase parental engagement.?THE JOURNAL OF HUMAN RESOURCES,?5(4), 4.

Mayer, S. E., Kalil, A., Oreopoulos, P., & Gallegos, S. (2019). Using behavioral insights to increase parental engagement: The Parents and Children Together intervention.?Journal of Human Resources,?54(4), 900-925.

Morisano, D., Hirsh, J. B., Peterson, J. B., Pihl, R. O., & Shore, B. M. (2010). Setting, elaborating, and reflecting on personal goals improves academic performance.?Journal of applied psychology,?95(2), 255.

Muralidharan, K., A. Singh, and A. J. Ganimian. 2019. “Disrupting Education? Experimental Evidence on Technology-Aided Instruction in India.” American Economic Review, Vol. 109(4): 1426-1460

O’Donoghue, T., & Rabin, M. (1999). Incentives for procrastinators.?The Quarterly Journal of Economics,?114(3), 769-816.

Paunesku, D., Walton, G. M., Romero, C., Smith, E. N., Yeager, D. S., & Dweck, C. S. (2015). Mindset interventions are a scalable treatment for academic underachievement.?Psychological science,?26(6), 784-793.

Riley, E. (2022). Role models in movies: the impact of Queen of Katwe on students’ educational attainment.?Review of Economics and Statistics, 1-48.

Roschelle, J., Shechtman, N., Tatar, D., Hegedus, S., Hopkins, B., Empson, S., ... & Gallagher, L. P. (2010). Integration of technology, curriculum, and professional development for advancing middle school mathematics: Three large-scale studies.?American Educational Research Journal,?47(4), 833-878.

Thaler, R. H., & Sunstein, C. R. (2009).?Nudge: Improving decisions about health, wealth, and happiness. Penguin.

York, B. N., Loeb, S., & Doss, C. (2019). One step at a time: The effects of an early literacy text-messaging program for parents of preschoolers.?Journal of Human Resources,?54(3), 537-566.

?