Simulation Model for Face-to-Face and Distance Education Options Post-COVID-19

The United Nations (2020) has reported that the COVID-19 pandemic affected 1.6 billion students from 190 countries: “Closures of schools and other learning spaces have impacted 94 per cent of the world’s student population, up to 99 percent in low and lower-middle income countries” (UN, 2020, p. 2). The World Bank (2021) has indicated that the short-term challenge is to reopen schools with the appropriate health and safety measures and ensure student retention, while stressing that girls are at greater risk of school dropout. The UN and World Bank agree that in the long term, the challenge is to reverse learning losses (World Bank, 2021; UN, 2020), taken to mean a decline in average overall learning; countries responded to the crisis differently and faced massive school dropout (Iqbal et al., 2020).

School dropout behaves like a dynamically complex system (Barragán, 2015) that should be frequently monitored to strengthen school retention plans. This monitoring, which was already important prior to the pandemic, gained new significance during and after the COVID-19 lockdowns. Explanatory variables can be monitored by adapting Tinto’s (1993) model, accepted by Colombia’s Ministry of National Education (MEN) to diagnose, address, and monitor dropout in higher education as a consequence of the interaction between students and institutions. Diagnostic and analytical reports on school dropout show that the explanatory variables most frequently used to compare determinants of dropout, across different approaches, are those outlined by Tinto in early work in 1975 (Ministry of National Education, 2009).

Tinto’s (1993) integration model categorizes variables into determinants that are endogenous to the education system (academic and institutional determinants) or exogenous (individual and personal determinants); it is therefore useful to apply this model to school dropout by grouping the explanatory variables (see Figure 1) as appropriate for the relevant levels of education and ages. For clarity, and in light of the differences between countries, in the Colombian education system early childhood education is between the ages of 0 and 5 years; primary education is for children from 6 to 10 years of age; lower secondary education is from 11 to 14 years of age; and upper secondary (known as educación media) is for 15 and 16-year-olds.

As Barragán-Moreno and Lozano-Galindo (2022) have noted, Tinto’s (1993) model makes it possible to identify and monitor at least 44 direct and indirect variables that affect student retention and school dropout at pre-higher education levels, to varying degrees depending on student age. For example, the parenthood variable is weighted 0 at young ages, but more heavily in adolescence.

The most predictable explanatory variables of school dropout during lockdowns include those in the socioeconomic determinant, given the challenges of the economic downturn and the preexisting economic vulnerability that exacerbated significant differences in the ability to cope with isolation (World Bank, 2021), in addition to child marriage, teenage pregnancy, and gender-based violence, which affect girls and young women (UN, 2020). Meanwhile, Ramírez and Hernández (2021) highlighted how the mandatory use of online learning for preschoolers demonstrated inequalities in access to this form of education, with preschool serving as the foundation for socioemotional and intellectual development through face-to-face interaction. It is against this backdrop that the following guiding research questions were formulated.

RQ1: What are the main explanatory variables of school dropout in preschool, basic, and upper secondary education in Colombia during the COVID-19 lockdown?

RQ2: What is the long-term behavior of the variables under the different education options – lying on a continuum between face-to-face and distance learning – for reducing school dropout in preschool, basic, and upper secondary education in Colombia, in the aftermath of COVID-19?

This article aimed to simulate the behavior of school dropout in preschool, basic, and upper secondary education in Colombia in the aftermath of COVID-19, based on the degree of online instruction under various alternatives on a continuum between face-to-face and distance learning.

1.1 Emergency changes, responses, and options

García (2021) has noted that the pandemic forced a drastic change in the provision of education services due to the mandatory closure of facilities. Governments found themselves compelled to turn to digital distance learning, using available information technologies, to prevent interruptions in education. This resulted in distinct responses both between and within countries due to limitations in technological infrastructure and planning within education systems (Herrera, 2021), coupled with the challenge of preparing teachers and students required to deliver and receive a “digital distance education” or “remote emergency education” that emerged as a solution to address the need for educational continuity, maintaining physical distance and enabling synchronous and asynchronous interaction.

One response to the change was to adjust the student-teacher ratio to monitor student progress, adapting in-person activities to the virtual classroom. This demanded a change in teaching and learning methods, with proposals for flexible teaching approaches adapted to the health and safety measures, student group sizes and readiness, and teacher readiness and attitudes. Teachers were quick to incorporate information and communication technology (ICT) and learning and knowledge technologies (LKTs) (Latorre et al., 2018) to support their work. In response to uncertainty about an unknown virus and the potential emergence of other viruses, schools underwent a transition from face-to-face learning (“normal” conditions) to hybrid learning (with some students attending in person, and others remotely with digital support), with a fully remote option.

Distance education has been recommended in higher education and has posed challenges for curricular adaptation and for evaluation in these new environments. In higher education, a lack of student and teacher readiness became clear and revealed huge differences in access to hardware, software, and connectivity, coupled with a poor ability to use educational platforms despite the users being digital natives (Díaz Barriga & Barrón, 2022; Guevara, 2021). These differences were shown more acutely in a georeferenced analysis of external test results, which found better outcomes in areas with better infrastructure and connectivity (cities), as opposed to more remote locations away from urban areas (Arias-Velandia et al., 2021).

Distance education is scarcely recommended in preschool, basic, and upper secondary education, where teacher-student and student-student relationships are key to furthering the education process, as it is through these relationships that affective and emotional bonds are developed that influence knowledge appropriation. Health and safety protocols and lessons learned made it necessary to consider flexible learning scenarios during lockdown and post-lockdown periods to reduce school dropout, maintain levels of quality, and mitigate the effect of economic or geographic variables or any other circumstance that may impede access to official education (Barragán-Moreno & Lozano-Galindo, 2022; García, 2021).

Figure 2 illustrates the range of options that were implemented, or which could be implemented, as ways to adapt the teaching-learning process and pedagogical approaches, in line with learning outcomes, the pedagogical model, and the school’s education plan. These alternatives depend on the contexts in which teaching practices take place, the availability of technological resources, and teachers’ and students’ technological skills and abilities, to limit the likelihood of reproducing the current disparities in education and dropout (Herrera, 2021).

The methodology consisted of two phases, which were carried out consecutively:

Phase 1: Meta-analysis. Analysis of 21 research articles published in 2020-2021, retrieved by searching 19 databases (see Figure 3), with 24 specific search strings related to student persistence and dropout. The sample of articles was obtained after identifying, selecting, screening, and excluding related documents, as per the PRISMA 2020 guidelines (Page et al., 2021).

Phase 2: Scenario modeling and simulation. For this phase of the research, a systemic approach was employed given the complex nature of the basic education system in Colombia, which comprises multiple stakeholders and countless interactions that can be distilled into two subsystems: educational demand and supply. It is the interaction between these two subsystems that determines the behavior of the education system as a whole.

In this phase, it is very important to view the feedback and effects that arise from the changes in the key variables in the system over time (see Appendix I). There is therefore a need to present the results of the system’s performance when social interventions (simulations) are performed, which can be designed and implemented with the support of computer tools, based on the historical context and future projections of the impacts of public education policy measures on the system (Bianchi & Salazar, 2020; Senge, 2010; Williams & Hummelbrunner, 2011). For this, we described the behavior using a causal loop diagram, in order to identify the hypotheses and roles played by the various elements in the subsystems.

The causal loop diagram shows both the system’s reference mode and the dynamic hypotheses and feedback structure. The interactions between the system’s variables are represented by arrows marked with positive or negative signs. A positive sign indicates a positive relationship between the variables (when one increases, the other also increases), while a negative sign indicates an inverse relationship (when one variable increases, the other decreases) (Bala et al., 2017). Lastly, a stock-and-flow model was used as a system dynamics simulation model, combining modeling as a scientific practice, modeling as model progression, and the instrumental dimension of modeling (Oliva, 2019).

In line with the model progression approach, the model designed by Serna and Flores (2015) to study educational policies in Peru was employed as a starting point. For the model progression, we included variables that were found in phase 1 to be related to the teacher’s role, the percentage of online instruction in the various educational modalities available following the COVID-19 lockdown, and the effects of the variables in the individual, institutional, and socioeconomic determinants, with the aim of leveraging the similarities between the Peruvian model and the dynamics of dropout in Colombia within this new context. Similarly, studies were included that reported multiple factors in the causes of dropout, such as COVID-19 infection, a rural education model ill-prepared for remote education, a lack of continuity in the inclusion processes of migrant students, and inequality in terms of technology access and availability to educate the population, both in Colombia and across Latin America (Mauris & Domínguez, 2022; Pérez, 2022; Rodríguez-Lizarralde et al., 2022; Sañudo, 2022). Then, as a scientific practice, the model was parameterized both to validate its operability and to perform simulations; this was done with data on preschool, basic, and upper secondary education in Colombia available in public databases. The simulation recorded behavior in three scenarios, revealing the effect on student dropout of different alternatives falling between in-person and fully remote instruction (Figure 2).

The phases of research described above identified variables that were incorporated into the model and yielded results from the system dynamics model.

3.1 Variables that explain dropout

Initially, our systematic review located 77,700 records in the 19 databases, with 24 specific search strings related to student persistence and dropout. Of these, 75,809 records were eliminated as duplicates, leaving 1,891 for screening in the title and abstract to determine eligibility for processing with Atlas.ti. A sample of 25 research articles published between January 2020 and December 2022 was included in the qualitative synthesis of the review (Figure 4).

Most of the variables identified in articles that made mention of the meaning, value, or effect of a specific variable fell under the academic determinant (57.14%), followed by the institutional (28.57%), socioeconomic (9.52%), and lastly, individual (4.76%) determinants.

Research on the academic determinant has identified “teachers” as the most important variable in explaining dropout. Findings have shown that “teachers” are viewed as individuals who engage and provide tools to students to support them in their learning process and facilitate new relationships between students, teachers, and parents. It is teachers who are responsible for ensuring the right to education in a context of isolation and physical distancing, which combine with Colombia’s socioeconomic inequalities, reflected in a Gini index of 0.57% and the country’s ranking as the eighth most unequal, according to data from the Economic Commission for Latin America and the Caribbean (ECLAC) and the World Bank (Herrera, 2021).

At the same time, it was found that simply providing technology was not sufficient to yield better learning outcomes; rather, learning should be facilitated by teachers well versed in technology (Reimers & Schleicher, 2020; Sanabria et al., 2020). Another key aspect in this determinant is the lack of teacher readiness to cater to groups with special educational needs, a critical issue that deserves attention given the need for interaction between families, schools, and professionals working together toward equity (Díaz-Piñeres et al., 2020). It was also established that this variable includes limited teacher readiness in the technical and administrative aspects of ICT and LKTs to apply these tools to solve problems and support engagement and active teaching to improve teaching practice. This improvement motivates teachers and students in the teaching-learning process, compelling teachers to acquire the necessary technical and pedagogical skills to integrate them into this form of education (Avendaño et al., 2021; García, 2021; Sánchez et al., 2020).

The “curriculum matrix” variable encompasses various curricular, pedagogical, didactical, and attitudinal aspects of teachers, bringing together student autonomy resulting from the lack of teacher contact, the courses offered, ICT or LKT support, and the necessary accommodations to leave no student behind for socioeconomic reasons or due to special educational needs. The intention is to capture, through the structure of this “curriculum matrix,” a proposal that allows for flexibility in resources, materials, deadlines, and assessments, with the goal of achieving pedagogical renewal and innovation. The objective is to respond to the crisis and equip educational stakeholders to migrate to new hybrid, in-person, and distance models, by prioritizing and defining what should be learned during physical distancing, but also to answer questions and verify learning (García, 2021; Díaz-Piñeres et al., 2020; Reimers & Schleicher, 2020; Serna et al., 2020).

Notable in the institutional determinant was “political environment.” In previous years, the political environment referred to a context of armed conflict, violence, and the Peace Agreement, but in 2020-2021, this variable took on a new meaning in reference to the Colombian government’s obligation to formulate and implement specific public policies to guarantee accessibility, opportunity, quality, and inclusion in education, taking into account specific regional characteristics and realities and helping to close the gap between urban and rural education in the pandemic and post-pandemic eras (García & Maldonado, 2020; Herrera, 2021; Maroscia & Ruiz, 2021). The “political environment” variable was associated with public policy in response to a need to prioritize support for students and schools most lacking in infrastructure and independent study skills, including wellbeing activities for both students and teachers (Findeter, 2021).

The “infrastructure” variable no longer referred to physical facilities and instead came to mean technological infrastructure, including the necessary adaptations for ICT to be accessed by students without the basic devices or connectivity necessary for academic progress, and for teachers to more easily fulfill their obligation to deliver education digitally, facilitating technical support for synchronous and asynchronous remote sessions that allow individual and group-based academic work and make it possible to deliver and assess distance learning. All of this also means overcoming obstacles like problems with connectivity and devices, and a rejection of or aversion to technology among teachers and students accustomed to learning in a physical space that has been transposed to a virtual room.

In the socioeconomic determinant, the variable “location of residence” was most cited, due to the differences in access between schools located in urban and rural areas. The switch from in-person to digital formats exposed even more clearly the historical education gap between towns and more rural settings. The lack of technological devices is more apparent in rural areas, where the ability to generate economies of scale in the education service is hampered by the geographical distances, which renders the provision of transportation, food, and accessibility costly and discourages students from continuing their studies due to the lack of physical and technological conditions or because a full range of courses is not offered (García & Maldonado, 2020), hindering the mitigation of school dropout (Díaz-Piñeres et al., 2020; Elacqua et al., 2020; Herrera, 2021; Osiobe, 2020).

In the individual determinant, the most frequently cited variable was “state of health,” within a context of concern and uncertainty about the consequences of the pandemic. This compounded preexisting problems of malnutrition, domestic violence, and migration, which exacerbated student dropout and which, in any crisis, have a greater impact on more vulnerable groups (Maroscia & Ruiz, 2021).

3.2 Simulation model for options on the continuum from in-person to distance education

After identifying the variables, we proceeded to the second phase of the methodology and obtained the following causal loop diagram as an initial result (Figure 5).

Figure 5 shows three loops that provide a qualitative representation of the following effects. First is the effect of teacher training on dropout: the greater the number of trained teachers, the better the curriculum matrix, which also improves results in government tests taken at the end of high school. This improvement has a positive effect on overall education outcomes and reduces the dropout rate. There are fewer dropouts, resulting in an increase in the active student population, which requires an increase in teaching staff. In turn, this demands an increase in training time so that a greater number of trained teachers are available.

Second is the effect of contextualized student-centered activities on students in schools. The greater the number of activities that involve students, the better their results, including the results of standardized government testing. This improves education outcomes and reduces the dropout rate and the number of dropouts, allowing more students to remain in school, which in turn requires a greater number of contextualized activities.

Third is the effect of students in schools on student dropout. When a school has a high number of students, there is a need for more teaching staff and hence more training. When more teachers are trained, this improves the curriculum matrix and government test results, thus improving education outcomes, which reduces the number of dropouts and increases the active student population in schools.

Figure 6 shows the model developed using system dynamics to simulate three scenarios that demonstrate the impact on school dropout of the percentage of remote instruction in the different options on the continuum from in-person to remote instruction in preschool, basic, and upper secondary education.

The model includes an important assumption: at any percentage of virtual instruction, including the extremes of 0% (fully face-to-face) and 100% virtual learning, teachers receive curricular training in ICT and LKTs to develop their sensitivity toward the teaching profession and teaching practice. The teaching profession requires ongoing professional development in both teaching and subject-matter skills.

3.3 First scenario (SIM-1): assumptions and initial values

To begin the long-term trend analysis, we employed the variables and parameters listed in Appendix I. The parameters for the system’s reference model (initial conditions) include, notably, 10% online instruction in planned activities and 40 hours of training per teacher per year for a potential student population of 8,613,646 and an annual dropout rate of 15.79%. SIM-1 showed that dropout displays horizontal asymptotic behavior at 129,281 students dropping out per year, nationwide (Figure 7).

The simulations for the second and third scenarios were performed ceteris paribus; the second scenario only varied the parameter corresponding to the percentage of online instruction in academic activities (five simulations), and in the third simulation, only the number of hours of teacher training varied (two simulations).

3.4 Second scenario: progressive implementation of online instruction

For this analysis, five simulations were performed, in which we introduced incremental increases to the percentage of online instruction received by children and adolescents in the primary and lower secondary stages of basic education and in upper secondary education. The simulations were performed for the period from 2021 to 2030, in line with the horizon for achieving the Sustainable Development Goals (SDGs). As shown in Table 1, student dropout becomes more frequent as the percentage of online instruction increases. For example, when 20% of academic activities are online, student dropout stabilizes at 254,498 students, while moving all activities online results in over one million children and adolescents dropping out of school.

Figure 8 shows the sensitivity of students who drop out of lower and upper secondary education to the percentages of online instruction, with the horizontal asymptotes (long-term trends) of student dropout behavior.

3.5 Third scenario: teacher training

In this instance, two computer simulations were run for the parameter corresponding to teacher training hours as a tool to improve teaching practice and develop teacher sensitivity toward the profession. Importantly, all the other parameters were kept as in Appendix I – that is, as in the SIM-1 model. The initial parameter value of 40 hours of annual training per teacher is what is required in one week of institutional development each year (see Appendix I). Simulations SIM-7 and SIM-8 were performed with 80 and 120 hours of annual training per teacher, which is equivalent to two and three weeks of institutional development, respectively. In the long term, the student dropout trend is unchanged, at 129,281 students. In other words, dropout is not sensitive to teacher training under the conditions given. No simulations were run for longer durations – for example, for five weeks of institutional development in the academic year, as established by Decree No. 1075 of 2015 (Decreto No. 1075, 2015) – because that would be for large-scale specialization courses or graduate programs taken by teachers, and that is not feasible here.

The systematic review found that during the COVID-19 lockdown, 57.1% of studies on the determinants of student dropout in basic and upper secondary education in Colombia focused on the academic determinant, within which the most frequently cited variable was “teachers,” who were identified firstly as being responsible for maintaining a link between students, teachers, and parents with the education service in a sudden, unexpected context of isolation and physical distancing, while also taking into account the conditions of social, economic, and technological inequality. Secondly, the studies note teachers’ lack of personal and professional skills in the use of tools specific to remote education (ICT and LKTs). These are factors that make it difficult to provide education in the space between in-person and fully remote learning during pandemic and post-pandemic times.

The second most frequently cited variable in the academic determinant was “curriculum matrix,” for which the aim is to provide an approach to teaching that allows flexibility in resources, materials, deadlines, and assessments, thus enabling the various stakeholders in the education process to migrate to the hybrid models of education proposed in the continuum between in-person and remote instruction in this new normal.

At 28.75% of the documents retrieved, the institutional determinant became the second most widely studied determinant and included “political environment” as the most frequently cited variable. Previously, this referred to a context of armed conflict, violence, and the Peace Agreement, but in 2020-2021 it took on a new meaning in the context of the Colombian government’s obligation to formulate and introduce public policies that guarantee accessibility, opportunity, quality, and inclusion in the education system, taking into consideration the specific characteristics of each local area, at all times endeavoring to close the gaps between urban and rural education. Lastly, within the institutional determinant, “infrastructure” was given a new meaning as it shifted from a physical to a technological connotation. This meant there was a need to introduce changes in order to provide education to students who did not all have access to electronic devices or an internet connection, which became indispensable to progress academically. As far as teachers were concerned, this meant further developing the compulsory digitalization of education, providing technical support for synchronous or asynchronous class sessions for individual or group work and making it possible to deliver and assess a new form of education remotely.

Finally, the socioeconomic determinant concerned 9.52% of studies and “location of residence” was the most frequently cited variable, in light of its implications for access and connectivity in the areas served by urban and rural schools.

By following the model progression approach and incorporating the variables into the baseline model proposed by Serna and Flores (2015), we found that student dropout was highly sensitive to the percentage of online instruction in the education received by children and adolescents in the Colombian system, and not sensitive to the duration of teacher training, which ranges from 40 to 200 hours a year. The difficulties of virtual learning are not offset by a focus on developing teachers’ sensitivity to their profession and acquiring and improving teaching skills that promote pedagogical, didactical, and ICT and LKT competencies that build support for students in each school and in the system more broadly. This also fails to take into account that during training weeks, time must also be devoted to the institutional educational project, curriculum, pedagogical research and refesher training, institutional evaluation, and other activities for delivering education.

This poses a challenge for the system used by Colombia’s Ministry of National Education (MEN) to train educators, which is designed to effectively provide the necessary pedagogical tools to deliver education, taking into account students’ abilities and needs to enable them to learn better and learn more, and develop competencies for life (Ministerio de Educación Nacional, 2021).

The highest dropout rate was observed in the simulations with 100% online learning, potentially mirroring a situation in which students are under the academic supervision of parents or caregivers and carry out activities and instructions provided by teachers through photocopies, online platforms, or e-mails. Under this extreme scenario, children and adolescents have no in-person contact with their teacher or classmates, which combines with persistent regional inequalities in education in Colombia (Herrera, 2021) to trigger massive levels of school dropout, a phenomenon that was already severe given that “out of every 100 students that enter 5^th grade, only 55 manage to reach 11^th grade” (García & Maldonado, 2020, p. 6) and, similarly, out of every 100 students who begin 6^th grade, 64 finish 11th grade (Forero & Saavedra, 2019), the last year of high school.

Lower levels of online instruction result in lower levels of dropout in the simulations. This could inform responses to adapt to any future lockdowns or peaks of infection, using alternatives or a combination of measures that ensure no child or adolescent is left behind (García, 2021). In any case, teacher attitudes make a difference in inclusive education (Rodríguez et al., 2021) and in understanding the contexts in which the teaching-learning process takes place (Herrera, 2021). Moreover, the social situation in Colombia in 2020 and 2021 due to the pandemic, which was further exacerbated by insecurity challenges in 2021, points to an increase in poverty rates and physical and mental health conditions, which should be tackled with proposals that acknowledge these deficiencies and work to strengthen education as a healthy life project – mentally, physically, economically, and socially.

The two phases of the research methodology successfully simulated the behavior of school dropout at preschool, basic, and upper secondary levels in Colombia in the aftermath of COVID-19. Using the system dynamics technique to simulate the behavior of the main variables in the proposed alternatives on the continuum between face-to-face and distance learning, we found that a higher level of online instruction (increased in 20% increments in each simulation) resulted in an increase in dropout, given the technological infrastructure and teacher and student ICT and LKT readiness. At the same time, an attempt to mitigate dropout by increasing teacher training hours, within the limits of teachers’ annual minimum training program set by education policy, revealed the insensitivity of dropout in this respect. Training just one set of actors in the education process is therefore insufficient. Instead, joint efforts are needed on several fronts to facilitate access to minimum ICT tools and connectivity.

Sustained teacher training in technological appropriation, skills, and abilities improves teacher competencies recommended by UNESCO; this recommendation should be thoroughly adopted by Colombia throughout all stages of education to address the specific challenges of the reality of education that may be both face-to-face and remote. This continuum from face-to-face to distance learning demands immediate efforts to develop connectivity, as part of the right to education enjoyed by children and adolescents worldwide.

Educational research helps policymakers to formulate wide-ranging, impactful solutions and evaluate how they function, in order to implement, adjust, or undertake measures to benefit the commmunity. However, a study like this has limitations due to the delay in publishing research findings (given the time required for the editorial process, which means that data are already at least one year old when work is published). Another limitation is the lack of data to describe schools and students in terms of ICT availability and access. Highly accurate data make it possible to assess the real-world situation, offering a window of opportunity for formal research and knowledge generation that benefits all stakeholders in the education process.

Translation: Joshua Parker

Contribution of each author

Sandra Patricia Barragán-Moreno: conceptualization, investigation (60%), formal analysis, writing original draft (60%).

Oscar Leonardo Lozano Galindo: methodology, investigation (40%), data curation, writing original draft (40%).

Declaration of no conflict of interest

The authors declare no conflict of interest.

Source of funding

Unfunded research.

References

Arias-Velandia, N., Rincón-Báez, W. U. y Cruz-Pulido, J. M. (2021). Diferencia de logro geolocalizado en educación presencial y a distancia en Colombia [Geolocated achievement differences in face-to-face and distance education in Colombia]. Revista Electrónica de Investigación Educativa, 23, e18, 1-22. https://doi.org/10.24320/redie.2021.23.e18.3711

Avendaño, W. R., Hernández, C. A. y Prada, R. (2021). Uso de las tecnologías de información y comunicación como valor pedagógico en tiempo de crisis [Use of information and communications technology as a pedagogical asset in times of crisis]. Revista Historia de la educación Latinoamericana, 23(36). https://doi.org/10.19053/01227238.11619

Bala, B. K., Arshad, F. M. y Noh, K. M. (2017). Systems thinking. System dynamics. Springer. https://doi.org/10.1007/978-981-10-2045-2_2 

Barragán, S. P. (2015). La Universidad de Bogotá Jorge Tadeo Lozano en el camino de la retención estudiantil: más cerca de la reducción del abandono estudiantil en la Tadeo [Jorge Tadeo Lozano University of Bogotá (UTADEO) on the path of student retention: Moving closer to reducing student dropout in Tadeo]. Fundación Universidad de Bogotá Jorge Tadeo Lozano.

Barragán-Moreno, S. P. y Lozano-Galindo, O. L. (2022). Explanatory variables of dropout in Colombian public education: Evolution limited to coronavirus disease. European Journal of Educational Research, 11(1), 287-304. https://doi.org/10.12973/eu-jer.11.1.287

Bianchi, C. y Salazar, R. (2020). A feedback view of behavioural distortions from perceived public service gaps at ‘street‐level’ policy implementation: The case of unintended outcomes in public schools. Systems Research and Behavioral Science, 39(1), 63–84. https://doi.org/10.1002/sres.2771

Decreto No.1075 (mayo 26 de 2015). Por medio del cual se expide el decreto único reglamentario del sector educación [By which the sole regulatory decree of the education sector is issued]. Congreso de la República de Colombia. https://bit.ly/3QD6ZOj

Díaz Barriga, F. y Barrón, M. C. (2022). Desafíos del currículo en tiempo de pandemia: innovación disruptiva y tecnologías para la inclusión y justicia social [Curriculum challenges in times of pandemic: Disruptive innovation and technologies for inclusion and social justice]. Revista Electrónica de Investigación Educativa, 24, e10 1-12. https://doi.org/10.24320/redie.2022.24.e10.4500

Díaz-Piñeres, A., Bravo-Rueda, C. y Sierra, G. E. (2020). Educación inclusiva en contexto: reflexiones sobre la implementación del Decreto 1421 de 2017 [Inclusive education in context: Reflections upon the implementation of Decree 1421 of 2017]. Revista Historia de la Educación Latinoamericana, 22(34), 265-290. https://doi.org/10.19053/01227238.9823

Elacqua, G. M., Sánchez, F. y Santos, H. (2020). School reorganization reforms: the case of multi-site schools in Colombia. School Effectiveness and School Improvement, 32(1), 141-172. https://doi.org/10.1080/09243453.2020.1797830

Findeter. (octubre, 2021). Estudio del sector educativo colombiano [Study of the Colombian education sector]. https://bibliotecadigital.findeter.metabiblioteca.com/server/api/core/bitstreams/68d7ffc5-9013-408b-b2d0-4a95b036be62/content

Forero, D. y Saavedra, V. (2019). Los 10 pasos para ser Colombia la mejor educada de América Latina [10 steps to make Colombia the best-educated country in Latin America]. Fedesarrollo. https://bit.ly/3vV29Ta

García, L. (2021). COVID-19 y educación a distancia digital: preconfinamiento, confinamiento y posconfinamiento [COVID-19 and digital distance education: Pre-lockdown, lockdown, and post-lockdown]. RIED. Revista Iberoamericana de Educación a Distancia, 24(1), 9-32. https://doi.org/10.5944/ried.24.1.28080

García, S. y Maldonado, D. (2020). Universalización de la educación media de calidad: el reto sigue para llegar a ser “Colombia la mejor educada” [Universalization of quality high school education: The next challenge in making Colombia “the best-educated country”]. Tribuna: Revista de Asuntos Públicos, (13), 4-8. https://repositorio.uniandes.edu.co/entities/publication/2dd30b99-837e-4fc6-8c2b-79e45118863b

Guevara, A. (2021). Evaluación de los aprendizajes en tiempos de COVID-19, el caso del estado de Chihuahua [Assessment of learning during COVID-19: The case of the state of Chihuahua]. Revista Electrónica de Investigación Educativa, 23, e17, 1-16. https://doi.org/10.24320/redie.2021.23.e17.4335

Herrera, D. Y. (2021). El modelo de la alternancia y la desigualdad educativa territorial en la educación en Colombia [The alternation model and regional inequality in education in Colombia]. Revista Internacional de Pedagogía e Innovación Educativa, 1(2), 61-86. https://doi.org/10.51660/ripie.v1i2.38

Iqbal, S., Azevedo, J., Geven, k., Hasan, A. y Patrinos, H. (2020, 13 de abril). We should avoid flattening the curve in education – Possible scenarios for learning loss during the school lockdowns. World Bank Blog: Education for Global Development. https://blogs.worldbank.org/en/education/we-should-avoid-flattening-curve-education-possible-scenarios-learning-loss-during-school

Latorre, E. L., Castro, K. P. y Potes, I. D. (2018). Las TIC, las TAC y las TEP: innovación educativa en la era conceptual [ICT, LKT, and EPT: Innovation in education in the conceptual era]. Universidad Sergio Arboleda. https://bit.ly/3QzAWyW

Maroscia, C. y Ruiz, P. C. (2021). Las organizaciones de la sociedad civil en época de pandemia. Reflexiones hacia una nueva normalidad: ¿Nuevos desafíos o mismas realidades? [Civil society organizations in times of pandemic. Reflections toward a new normal: New challenges or the same realities?] Ciencias Administrativas, (17), 97-107. https://doi.org/10.24215/23143738e079

Mauris, L. A. y Domínguez, B. P. (2022). Los efectos de la crisis sanitaria del Covid-19 en la educación rural de Colombia [The effects of the COVID-19 health crisis on rural education in Colombia]. Panorama, 16(30), 1-21. https://doi.org/10.15765/pnrm.v16i30.3023

Ministerio de Educación Nacional. (2009). Deserción estudiantil en la educación superior colombiana [Student dropout in higher education in Colombia]. https://www.mineducacion.gov.co/sistemasdeinformacion/1735/articles-254702_libro_desercion.pdf

Ministerio de Educación Nacional. (2021, 23 de Junio). Sistema colombiano de formación de educadores [The Colombian educator training system]. https://cms.mineducacion.gov.co/static/cache/binaries/articles-345485_anexo1.pdf?binary_rand=4264

Oliva, J. M. (2019). Distintas acepciones para la idea de modelización en la enseñanza de las ciencias [Different definitions for the idea of modeling in science education]. Enseñanza de las Ciencias, 37(2), 5-24. https://doi.org/10.5565/rev/ensciencias.2648

Osiobe, E. U. (2020). Understanding Latin America´s educational orientations: Evidence from 14 nations. Education Quarterly Reviews, 3(2), 249-260. https://ssrn.com/abstract=3620652

Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., Chou, R., Glanville, J., Grimshaw, J. M., Hróbjartsson, A., Lalu, M. M., Li, T., Loder, E. W., Mayo-Wilson, E., McDonald, S.... y Moher, D. (2021). The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Systematics Reviews, 10(89). https://doi.org/10.1186/s13643-021-01626-4

Pérez, A. (2022). Educación, desigualdad sociodigital en pandemia [Education, sociodigital inequality, and the pandemic]. El Cotidiano, 233, 69-76. https://www.researchgate.net/publication/389135541_Educacion_desigualdad_sociodigital_y_pandemia_El_Cotidiano

Ramírez, G. y Hernández, S. M. (febrero, 1 de 2021). Una mirada a la educación preescolar desde la diversidad en pandemia Covid-19 y sus afectaciones [A look at preschool education from a diversity perspective in the COVID-19 pandemic and its impacts]. [Degree project, Bachelor’s Degree in Preschool Education]. Universidad Santo Tomás, Colombia. http://hdl.handle.net/11634/31926

Reimers, F. M. y Schleicher A. (2020). Un marco para guiar una respuesta educativa a la pandemia del COVID-19 [A framework to guide an education response to the COVID-19 pandemic]. Enseña Perú. https://globaled.gse.harvard.edu/files/geii/files/un_marco_para_guiar_una_respuesta_educativa_a_la_pandemia_del_2020_del_covid-19_.pdf

Rodríguez, A., Gallego, J. L., Navarro, A. y Caurcel, M. J. (2021). Perspectivas actitudinales de docentes en ejercicio y en formación hacia la educación inclusiva [Attitudinal perspectives of in-service and trainee teachers toward inclusive education]. Psicoperspectivas, 20(1), 1-13. https://doi.org/10.5027/psicoperspectivas-Vol20-Issue1-fulltext-1892

Rodríguez-Lizarralde, C., López-Villamil, S. y Barrera-García, A. D. (2022). Inclusión educativa en pandemia: interseccionalidad y situación de menores venezolanos en Bogotá [Educational inclusion in a pandemic: Intersectionality and the situation of Venezuelan minors in Bogotá]. Revista Latinoamericana de Ciencias Sociales, Niñez y Juventud, 20(2), 1-27. https://dx.doi.org/10.11600/rlcsnj.20.2.5123

Sanabria, L. A., Pérez, M. C. y Riascos, L. E. (2020). Pruebas de evaluación Saber y PISA en la Educación Obligatoria de Colombia [The Saber and PISA assessment tests in compulsory education in Colombia]. Education Siglo XXI, 38(3), 231-254. http://dx.doi.org/10.6018/educatio.452891

Sánchez, P. A., Saavedra, C. E. y Figueroa, C. (2020). Efectos de las tecnologías de la información y las comunicaciones en la educación en Colombia [Effects of information and communications technology in education in Colombia]. Revista Espacios, 41(48), 303-313. https://www.revistaespacios.com/a20v41n48/a20v41n48p22.pdf

Sañudo, L. (2022). Del abandono a la permanencia escolar en secundaria [From dropout to student persistence in secondary school]. Profesorado, Revista de Currículum y Formación de Profesorado, 26(1), 213- 233. https://revistaseug.ugr.es/index.php/profesorado/article/view/13535

Senge, P. (2010). La quinta disciplina (2.ª ed.) [The fifth discipline (2nd ed.)]. Granica.

Serna, J. S., Martínez, D. J. y Arrubla, J. A. (2020). Educación rural y dispositivo evaluación en tiempos de ‘COVID-19’: voces de profesores de matemática [Rural education and evaluation apparatus in times of COVID-19: The voices of mathematics teachers]. Revista Latinoamericana de Etnomatemática, 13(1), 86-103. https://doi.org/10.22267/relatem.20131.43

Serna, R. y Flores, Z. A. (2015, 22 de Julio). Modelación dinámica del efecto de las políticas del gobierno que se implementaron para la educación en Perú desde 1990 hasta 2010 [Dynamic modeling of the effect of government policies implemented in education in Peru from 1990 to 2010]. 13th LACCEI Annual International Conference: “Engineering Education Facing the Grand Challenges, What Are We Doing?”, Santo Domingo, Dominican Republic. https://www.laccei.org/LACCEI2015-SantoDomingo/StudentPapers/SP054.pdf

Tinto, V. (1993). Leaving college. Rethinking the causes and cures of student attrition (2.ª ed.). University Chicago Press.

United Nations. (2020, August). Policy brief: Education during COVID-19 and beyond. https://bit.ly/3Xoa0Eg

Williams, B. y Hummelbruner, R. (2011). Systems concepts in action. Stanford University Press.

World Bank. (2021, Abril). Global economic prospects. https://www.worldbank.org/en/publication/global-economic-prospect