Next Article in Journal
Exploring the Spatial Effects of Built Environment on Quality of Life Related Transportation by Integrating GIS and Deep Learning Approaches
Previous Article in Journal
Effect of Sludge Content on the Decomposition of Different Types of Food Waste
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Sustainability
Volume 15
Issue 3
10.3390/su15032784
sustainability-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

A Heterogeneity Study on the Effect of Digital Education Technology on the Sustainability of Cognitive Ability for Middle School Students

by
Desheng Yan
1 and
Guangming Li
2,3,*
1
School of Education, South China Normal University, Guangzhou 510631, China
2
School of Psychology, South China Normal University, Guangzhou 510631, China
3
Center for Studies of Psychological Application, Guangdong Key Laboratory of Mental Health and Cognitive Science, South China Normal University, Guangzhou 510631, China
*
Author to whom correspondence should be addressed.
Sustainability 2023, 15(3), 2784; https://doi.org/10.3390/su15032784
Submission received: 28 December 2022 / Revised: 28 January 2023 / Accepted: 1 February 2023 / Published: 3 February 2023
Browse Figure
Versions Notes

Abstract

:
Digitalization gradually transforms digital education technology from being a teaching means to focusing on the student’s abilities. This study analyzes the data from the China Education Baseline Survey of the Renmin University of China using Coarsened Exact Matching (CEM) and quantile regression methods. The Ordinary Least Squares (OLS) regression is used to test the net effect of digital education technology on students’ academic and cognitive abilities. The OLS result shows that digital education technology has a significantly positive impact on the cognitive ability of middle school students. However, schools focusing on using digital education technology as a means of school management will lower students’ cognitive abilities. Second, the CEM method result shows a significant difference in the cognitive ability scores between students in classrooms with and without digital education technology. This indicates that digital education technology can inspire students’ internal drive, motivate them to learn, and enhance their cognitive abilities. Last, the quantile regression result shows that the use of digital education technology has heterogeneity in the development of the cognitive ability of middle school students. When the cognitive ability quantile point increases, the influence becomes more obvious, changing from not significant to significant, then to very significant. This provides some inspiration for understanding the application of digital education technology. It can be concluded that this is an effective method to study the sustainable development of cognitive abilities through heterogeneity. The method ensures individuals’ holistic and comprehensive development, thus promoting groups’ development and contributing intellectual support to them to adapt to future social requirements and lead future social development. Digital education technology endows students with the cognitive abilities for lifelong learning to solve various problems in future social life, reserve high-quality talent resources for the future, and build a learning society with extensive significance. This paper analyzes the sustainable development of students’ cognitive abilities using emerging digital education technologies. It not only deepens the understanding of sustainable development of education and humans but also provides intellectual support for society’s sustainable development.

1. Introduction

In the rapid development era of information technology, digital education technology characteristics, including human–machine collaboration, cross-border integration, and collective creation, can significantly impact student development, the educational system, and the teaching mode [1]. Some scholars believe that digital education technology can help students to improve their learning ability [2], inspire their learning enthusiasm, enhance their understanding of knowledge shown from the cognitive aspect [3], and improve their advanced thinking ability [4]. Thus, digital education technology is recognized as an important instrument for the development of children’s perceptual and cognitive skills. However, studies on the effects of digital education technologies on students’ cognitive abilities are debatable. There is an ongoing debate regarding the impact of Internet use, computers, and digital tools on students’ cognitive abilities [5]. George et al. have found that continuous multitasking using digital devices damages adolescents’ cognitive abilities [6]. Tarus et al. believe that the benefits brought by digital learning far outweigh the challenges [7]. It is clear that the impact of digital education technology on middle school students’ cognitive influence has not reached a universal agreement. Further rethinking can reduce the uncertainty in this field because little attention has been paid to the heterogeneity of this impact. In other words, most studies are based on the experience of technology applications or limited sample survey methods. These studies mainly use the data selected from a specific range of small samples by using structural equations [8], multiple regression, mediation effect analysis, and other methods to research, such as using structural equation modeling to analyze the effectiveness of digital learning technology [9]. However, there are often highly complex relations in the teaching field. Further expansion is needed regarding sample typicality, representativeness, data coverage, randomness, or analysis techniques. This study explains a relatively accurate net effect mechanism using the Coarsened Exact Matching (CEM) method to reduce the selection bias of the sample. After analyzing the heterogeneity of this effect, this study provides an accurate answer for the impact of digital education technology on cognitive ability. It has significant theoretical and practical value for improving the efficacy of digital teaching and developing high-level cognitive ability. The research questions include: (1) Does digital education technology affect the cognitive ability of middle school students? (2) Is there any heterogeneity in the impact of digital education technology?
In terms of technical characteristics, digital education technology provides a comprehensive, dynamic, and opening-up learner model to meet the sustainable development needs of students’ cognitive abilities. Mobility, novelty, intelligence, and adaptability constitute the primary technical framework of digital education technology. In terms of educational characteristics, digital education technology can effectively enhance learners’ personalized learning experience, help learners develop advanced cognitive skills, and cultivate the sustainable ability of autonomous learning and lifelong learning [10]. Supported by digital education technology, the sustainable development of students’ cognitive abilities is conducive to the digitalization of social individuals. Meanwhile, it is instrumental in building a smart society to realize the participation of all the people and the free development of humans. Therefore, in the information age, the sustainable development of cognitive abilities brings unlimited possibilities for the future society. Cognitive and non-cognitive abilities provide positive interventions on students’ performances after entering the labor markets and determine social and economic success. Cognitive abilities affect market productivity, skill acquisition, and various behaviors. The diversities in these abilities are permanent and crucial to social and economic success [11].
The current debate on digital education technology and students’ cognitive ability mainly focuses on idealized analyses based on the net effect of small-size samples. The linear regression analysis used in these studies can easily cause selection bias in estimating digital education technology. Therefore, this study uses CEM and quantile regression methods. First, it can effectively avoid the selection bias brought by the existing linear regression. Second, it can accurately understand the range or difference in the impact of digital education technology on middle school students’ cognitive ability. Thus, it can help academia better understand students’ cognitive ability development in the digital era.

2. Literature Review

2.1. Digital Technology, Cognitive Ability: The Core of 21st Century Skills

Twenty-first century skills include core skills such as technology application ability and information processing capability. Twenty-first century digital skills include the cognitive ability for information technology applications and critical thinking-based cognitive abilities [12]. Digital competence and critical thinking-based cognitive abilities are included in 21st century skills. The 21st century digital skills are essential skills for teachers and students. Digital technology is at the core of 21st century technology. Studies have shown that 21st century digital skills start with ICT skills, followed by collaboration, critical thinking, and creative digital skills [13]. Digital skills play a role in the process of cognitive development. The heterogeneity exists within digital skills and students’ cognition.
Especially after entering the 21st century, digital technology brought teenagers complex technical skills and learning preferences. These teenagers are vividly called the “digital natives”. Their life and learning cannot do without digital technology, which has become an essential factor affecting their learning and cognitive styles [14,15]. Digital education technology provides students with the coordination of learning content, learning sequence, and learning efficiency, enabling them to customize their experience. It offers personalized adaptive learning opportunities and environments to suit students’ different learning styles and promotes the flexibility of students’ cognitive effects and learning methods. Students acquire knowledge, skills, and attitudes through digital education technology faster than through traditional methods of teacher guidance [16].

2.2. Digital Education Technology

Digital skills appear broad because any member of society can use digital skills. As a result, digital skills began to shift to the field of learning. Digital learning technology (DLT) is defined as any type of learning facilitated by education technology [9]. This concept has a broad meaning and can be applied to all learners. Focusing on teacher-guided in-classroom teaching will generate digital technology within classroom teaching. Digital education technology refers to using digital technology tools to improve teaching and learning processes and students’ development [17], such as computers, the Internet, animation, hypertext, games, virtual reality, and so on [18].
Digital education technology includes both digital technology tools and approaches to apply these tools. With the rapid development of information technology in the past 30 years, applying multimodal computer technology, artificial intelligence, virtual technology, and big data technology to education has changed the current teaching form, which causes teaching technology to upgrade to digital education technology. The difference when comparing digital education technology to traditional teaching technology includes the intellectualization, cyberization, individualization, and self-adaptivity of digital education technology. Digital education technology is primarily characterized by using suitable digital technology instruments and methods by teachers in teaching to facilitate students’ cognitive development and other abilities. From the perspective of learning, digital learning skills, usability and usefulness of digital tools, and Internet-based skills are the core components of digital education technology [19].
No matter the purpose of educational technology research, the core of digital education technology is that digital tools being essential for teaching and learning is a fact that cannot be questioned, even from the learning perspective. Tondeur et al. argue that digital tools, represented by computers, have three main pedagogical applications: operational technology, information tools, and learning tools [20]. Pozo et al. addressed teaching experience, previous ICT use, and other relevant influential factors that affect the usage of digital tools [21]. This digital education technology study focuses on the educational utilization of digital tools, such as computers or the Internet as teaching tools.
Technological innovation theory requires the subject of innovation to have high-level innovative thinking ability. However, technology will not innovate immediately but needs accumulation. Innovation can come out only if the accumulation reaches a certain level. Therefore, before accepting technological innovation, a person first needs to accept technology, so the theory of technology acceptance emphasizes how to apply technology. Digital education technology first focuses on how teachers and students use it to improve their cognitive ability. Innovation is a high-level cognitive ability. Technology inspires the creativity of teachers and students. Creativity reveals how digital education technology allows students to have more opportunities to use technology creatively [22].

2.3. Cognitive Ability

Cognitive ability is the ability of the human brain to extract, process, store, convey, and reproduce information from the objective world. Cognitive ability usually consists of verbal ability, numerical ability, technological ability, logical reasoning skills, and so on. The creative ability to process information is one of its most representative features [23]. Peter et al. believe that elementary and middle school students with high cognitive ability use computers and the Internet to obtain information more frequently than as a tool for entertainment. In contrast, students with lower cognitive abilities are usually on the opposite end of the spectrum [24]. Therefore, digital education technologies represented by computers and the Internet are highly correlated with students’ cognitive abilities and have a heterogeneous influence.

2.4. The Positive Effects of Digital Education Technologies on Students’ Cognitive Abilities

Teaching is and has been in the process of transforming from a traditional paper-mediated instruction model to a digital instructional model. Digital education technology has become both a disruptive and an opportunity-providing catalyst [25]. This catalyst may be perceived as beneficial rather than disadvantageous to the cognitive abilities of middle school students. Digital technology is naturally intuitive and user-friendly [26]. However, whether digital education technology can improve students’ cognitive development lies not only in the technology per se but in the subjective motivation of teachers and students. Technology Acceptance Theory (TAM) emphasizes that important factors influencing user behavior intention are perceived usefulness and perceived ease of use [27]. The usefulness answers the functionality of technology for people, and user-friendliness answers the humanity of technology for people. This completes the transformation from usefulness to user-friendliness. Therefore, only when students consciously perceive the technology, will the realization of cognition be easier, more convenient, and smoother [9]. New digital education technology, such as smartphones and tablets, has increased student engagement in instruction through applications to play a more active role [28]. Because usefulness is what generates value, it is what increases efficiency. Digital technology can help students build a conceptual basis through the topic analysis of interviews [29], thereby eventually enhancing their comprehensive ability. Educational teaching researchers and practitioners are keen on researching and applying new teaching technologies. Some scholars argue that digital education technology changes the quality of teaching because of its ubiquity [30]. Thinking skills are fundamental to cognitive ability and are more challenging to acquire than knowledge. Skills and knowledge can be easily acquired through technology. The students’ thinking ability and innovation abilities determine how information technology is used and the result of realization [31]. A study evaluating cognitive skills in information technology among 15-year-old Chilean students showed that nearly half of the participants could organize, utilize, and manage digital information [32]. Later, researchers focused on the smart teaching technology system and found that the technology is able to adapt the learning algorithm intelligently to the new requirements of the students based on the competency level that the student has already achieved [33].
Hypothesis H1.
Digital education technology positively affects junior high school students’ cognitive abilities.

2.5. Heterogeneity and the Sustainable Development of Students’ Cognitive Abilities

Heterogeneity means inhomogeneity and uneven distribution. Heterogeneity is mainly a biological disciplinary concept and primarily refers to the diversity of essences or species. Since the 1970s, heterogeneity has gradually become an essential concept in social science research, where it mostly indicates the diversity between human individuals or between human groups [34]. Heckman considered that heterogeneity was a different unmeasured or measured variable between individuals and may vary with the same individual as time went by [35]. With the enriching of connotation, heterogeneity has been increasingly extensively applied, and concepts such as political heterogeneity, economic heterogeneity, cultural heterogeneity [36], and network heterogeneity [37] have emerged. Therefore, heterogeneity can be understood as the diversity between individuals or between groups. On the one hand, diversity is a necessary condition for harmony and a necessary basis for a system to develop powerful functions; on the other hand, diversity is a significant characteristic of a complex system.
Heterogeneity in this study indicates the diversity or non-uniformity of junior high school students’ cognitive abilities. The research on heterogeneity reveals the dissimilar effects of digital education technology on different levels of cognitive abilities to provide personalized teaching with a valuable reference, thus promoting the sustainable development of students’ cognitive abilities. Digital technology enables interaction in the process of online culture, which may transform our cognitive function [38]. We should coordinate digital education technology with the sustainable development of students’ cognitive abilities. Meanwhile, we should constantly refine digital education technology to meet the development needs of students’ cognitive abilities at different levels.
Hypothesis H2.
The impact of digital education technology on junior high school students’ cognitive abilities is heterogeneous.
In 1987, the World Commission on Environment and Development (WCED) published “Our Common Future” and put forward the concept of sustainable development. It met the development needs of not only contemporary but also successive generations. Follwing this, with its connotations constantly enriched, sustainable development has been regarded as a behavior vector in a complex system of nature, society, and economy. Sustainable development reveals the systematic essence of “development, coordination, and sustainability ” [39]. Digital education technology is the “hard support” for the sustainable development of junior high school students’ cognitive abilities and sustainable development theory is the “soft support” for dealing with the relationship between “teachers’ teaching using digital technology” and “students’ learning using digital technology.” Only when students’ cognitive abilities match digital teaching can the realization of sustainable development of students’ cognitive abilities possesses technical basis and environmental conditions.
Holling proposed the concept of resilience. Resilient thinking explains the sustainability, unity, and endogenesis of sustainable development. Resilience determines relationships’ persistence in the system. Sustainable development based on resilience emphasizes the need to maintain the openness of choice and heterogeneity [40]. The view of resilience originates from ecology and is an essential way to comprehend the social ecosystem. The resilient method accentuates nonlinear dynamics, thresholds, and uncertainty. Recent research progress includes understanding social processes, such as social networks, adaptability, and changeability [41]. The impact of the social ecosystem’s dynamic interpretation on sustainable development includes shifting the focus from seeking the decisive factors of the optimum state and maximum sustainable output to the resilient analysis [42]. The effects of digital education technology on students’ cognitive abilities are continuous, comprehensive, and endogenous. It demands resilient thinking to examine the technical forces at different levels of cognitive abilities.
Education concerns students’ lifelong development. Lifelong development is a significant manifestation of students’ sustainable development, and the sustainable development of cognitive abilities is the key to students’ comprehensive and sustainable development. In the information age, digital educational technology’s teaching space has expanded everywhere and is not restricted to schools. To realize the sustainable development of junior high school students’ cognitive abilities, we must have digital education technology play an essential role in cognitive ability and attach importance to the heterogeneous relationship between them.
Education modernization represents an efficient and systematic method of sustainable development. Digital education technology is the brightest pearl of education modernization. Therefore, it is the battle position for the sustainable development of students’ cognitive abilities.

2.6. Review

Studies regarding digital education technology have mostly focused on the usage of instructional media and the operation of instructional tools. In contrast, digital education technology indicates the application of teaching media instrumentality and the impact on students’ cognitive ability development. Fewer studies have addressed this area, particularly concentrating on the impact of digital education technologies on the cognitive abilities of middle school students. All areas of teaching and learning and all aspects of teaching and learning are in full swing and in full bloom with digital teaching and learning. The seemingly lively and interesting classroom becomes a training ground for low-quality thinking. Students in the exciting classroom are unconsciously dragged by digital technology into the domain of cognitive homogenization and mechanization of thinking activities. The value of digital education technology is to serve the needs of teachers’ teaching and students’ learning. However, because traditional education technology dominates the current teaching process, problems such as large gaps between the implementation effect and the expectation of digital education technology need to be resolved urgently [43]. Some studies have found that the cognitive benefits of using digital technology are not significant, with only 13.5% improving students’ understanding of knowledge and 7.2% improving the ability to think and ask questions [44]. Because no information technology can play a teaching role, it is restricted by many factors such as students, teachers, environment, and teaching practice. For example, the impact of mobile technology is inconsistent, because for some shy students, there appears to be a beneficial effect of building skills among them; for others, the risk is increased. Excessive digital use may cause damaging of the cognitive ability [6].
In summary, the impact of digital education technology on students’ cognitive abilities is influenced not only by digital technology and digital skills but also by self-efficacy, gender, and teaching expectations. Only by taking these influencing factors into account as comprehensively as possible will the reliability of the study be ensured. Because digital technologies are constantly updating and becoming more intelligent, the impact of digital education technologies on middle school students’ cognitive abilities is constantly changing. Whether digital education technologies improve students’ thinking or delay students’ understanding is a debate on which established studies have not yet reached an agreement. The exact mechanism is unclear. Based on the sustainable development perspective, this study identifies the heterogeneity of digital education technologies affecting cognitive abilities. In terms of research samples, most researchers used questionnaires to obtain small samples, with most sample sizes below 1000. Most of the research methods did not test the existence of heterogeneity. Therefore, this study selects the China Education Panel Survey (CEPS), a comprehensive and representative large-scale tracking survey project, as the research sample. The selection bias was controlled by using methods such as coarsened exact matching to ensure the study’s representativeness and accuracy [45]. Although there are many existing studies related to digital technology, or even digital learning and digital ability, there are relatively smaller numbers of studies focusing on the impact of digital education technology on the cognitive ability of middle school students. Only a few studies have addressed the heterogeneity in the impact of digital education technologies on students’ cognitive abilities.

3. Materials and Methods

3.1. Data Collection

The sample data of this study were obtained from a comprehensive and representative large-scale tracking survey project designed and implemented by the National Survey Research Center (NSRC) of the Renmin University of China. The China Education Panel Survey (CEPS) set the 2013–2014 school year as the baseline and took the cohorts of the first grade (seventh grade) and the third grade (ninth grade) of middle school as the survey starting point and 28 county-level units (county, district, city) as the survey point. A total of about 20,000 students were surveyed in the baseline survey [46]. Two rounds of data from the baseline survey and the second round of follow-up surveys have been published, which were conducted in the 2013–2014 and 2014–2015 academic years. The missing value ratio is less than 10%, and generally, using any missing data treatment method, the difference may not be large [47]. Cases with more missing values have been deleted directly. For cases with few missing values, mean imputation is used to fill out the cases. After processing, a total of 18,978 samples remained at the end.

3.2. Measures

The variables used in this study were selected based on existing research. Digital technology is characterized by using the school’s network to access the Internet and complete learning tasks, such as homework and active learning. Specific digital technology design and factors influencing student attitudes and beliefs can affect learning implementation and success [30]. It is necessary choose whether to use the Internet and personal website as the core explanatory variables of this study. In the corresponding questionnaire, the indicator for measuring digital education technology is: “Did you use the following teaching media when teaching the class under investigation?”, including three items: 1. Internet; 2. Wall charts, models, and posters; and 3. Personal teaching websites or blogs, microblogging teaching. The answer is “never, occasionally, sometimes, often, always,” represented by 0–4, respectively. Since the second item uses wall charts, models, and posters as teaching media, which do not belong to digital education technology, the average value of the first and third items is used as the index of digital education technology.
The cognitive ability of middle school students is the explanatory variable. By adopting the research method of Fang et al., using the cognitive index provided by CEPS 2013–2014, the cognitive ability test consists of test questions involving 11 concepts in three dimensions: verbal, graphical, and computational and logical [48]. Cognitive skills measure the students’ abilities in terms of logical thinking and problem-solving skills. These questions cover three areas: verbal, graphical, and calculative and logical. Cognitive ability tests involving CEPS are nationally standardized [49].
To ensure the reliability and validity of digital education technology‘s effect on middle school students’ cognitive ability, the control variables in this study were selected based on the basic characteristics of middle school students and the characteristics of their families, classes, and schools. 1. Grade level, gender, self-efficacy, and academic level characterize the personality traits of middle school students and are factors closely related to cognitive ability. 2. Family characteristics include the level of family education in terms of parents’ education, while the Internet and computer accessibility are used to characterize the family information and cultural environment. 3. Class characteristics consisted mainly of teachers’ attitudes toward salary and the school management styles. At the same time, class capacity is included in the class characteristics variables. 4. School characteristics are mainly characterized by educational informatization, i.e., whether the school achieves the construction of Three Supplies and Two Platforms. For details, see Table 1.

3.3. Statistical Description of Variables

According to Table 2, the average value of junior high school students’ cognitive abilities is 9.9788. Overall, the cognitive scores are relatively low, with a standard deviation of 3.7444, indicating that there are certain diversities in the cognitive scores of junior high school students. It provides a data basis for the research on heterogeneity. The average value of digital education technology is 0.4123. On the whole, nearly half (41.23%) of junior high school students accept learning with digital educational technology, but this proportion is relatively low. It may be related to the following three factors: the informatization construction of the school, teachers’ application ability of computers and Internet networks, and students’ characteristics. The average value of schools’ educational informatization is 0.8499, which means the rate of schools’ educational informatization has reached 85%. This is in deep contrast with the 41.23% of students who accepted digital education technology, indicating that the applications of computers and Internet networks, accompanied by the characteristics of teachers and students, have enormous impacts on digital education technology. For details, see Table 2.

3.4. Data Analyses

First, ordinary least squares (OLS) regression was used to describe the relationship between the acquisition of digital education technology and the cognitive ability of middle school students. Second, Coarsened Exact Matching was used to avoid the selective estimation bias brought by OLS and analyze the heterogeneity of the different distributions of both experimental and control groups.

3.4.1. OLS Model Setting

The predicted variable is the cognitive ability of middle school students. The core explanatory variable is whether or not to accept digital education technology. The control variables consist of individual student characteristics, classroom characteristics, school characteristics, and parent characteristics, as well as factors that may affect the cognitive ability of middle school students. The following model was set to test whether digital education technology has any impact on the cognitive ability development of middle school students. The OLS model can be seen in Equation (1):
Cogi = β0 + β1Digteci + β2Xi + ɛi
In Equation (1), the subscript i represents middle school students, and the predicted variable, Cogi, is the cognitive ability of student i. The core explanatory variable, Digteci, is whether i students are taught with digital education technology, Xi is the influencing factor of cognitive development of middle school students, and ɛi is the residual term in the model.

3.4.2. Coarsened Exact Matching

Although Equation (1) can describe the relationship between the explanatory variables and the predicted variables, ordinary least squares estimation may lead to selection bias in the estimation. The CEM method can reduce the estimation bias brought by OLS and precisely match the individual selection results of the experimental and control groups. Therefore, the CEM method can more reliably explain the heterogeneity between digital education technology and middle school students’ cognitive ability. In addition, it can effectively exclude control variables’ interference in the heterogeneity. The CEM weights brought into the weighted linear regression model can control the selection bias and reduce the imbalance of characteristic variables among groups [45]. The CEM model of these constructs can be seen in Equation (2):
Cogi = β0 + β1Digteci[cem_weights] + β2Xi + ɛi
The predicted variable, Cogi, is the cognitive ability of student i, β0 is the intercept. Digteci is whether student i is taught with digital education technology, and Xi is the score of the control variable that has an effect on cognitive ability. β1 and β2 are the regression coefficients of the corresponding variables, ɛi is the residual term in the model, and cem_weights are the weight.

3.4.3. Quantile Regression and Heterogeneity

Proposed by Koenker and Bassett, quantile regression regresses different distributions or diversities of dependent variables. It is more robust than mean regression. Quantile regression can reveal the heterogeneity of the influence on explained variables in their different intervals from explanatory variables [50]. In contrast with traditional linear regression, quantile regression is more effective and reliable. The ordinary least square method can only explain impacts based on the average value instead of the different levels of relevant data, while quantile regression can make up for this deficiency [51].
On the individual level, quantile regression is applied to explore whether the influence of digital education technology on junior high school students’ cognitive abilities varies with the distributions of cognitive abilities. Compared with OLS regression, quantile regression can reveal the effect of explanatory variables on explained variables at different levels more intuitively and effectively. Therefore, this study selected nine quantiles of 10–90% and estimated, respectively, to investigate the effect of digital education technology on the cognitive abilities of junior high school students at different levels [52]. Based on this, this paper studies the heterogeneous impact of digital education technology on cognitive abilities and provides the foundation for teaching students in accordance with their aptitude.
The formula of the quantile regression model is:
Quantileτ(Cog|Digtec) = β0τ + β1τDigteci + β2τXi + ɛiτ
In Equation (3), τ represents the quantile; β0τ, β1τ, and β2τ represent the parameter of the τ-th quantile, respectively; ɛiτ represents the random error of different quantiles.

4. Results

4.1. Baseline Results

First, the variables were tested, and none of them had collinearity and heteroskedasticity problems. To ensure the accuracy and scientificity of OLS estimation results, we carried out the collinearity and heteroscedasticity tests. The results of VIF being less than 10 indicate no collinearity between variables (Table 3); the results of the Breusch–Pagan test (p = 0.2446) indicate that no heteroscedasticity exists in this study. At the same time, it can be seen from the residual distribution scatter diagram that the residual is evenly distributed and there is no heteroscedasticity in the model.
Second, the OLS method was used to estimate the impact of digital education technology on middle school students’ cognitive abilities. From the estimated parameter, digital education technology was 0.261, and it was highly significant, indicating that digital education technology positively affects middle school students’ cognitive abilities. Therefore, hypothesis H1 proves correct: after controlling the factors of students, families, teachers, and schools, digital education technology has a significant positive effect on cognitive ability. For details, see Table 3.

4.2. Main Results from Coarsened Exact Matching Method

The OLS regression results have shown that digital education technologies significantly impact middle school students’ cognitive abilities. However, such estimates tend to have large errors and sample selection bias. The CEM method can improve the accuracy of the result assessment and effectively reduce the sample selection bias. In the Multivariate Imbalance Measure (L1 = 0.536 before matching and L1 = 0.016 after matching), the L1 value is significantly reduced, indicating that the CEM method is very effective, and further indicating that the OLS estimation results have a significant bias, as shown in Table 4. The total sample was 18,978, the matched successful sample was 12,132, and the unmatched successful sample was 6846, with a matching success rate of 63.9%, indicating a good matching effect. The primary goal of CEM matching is to eliminate the observation value from the data so that the observation value has a better balance between the treatment group and the control group [53]. Students who received digital education technology were the treatment group, and those who did not receive digital education technology were the control group. CEM can balance the treatment and control group, remove redundant invalid observations, and predict the cognitive abilities of junior high school students more accurately than OLS regression. Therefore, after CEM matching, observation values are less than OLS regression. Because no matching method helps to generate high-quality partial balance, CEM tends to generate a reasonable number of matches in practice, which is regarded as an advantage in practical applications. CEM has a lower root mean square error than propensity score, Markov distance, nearest neighbor, and optimal matching [54]. Therefore, using the matched samples, the impact of digital education technology on the cognitive ability of middle school students is estimated again based on the weights of CEM, which can be estimated accurately [45].
The CEM weighted regression results (see Table 5) showed that digital education technology still positively influenced middle school students’ cognitive ability, and the matched digital education technology regression coefficient (0.239) was slightly less than the OLS regression coefficient (0.261), indicating that the OLS method overestimated the influence of the core predictor variable. Computer accessibility changed from significantly influencing students’ cognitive abilities to no longer significantly influencing them. In contrast, the impact of Internet accessibility on middle school students’ cognitive ability increased significantly. For each unit increase in Internet use score, the cognitive ability score would increase by 1.225 points. This shows that computers have been integrated into the Internet and become a tool for using the Internet, mainly to obtain information from the Internet, exchange information, and use the information to open their minds and enrich themselves. In addition, among the four indicators reflecting the personality characteristics of middle school students, grade and gender negatively predicted middle school students’ cognitive abilities, while self-efficacy and academic performance positively predicted middle school students’ cognitive ability. The family characteristics variables of parents’ education and family information culture environment (Internet and computer accessibility) both positively predicted middle school students’ cognitive ability. The class characteristics variables of teachers’ salary attitude and class capacity all negatively predicted middle school students’ cognitive ability. The school characteristics, such as educational informatization, also correctly predicted middle school students’ cognitive ability. It can be seen that the CEM method can reflect the net effect of impact more accurately than the OLS method.

4.3. Quantile Regression

Although it can be concluded that digital education technology can positively predict middle school students’ cognitive ability through CEM regression analysis, it is still based on the analysis of the overall matched sample. Whether or not the same effect exists in different cognitive ability groups is unknown. Therefore, quantile regression was used to further understand the effect of digital education technology on middle school students’ cognitive ability to test the heterogeneity of the cognitive ability. The results (see Table 6) of the study showed that the effect of digital education technology on middle school students’ cognitive ability increased as the quantile rose from insignificant (q10 = 0.114) to more significant (q20 = 0.179, q30 = 0.152) to very significant (q80 = 0.390, q90 = 0.500). In other words, the higher the cognitive ability score, the more significant the effect of digital education technology would be, and there was significant heterogeneity. Therefore, hypothesis H2 is established. The positive effect of digital education technology is stronger on students with higher cognitive abilities.
The variation diagram of the quantile coefficients depicts the variation trend of digital education technology and other coefficients at different quantiles (see Figure 1). The dotted line represents the OLS regression estimate value of the explanatory variable and the solid line represents the quantile regression estimate value of the explanatory variable. It can be seen that, generally, the effects of digital education technology on cognitive abilities keep increasing from the low quantile (10%) to the high quantile (90%). The effects are not remarkable at the 10% level, but after the 70% quantile, the effects of digital education technology increase significantly, far higher than the estimated value of OLS regression. This indicates that the effects of digital education technology on junior high school students’ cognitive abilities are heterogeneous, which provides vital enlightenment to promote the sustainable development of junior high school students’ cognitive abilities. Sustainable development places a particular emphasis on the characteristics of integrity, endogenesis, and comprehensiveness [39]. For individual students, the sustainable development of cognitive abilities is a continuous development process from low to high. The heterogeneous results of quantile regression imply that the effects of digital education technology on the sustainable development of individual students’ cognitive abilities have the characteristics of vertical integrity and endogeneity; for student groups, the sustainable development of cognitive abilities represents that all students can gain development. The heterogeneous results of quantile regression indicate that the influence of digital education technology on the sustainable development of students’ individual cognitive abilities is characterized by horizontal integrity and comprehensiveness.

5. Discussion

In summary, digital education technologies are primarily designed to improve the teaching and learning processes and promote the sustainable development of students’ cognitive abilities [17]. Either a large number of educational theorists or pedagogical practitioners have made digital education technology a trending topic. Plus, the impact of digital instructional technologies on students’ cognitive development is debated in different ways.

5.1. The Impact of Digital Education Technology on Students’ Cognitive Ability

Although digital education technology is good, it needs to be adapted to the students’ needs. Based on the CEM methods’ results, students’ score with digital education technology in classroom instruction is 0.239 higher than the cognitive ability score of students without digital education technology. Silveira believes that digital educational technologies improve the use of active learning methods and that adapting them is a flexible way to develop critical thinking [18]. Digital technology is considered to be the driving force behind pedagogical innovation and the most direct influence in empowering students’ cognitive sustainable development [55]. Digital education technology in the digital environment has the essential characteristics of pedagogical intelligence.

5.2. Heterogeneity

According to the results of the quantile regression, it was shown that there was heterogeneity in the effect of digital education technology on the cognitive ability of middle school students. This means there were differences in the impact among different groups. The impact of cognitive ability of middle school students in the 10th quartile was not significant. Those in the 20th and 30th quartiles were significant at the p < 0.1 level, and the latter 50% were significant at the p < 0.01 level, showing an overall increasing trend. It can be seen that the higher the cognitive ability score, the greater the impact of digital education technology (see Figure 1). This is consistent with the findings of Zheng et al. Home Internet accessibility significantly predicted students’ cognitive abilities. This suggests that access to and use of information technology can improve learning outcomes. Guiding students to use digital technology correctly is the key to the “digital benefits” of education [56]. However, an area where this finding needs improvement is the direct hypothesis of empowering students’ cognitive abilities, as this ignores the dynamic, constructive role of digital education technologies. One way to examine this question is to look at how teachers view the potential of digital education technology for learning. For example, we found that students’ cognitive abilities guided teachers to evaluate the of using digital education technologies differently based on how they categorized students’ characteristics in their groups [57]. Cao et al. also confirmed that Internet use affects the cognitive ability gap between urban and rural adolescents. In other words, urban adolescents have significantly higher Internet ownership and return rates than rural adolescents, which, in turn, will widen the cognitive ability gap between urban and rural adolescents [58]. However, existing studies do not provide different cognitive abilities influenced by digital education technologies. This paper further points out that there is heterogeneity in the influence of digital education technologies on middle school students’ cognitive abilities via an exact-matching quantile regression analysis method, which is an important instruction based on their needs.
This study found heterogeneity in the impact of different digital education technologies. It is noteworthy that the marginal effect value of the decile was 0.1, and there was no significant effect (p > 0.1). The ninetieth decile was 0.5, and the effect was very significant (p < 0.01), indicating a 0.4 difference between middle school students with low cognitive ability scores and those with high scores among students who received digital education technology. Thus, it is necessary to be aware of the ineffectiveness of digital education technology in teaching and avoiding the “enrichment effect” are both important because students with high cognitive ability scores benefit more from digital education technology. They have good information literacy, comprehensive abilities, and the knowledge base to use digital education technology to solve problems in learning. This will increase interest in learning and develop various thinking skills. At the same time, students with low cognitive ability scores are susceptible to interference or negative effects of digital education technology. Although they are curious and interested in digital education technology, it is easy for them to convert the originally good and expected learning technology into a gaming tool for their own entertainment because they are not skilled in operating it or because they cannot understand the learning resources provided by digital education technology. Therefore, teachers using digital education technology should be aware of students, understand students, facilitate students, and implement individualized teaching according to students’ needs to fully utilize digital education technology’s efficacy [45]. The sustainable development of cognitive ability supported by digital technology endows students with stronger adaptability for future society, thus reflecting the sustainability and integrity of students’ cognitive abilities development. Research has confirmed that improving cognitive skills may increase students’ various benefits [59]. The sustainable development of cognitive ability is also closely related to personal happiness, family, marriage, and interpersonal relationships. General cognitive ability is proven highly effective in predicting merit ratings for all jobs [23].
The heterogeneous impact of digital education technology on students’ cognitive abilities is embodied in different digital education technologies promoting the development of cognitive abilities at different levels. Clements et al. revealed the effect of programming technology on children’s cognitive style, meta-cognitive ability, and cognitive development. Programming spurred divergent thinking in the graphical environment. As a result, children in the programming group improved their creative ability. In contrast, the CAI group did not show any improvement because CAI usually paid attention to convergent thinking rather than divergent thinking [60]. This is because of the diverse expressiveness and pluralistic types of digital education technology that can meet the needs of different levels of cognitive abilities and multiple cognitive sustainable developments. Through experiments, Yang confirmed that digital game-based learning (DGBL) had greatly improved the ability of creative thinking, critical thinking, and problem-solving [61]. In mathematical teaching, digital education technology contains dynamic geometry tools, computer-based laboratory equipment, graphic calculators, etc. Dynamic geometry tools provide students with a computer world to express and study with mathematical thinking; graphic calculators enable them to access various geometric figures easily [62]. Different digital education technologies promote students’ corresponding cognitive abilities in different ways. When the student’s cognitive ability is at the pre-operation stage or relatively low-level, combining multi-modal technology and intuitive education technology can realize the leapfrog development of image thinking and promote mental co-evolution. When the student’s cognitive ability is at high-level, mainly abstract thinking, the combination of computer-aided education and artificial intelligence technology can realize the sustainable development of creative thinking. Therefore, the heterogeneity of cognitive ability requires applying reasonable and appropriate digital education technology to realize the sustainable development of students’ cognitive abilities.
There was also heterogeneity in personality traits by grade level, gender, learning performance, and self-efficacy. There were two grades in the sample, seventh and ninth. The regression coefficient is negative, indicating that because ninth-grade students are under relatively high academic pressure, using digital education technology can squeeze knowledge learning if it is too frequent. Overall, both seventh and ninth grades benefit from digital education technologies. Family characteristics, such as parental education level and whether the family owns a computer and the Internet, affect students’ cognitive abilities. The sample after CEM method, whether the family owns a computer, becomes insignificant, indicating that the Internet’s vast resources are what teachers and students utilize. This also means smartphones are miniaturized computers and middle school students spend relatively little time utilizing them. It also suggests that selective bias overestimates the impact of computers on students’ cognitive abilities. Because these suggested findings are based on a matched sample, the impact of potential selection bias is reduced. Thus, the effect of digital education technology on student cognition can be considered a net positive effect. This can be explained by digital education technologies improving students’ comprehensive ability and thinking skills [63].

5.3. Limitations

Although the selection bias has been eliminated to a certain extent by using the CEM method, there are still some limitations as follows: first, digital education technology has various characteristics such as diversification, intelligence, humanization, and resourcefulness, and further research is needed in this regard; second, digital education technology includes methods of using digital education technology in addition to tools and equipment, and further improvement is needed in this regard; last, digital education technology and middle school students’ cognitive development are complex multi-factor systems composed of multiple elements, and new models need to be constructed in future studies to grasp the interactivity and the dynamic mechanism between digital education technology and middle school students’ cognitive ability in a more comprehensive, dynamic, and sustainable way.

6. Conclusions and Implications

6.1. Conclusions

This study actively responded to this debate, and the results of this study showed that under the case of controlling relevant variables, digital education technology has a positive effect on students’ cognitive ability and there is heterogeneity. Digital education technologies significantly enhance students’ cognitive abilities. The effective use of digital devices such as school information technology teaching platforms, multimedia classrooms, online classrooms, and smart classrooms can improve students’ cognitive abilities. It was shown that there was heterogeneity in the effect of digital education technology on the cognitive ability of middle school students. It can be seen that the higher the cognitive ability score, the greater the impact of digital education technology. Consistent with recent research, the continuous application of technological equipment has developed various changes in human cognitive abilities, enabling people to reflect on the optimal application of technological means, rather than uncritically use them [64]. Digital education technology promotes the accessibility of junior high school students’ cognitive development. However, it still needs to pay attention to digital education technology’s ineffective application, even negative impact, which is a primary problem to be solved in this study. The heterogeneous impact of digital educational technology on cognitive abilities at different levels has been confirmed through quantile regression analysis. On the one hand, digital education technology has ineffective or negative impacts on junior high school students with low-level cognitive abilities, specifically related to their willpower, information literacy, etc. On the other hand, digital educational technology has positive impacts on junior high school students with high-level cognitive abilities. There are three major viewpoints in the existing research. First, digital education can deepen students’ cognitive understanding of the learning process and cultivate and promote students’ cognitive abilities and innovation abilities [65]. Second, the continuous multitasking of digital devices will damage the cognitive abilities of teenagers [6]. Third, digital education technology is a tool whose effect depends on the operation and application of teachers and students. These three views are reflected in the opposition between the two theories of “technology only” and “technology is useless” of digital education technology. This study answered this question well through heterogeneity research. Heterogeneity is a significant embodiment of the unity, comprehensiveness, and endogeneity of sustainable development. The resilient thinking of sustainable development emphasizes heterogeneity and is used to interpret the sustainability of development, the integrity of factors, and the persistence of relationships. The impacts of digital education technology on cognitive abilities at different levels can be positive, negative, and uncertain. This study analyzed and resolved these inconsistent results with heterogeneity analysis. Therefore, this paper attempts to solve the contradiction between the theories of “technology only” and “technology is useless” of digital education technology and provide a solution for the debate on the impact of digital education technology on students’ development in existing research. We should provide students of different cognitive abilities with different guidance, development, and applications of digital education technology. The development and application of digital education technology focus on the entire ecosystem revolved around computers, Internet networks, and smartphones, including applications and educational content, games, and social networks. At the same time, it provides an immersive multi-modal environment, designs resources for multi-sensory learning, enhances learning pleasure, and reduces cognitive load [66].
Heterogeneous research method pulls together the above three contradictory viewpoints well and provides constructive suggestions for effectively promoting the sustainable development of cognitive ability. Heterogeneous thinking not only respects individuals’ diverse development but also considers the collaborative development of groups. It reflects the inherent requirement of sustainable development. Therefore, the diversity and flexibility of digital education technology satisfy the various development possibilities of students’ cognitive abilities and cultivate a large number of high-quality talents with innovative abilities for society.

6.2. Implications

The CEM method controls selection bias and reduces errors, verifying the validity of this paper’s findings. This study, on the one hand, confirms the significant impact of digital education technology on middle school students’ cognitive abilities; on the other hand, it reveals that this impact is significantly heterogeneous and provides important indications for tailoring teaching to students’ needs. We suggest that teachers should pay attention to how different types of digital technologies work in different ways for teaching and learning and implement categorized instruction based on different standards.

Author Contributions

Methodology, D.Y. and G.L.; Software, D.Y.; Validation, G.L.; Data curation, D.Y.; Writing—original draft, D.Y.; Supervision, G.L. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported in part by Grant No. 2021A1515012516 from the Natural Science Foundation of Guangdong Province.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The data are not publicly available due to privacy restrictions.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Zheng, Y.; Wang Yang, C.; Yan, X.; Zhang, Z.; Zheng, Y. The development of intelligent education equipment in China: Status quo and directions for future R&D. Chin. J. Distance Educ. 2020, 42, 11–19+76. [Google Scholar] [CrossRef]
  2. Alakrash, H.M.; Razak, N.A. Technology-Based Language Learning: Investigation of Digital Technology and Digital Literacy. Sustainability 2021, 13, 12304. [Google Scholar] [CrossRef]
  3. Zhang, R.Y. Digital Media Teaching and Effectiveness Evaluation Integrating Big Data and Artificial Intelligence. Comput. Intell. Neurosci. 2022, 2022, 1217846. [Google Scholar] [CrossRef]
  4. Ramlee, N.; Rosli, M.S.; Saleh, N.S. Mathematical HOTS Cultivation via Online Learning Environment and 5E Inquiry Model: Cognitive Impact and the Learning Activities. Int. J. Emerg. Technol. Learn. 2019, 14, 140–151. [Google Scholar] [CrossRef]
  5. Xu, Z.; Jang, E.E. The role of math self-efficacy in the structural model of extracurricular technology-related activities and junior elementary school students’ mathematics ability. Comput. Hum. Behav. 2017, 68, 547–555. [Google Scholar] [CrossRef]
  6. George, M.J.; Odgers, C.L. Seven Fears and the Science of How Mobile Technologies May Be Influencing Adolescents in the Digital Age. Perspect. Psychol. Sci. 2015, 10, 832–851. [Google Scholar] [CrossRef] [PubMed]
  7. Tarus, J.K.; Gichoya, D.; Muumbo, A. Challenges of implementing e-learning in Kenya: A case of Kenyan public universities. Int. Rev. Res. Open Distrib. Learn. 2015, 16, 120–141. [Google Scholar] [CrossRef]
  8. Zhang, W.; Chen, M.; Zhao, X.; Bai, X. Influence of Teacher-student Interaction on Classroom Learning in Special Delivery Classroom: A Case of Art Delivery Class in Chongyang Primary School. E-Educ. Res. 2020, 41, 90–96. [Google Scholar] [CrossRef]
  9. Kashada, A.; Li, H.G.; Koshadah, O. Analysis Approach to Identify Factors Influence Digital Learning Technology Adoption and Utilization in Developing Countries. Int. J. Emerg. Technol. Learn. 2018, 13, 48–59. [Google Scholar] [CrossRef]
  10. Hwang, G.J.; Fu, Q.K. Advancement and research trends of smart learning environments in the mobile era. Int. J. Mob. Learn. Organ. 2020, 14, 114–129. [Google Scholar] [CrossRef]
  11. Heckman, J.J.; Stixrud, J.; Urzua, S. The Effects of Cognitive and Noncognitive Abilities on Labor Market Outcomes and Social Behavior. J. Labor Econ. 2006, 24, 411–482. [Google Scholar] [CrossRef]
  12. Van Laar, E.; van Deursen, A.J.A.M.; van Dijk, J.A.G.M.; de Haan, J. The relation between 21st-century skills and digital skills: A systematic literature review. Comput. Hum. Behav. 2017, 72, 577–588. [Google Scholar] [CrossRef]
  13. Van Laar, E.; Van Deursen, A.J.A.M.; Van Dijk, J.A.G.M.; de Haan, J. The Sequential and Conditional Nature of 21st-Century Digital Skills. Int. J. Commun. 2019, 13, 3462–3487. [Google Scholar]
  14. Wang, Q.; Myers, M.D.; Sundaram, D. Digital Natives und Digital Immigrants. Wirtschaftsinf. 2013, 55, 409–420. [Google Scholar] [CrossRef]
  15. Bennett, S.; Maton, K.; Kervin, L. The “digital natives” debate: A critical review of the evidence. Br. J. Educ. Technol. 2008, 39, 775–786. [Google Scholar] [CrossRef]
  16. Ruiz, J.G.; Mintzer, M.J.; Leipzig, R.M. The Impact of E-Learning in Medical Education. Acad. Med. 2006, 81, 207–212. [Google Scholar] [CrossRef]
  17. Damascena, S.C.C.; Santos, K.C.B.; Lopes, G.S.G.; Gontijo, P.V.C.; Paiva, M.V.S.; Lima, M.E.S.; Alves, J.M.F.; Campos, R.S. Use of Digital Educational Technologies as a Teaching Tool in the Nursing Teaching Process. Braz. J. Dev. 2019, 5, 29925–29939. [Google Scholar] [CrossRef]
  18. Silveira, M.S.; Cogo, A.L.P. The contributions of digital technologies in the teaching of nursing skills: An integrative review. Rev. Gauch. Enferm. 2017, 38, e66204. [Google Scholar] [CrossRef]
  19. Sayaf, A.M.; Alamri, M.M.; Alqahtani, M.A.; Alrahmi, W.M. Factors Influencing University Students’ Adoption of Digital Learning Technology in Teaching and Learning. Sustainability 2022, 14, 493. [Google Scholar] [CrossRef]
  20. Tondeur, J.; Hermans, R.; van Braak, J.; Valcke, M. Exploring the link between teachers’ educational belief profiles and different types of computer use in the classroom. Comput. Hum. Behav. 2008, 24, 2541–2553. [Google Scholar] [CrossRef]
  21. Pozo, J.I.; Pérez Echeverria, M.P.; Cabellos, B.; Sanchez, D.L. Teaching and Learning in Times of COVID-19: Uses of Digital Technologies during School Lockdowns. Front. Psychol. 2021, 12, 1–13. [Google Scholar] [CrossRef]
  22. Meng, C. How Does Technology Integration Theory Guide Teaching Innovation? A Reflection on the HPC Theory`s Support of “High Possibility” Teaching. J. Nanjing Norm. Univ. 2021, 1, 120–129. [Google Scholar]
  23. Hunter, J.E. Cognitive ability, cognitive aptitudes, job knowledge, and job performance. J. Vocat. Behav. 1986, 29, 340–362. [Google Scholar] [CrossRef]
  24. Peter, J.; Valkenburg, P.M. Adolescents’ internet use: Testing the “disappearing digital divide” versus the “emerging digital differentiation” approach. Poetics 2006, 34, 293–305. [Google Scholar] [CrossRef]
  25. Hashemi, S.S.; Cederlund, K. Making room for the transformation of literacy instruction in the digital classroom. J. Early Child. Lit. 2017, 17, 221–253. [Google Scholar] [CrossRef]
  26. Loureiro, F.; Sousa, L.; Antunes, V. Use of Digital Educational Technologies among Nursing Students and Teachers: An Exploratory Study. J. Pers. Med. 2021, 11, 1010. [Google Scholar] [CrossRef]
  27. Zhu, G.; Li, F.; Shen, Y.; Bian, S. Research on Reading Willingness of Privacy Policy in Social Media Context—Based on the Perspective of TAM Model and Self-efficacy Theory. J. Mod. Inf. 2022, 42, 150–166. [Google Scholar]
  28. Martin, S.; Diaz, G.; Sancristobal, E.; Gil, R.; Castro, M.; Peire, J. New technology trends in education: Seven years of forecasts and convergence. Comput. Educ. 2011, 57, 1893–1906. [Google Scholar] [CrossRef]
  29. Brink, H.; Kilbrink, N.; Gericke, N. Teaching digital models: Secondary technology teachers’experiences. Int. J. Technol. Des. Educ. 2022, 32, 1755–1775. [Google Scholar] [CrossRef]
  30. Penuel, W.R. Implementation and effects of One-to-One Computing Initiatives. J. Res. Technol. Educ. 2006, 38, 329–348. [Google Scholar] [CrossRef]
  31. Wrahatnolo, T.; Munoto. 21st centuries skill implication on educational system. IOP Conf. Ser. Mater. Sci. Eng. 2018, 296, 012036. [Google Scholar] [CrossRef]
  32. Claro, M.; Preiss, D.D.; San Martin, E.; Jara, I.; Hinostroza, J.E.; Valenzuela, S.; Cortes, F.; Nussbaum, M. Assessment of 21st century ICT skills in Chile: Test design and results from high school level students. Comput. Educ. 2012, 59, 1042–1053. [Google Scholar] [CrossRef]
  33. Ebner, M.; Edtstadler, K.; Ebner, M. Tutoring writing spelling skills within a web-based platform for children. Univ. Access Inf. Soc. 2018, 17, 305–323. [Google Scholar] [CrossRef] [Green Version]
  34. Gong, L.; Jiang, Y. Heterogeneity and Organizational Performance: A Research Approach. J. Guangxi Univ. (Philos. Soc. Sci.) 2020, 42, 137–144. [Google Scholar] [CrossRef]
  35. Heckman, J.J.; Borjas, G.J. Does Unemployment Cause Future Unemployment? Definitions, Questions and Answers from a Continuous Time Model of Heterogeneity and State Dependence. Economica 1980, 47, 247–283. [Google Scholar] [CrossRef]
  36. Vedeld, T. Village politics: Heterogeneity, leadership and collective action. J. Dev. Stud. 2000, 36, 105–134. [Google Scholar] [CrossRef]
  37. Reagans, R.; McEvily, B. Network structure and knowledge transfer: The effects of cohesion and range. Adm. Sci. Q. 2003, 48, 240–267. [Google Scholar] [CrossRef]
  38. Bruni, L.E. Sustainability, cognitive technologies and the digital semiosphere. Int. J. Cult. Stud. 2015, 18, 103–117. [Google Scholar] [CrossRef]
  39. Niu, W.Y. The Theoretical Connotation of Sustainable Development: The 20th Anniversary of UN Conference on Environment and Development in Rio de Janeiro, Brazil. China Popul. Resour. Environ. 2012, 22, 9–14. [Google Scholar] [CrossRef]
  40. Holling, C.S. Resilience and Stability of Ecological Systems. Annu. Rev. Ecol. Syst. 1973, 4, 1–23. [Google Scholar] [CrossRef]
  41. Folke, C. Resilience: The emergence of a perspective for social–ecological systems analyses. Glob. Environ. Chang. 2006, 16, 253–267. [Google Scholar] [CrossRef]
  42. Walker, B.; Holling, C.S.; Carpenter, S.R.; Kinzig, A. Resilience, Adaptability and Transformability in Social-ecological Systems. Ecol. Soc. 2004, 9, 5. [Google Scholar] [CrossRef]
  43. Shi, W.; Li, H. Reflection and Improvement of Teachers’ Teaching Technology Literacy at Digital Age. Curric. Teach. Mater. Method 2017, 37, 95–100. [Google Scholar] [CrossRef]
  44. Lai, K.-W.; Pratt, K. Positive to a Degree: The Effects of ICT Use in New Zealand Secondary Schools. Comput. Sch. 2007, 24, 95–109. [Google Scholar] [CrossRef]
  45. Li, W.; Yao, J. Effects of E-reading on Reading Literacy. Open Educ. Res. 2021, 27, 110–120. [Google Scholar] [CrossRef]
  46. China Education Panel Survey. Available online: http://ceps.ruc.edu.cn/xmjs/xmgk.htm (accessed on 7 December 2022).
  47. Liu, H.Y. Advanced Statistics for Psychology; China Renmin University Press: Beijing, China, 2019; p. 7. [Google Scholar]
  48. Fang, C.; Wang, G.; Huang, B. Can Information Technology Promote the Development of Students’ Cognitive Ability? Net Effect Estimation Based on CEPS. Open Educ. Res. 2019, 25, 100–110. [Google Scholar] [CrossRef]
  49. Zhou, L.; Jiang, B.; Wang, J.X. Do cash transfers have impacts on student Academic, cognitive, and enrollment outcomes? Evidence from rural China. Child. Youth Serv. Rev. 2020, 116, 105158. [Google Scholar] [CrossRef]
  50. Koenker, R.; Bassett, G. Regression quantiles. Econometrica 1978, 46, 33–50. [Google Scholar] [CrossRef]
  51. Wei, F.; Feng, N.; Zhang, K. Innovation Capability and Innovation Talents: Evidence from China Based on a Quantile Regression Approach. Sustainability 2017, 9, 1218. [Google Scholar] [CrossRef]
  52. Zheng, L.; Sun, Y. Where there is a Will, there is a Way: Heterogeneous Effects of Students’ Willpower on Cognitive Abilities. J. East China Norm. Univ. (Educ. Sci.) 2022, 40, 83–95. [Google Scholar]
  53. Blackwell, M.; Iacus, S.; King, G.; Porro, G. Cem: Coarsened Exact Matching in Stata. Stata J. 2009, 9, 524–546. [Google Scholar] [CrossRef]
  54. Iacus, S.M.; King, G.; Porro, G. Multivariate Matching Methods That Are Monotonic Imbalance Bounding. J. Am. Stat. Assoc. 2011, 106, 345–361. [Google Scholar] [CrossRef]
  55. Lin, R.; Yang, J.; Jiang, F.; Li, J. Does teacher’s data literacy and digital teaching competence influence empowering students in the classroom? Evidence from China. Educ. Inf. Technol. 2022. [Google Scholar] [CrossRef]
  56. Zheng, L.; Qi, X.; Zhu, Z.; Zhang, D. Home Internet Access and Urban-Rural Cognition Gap of Middle School Students. Res. Educ. Dev. 2021, 41, 10–18. [Google Scholar] [CrossRef]
  57. Rafalow, M.H.; Puckett, C. Sorting Machines: Digital Technology and Categorical Inequality in Education. Educ. Res. 2022, 51, 274–278. [Google Scholar] [CrossRef]
  58. Cao, D.; Luo, S.; Yang, X.; Wang, W. Effect of Internet on the Development of Cognitive Ability of Urban and Rural Teenagers. China Educ. Technol. 2018, 11, 9–17. [Google Scholar] [CrossRef]
  59. Bowles, S.; Gintis, H.; Osborne, M. The Determinants of Earnings: A Behavioral Approach. J. Econ. Lit. 2001, 39, 1137–1176. [Google Scholar] [CrossRef]
  60. Clements, D.H.; Gullo, D.F. Effects of computer programming on young children’s cognition. J. Educ. Psychol. 1984, 76, 1051–1058. [Google Scholar] [CrossRef]
  61. Yang, Y.T.C. Virtual CEOs: A blended approach to digital gaming for enhancing higher order thinking and academic achievement among vocational high school students. Comput. Educ. 2015, 81, 281–295. [Google Scholar] [CrossRef]
  62. Heid, M.K. The Technological Revolution and the Reform of School Mathematics. Am. J. Educ. 1997, 106, 5–61. [Google Scholar] [CrossRef]
  63. Fang, G.; Chan, P.W.K.; Kalogeropoulos, P. The Effects of School-to-School Collaboration on Student Cognitive Skills: Evidence from Propensity Score Analysis. Asia-Pac. Educ. Res. 2022, 31, 193–203. [Google Scholar] [CrossRef]
  64. Pattier, D.; Reyero, D. Contributions from the theory of education to the investigation of the relationships between cognition and digital technology. Educacion XX1 2022, 25, 223–241. [Google Scholar] [CrossRef]
  65. Liu, S. University Curriculum Setting and Teaching Governance System Based on Distributed Sensor Network. J. Sens. 2022, 2022, 2395417. [Google Scholar] [CrossRef]
  66. Papanastasiou, G.; Drigas, A.; Skianis, C.; Lytras, M.; Papanastasiou, E. Virtual and augmented reality effects on K-12, higher and tertiary education students’ twenty-first century skills. Virtual Real. 2019, 23, 425–436. [Google Scholar] [CrossRef]
Sustainability 15 02784 g001 550
Figure 1. Trend of each variable at the cognitive ability quantile.
Figure 1. Trend of each variable at the cognitive ability quantile.
Sustainability 15 02784 g001
Table
Table 1. Variables instructions.
Table 1. Variables instructions.
Name of the VariableVariable TypeData TypeVariable Instructions
Cognitive abilityPredicted variableContinuous variableraw score
Digital education technologyExplanatory variableCategorical variableNot using digital education technology = 0, using digital education technology = 1.
Educational informatizationControl variableCategorical variableNot Using Three Supplies Two Platforms = 0, Using Three Supplies Two Platforms = 1.
Educational background of motherContinuous variableContinuous variable generated based on educational level (0–8)
Educational background of fatherContinuous variableContinuous variable generated based on educational level (0–8)
Internet Accessibility Categorical variableFamily does not have the Internet = 0, family has the Internet = 1.
Computer AccessibilityCategorical variableFamily does not have a computer = 0, family has a computer = 1.
GradeCategorical variable0 = grade 7, 1 = grade 9
GenderCategorical variable0 = male, 1 = female
Self-efficacyCategorical variable0 = no 1 = yes
Academic performanceCategorical variable0 = below average, 1 = above average
Class capacityCategorical variable0 = less than 40, 1 = more than 40
School management styleCategorical variable0 = not related, 1 = related
Salary attitudeCategorical variable0 = not related, 1 = related
Table
Table 2. Descriptive data.
Table 2. Descriptive data.
VariableMeanSDMinMax
Cognitive ability9.97883.7444022
Digital education technology0.41230.492301
Educational informatization0.84990.355601
Educational background of mother2.81131.950308
Educational background of father3.18081.966908
Internet Accessibility0.60550.486201
Computer Accessibility0.68990.460101
Grade0.47120.499201
Gender0.48850.496601
Self-Efficacy0.85330.352501
Academic performance0.39460.487801
Class capacity0.78820.408601
School management style0.47840.499501
Salary attitude0.47300.499301
Table
Table 3. OLS estimation outcome.
Table 3. OLS estimation outcome.
VariablesCoefficientRobust SEtVIFTolerance
Digital education technology0.261 ***0.0495.301.0300.971
Educational informatization0.671 ***0.0729.291.1100.900
Educational background of mother0.177 ***0.01710.591.8600.538
Educational background of father0.139 ***0.0178.431.8400.543
Internet Accessibility0.810 ***0.0899.063.2900.304
Computer Accessibility0.464 ***0.0954.893.3200.301
Grade−1.495 ***0.049−30.611.0300.966
Gender−0.241 ***0.049−4.971.0100.991
Self-Efficacy0.470 ***0.0706.741.0500.949
Academic performance2.100 ***0.05141.421.0600.941
Class capacity−0.178 ***0.059−3.001.0300.974
School management style−0.417 ***0.050−8.351.0900.921
Salary attitude−0.148 ***0.050−2.981.0700.936
Constant7.549 ***0.11366.53
Observations18,579
Adjusted R20.240
F450.53
*** p < 0.01.
Table
Table 4. CEM result.
Table 4. CEM result.
Non-Digital Education TechnologyDigital Education TechnologyTotal
Total Sample11,153782518,978
Matched Sample6572556012,132
Unmatched sample458122656846
L1 Before matching0.536
L1 After matching0.016
Table
Table 5. Regression result (CEM weighted).
Table 5. Regression result (CEM weighted).
Outcome: Cognitive AbilityCoefficientStd.err.tp
Digital education technology0.239 ***0.0604.010.000
Educational informatization0.668 ***0.1235.440.000
Educational background of mother 0.134 ***0.0245.610.000
Educational background of father0.151 ***0.0236.460.000
Internet accessibility1.225 ***0.1319.370.000
Computer accessibility0.0580.1370.420.672
Grade−1.606 ***0.063−25.660.000
Gender−0.221 ***0.060−3.690.000
Self-Efficacy0.560 ***0.1075.250.000
Academic performance2.003 ***0.06431.460.000
Class capacity−0.284 ***0.091−3.140.002
School management style−0.315 ***0.063−4.970.000
Salary attitude−0.195 ***0.062−3.150.002
Constant7.743 ***0.18242.500.000
Observations12,132
Adjusted R20.239
F293.92
*** p < 0.01.
Table
Table 6. Quantile regression results.
Table 6. Quantile regression results.
(1)(2)(3)(4)(5)(6)(7)(8)(9)
VARIABLESq10q20q30q40q50q60q70q80q90
Marginal effects0.114
(0.0898)		0.179 *
(0.0950)		0.152 *
(0.0875)		0.189 **
(0.0814)		0.245 ***
(0.0803)		0.194 **
(0.0808)		0.306 ***
(0.0836)		0.390 ***
(0.103)		0.500 ***
(0.114)
Control variablesYesYesYesYesYesYesYesYesYes
Constant4.486 ***
(0.250)		5.619 ***
(0.274)		6.516 ***
(0.229)		7.183 ***
(0.219)		7.943 ***
(0.223)		8.545 ***
(0.234)		9.194 ***
(0.237)		10.07 ***
(0.264)		11.50 ***
(0.334)
Pseudo R20.1280.1450.1470.1480.1470.1430.1370.1290.110
Observations12,132
reps5000
*** p < 0.01, ** p < 0.05, * p < 0.1.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Yan, D.; Li, G. A Heterogeneity Study on the Effect of Digital Education Technology on the Sustainability of Cognitive Ability for Middle School Students. Sustainability 2023, 15, 2784. https://doi.org/10.3390/su15032784

AMA Style

Yan D, Li G. A Heterogeneity Study on the Effect of Digital Education Technology on the Sustainability of Cognitive Ability for Middle School Students. Sustainability. 2023; 15(3):2784. https://doi.org/10.3390/su15032784

Chicago/Turabian Style

Yan, Desheng, and Guangming Li. 2023. "A Heterogeneity Study on the Effect of Digital Education Technology on the Sustainability of Cognitive Ability for Middle School Students" Sustainability 15, no. 3: 2784. https://doi.org/10.3390/su15032784

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop