The impact of ICT characteristics on students’ learning outcomes during COVID-19 lockdown period: The crucial role of techno eustress perception

1 Introduction

Digital learning environments open up new pedagogical opportunities to educational institutions. Many studies have shown that the use of these technologies in education can bring multiple benefits [1–3], such as enhanced learning outcomes, increased learner engagement, and more efficient management and organization of learning [4]. These environments impact learning positively and have long been understood as a way “to think” [5], “to represent"[7], “to communicate” [7] or “to teach” [8].

Although digital technology undoubtedly motivates learners of all ages, including younger learners [9], there is gap in the literature regarding identifying the extant drivers influencing students’ learning performance during a situation of forced remote learning, as experienced during the COVID-19 pandemic [1]. Recent editorials and a number of scholars have explicitly called for a better understanding of this virtual transition for learners during the pandemic [86]. Some prior work claims that the relationship between the use of digital learning environments and academic performance is unclear, and conflicting results are presented in the literature [10, 11]. Similarly, another study has called for the development of a nuanced understanding of the use of digital technologies in education [12]. Researchers are urged to look into ways to ameliorate the negative aspects of digital technology usage, such as technology addiction, digital distraction, or cyberbullying in order to make digital education sustainable. Some researchers posit that the learner’s appraisal of the features of a digital learning system can have a positive influence on learning [1, 13]. However, the relationship between information communication technology characteristics (1) and students’ appraisals in a digital learning environment is yet to be empirically examined. Filling this research gap will help us to better understand how human systems perceive digital learning environments that have become a norm in many educational institutions around the world.

Recent studies have also called for a more integrative and holistic understanding of digital learning environments [14]. Specifically, understanding the affordances of various technologies (both functional and non-functional), their appraisal by learners, and their impact on learning is an important avenue for further theorization [14]. In the context of both positive and negative impacts of digital technology usage, it becomes interesting to develop a holistic approach on student learning outcomes and examine the conditions leading to positive impacts of digital learning environments on students’ performance during COVID lockdowns [87]. The findings of our study offer solutions for improving the effectiveness of educational programs, products, and learning outcomes through the successful appropriation of the digital learning environment by human systems.

Past research has shown a negative impact of digital technologies on individuals but few studies have analyzed the positive appraisal of technology stressors [15], apart from one study that addresses this aspect in the educational domain [1]. We attempt to contribute to this emerging stream of literature. In order to address the research gaps identified, we pose the following research questions:

To what extent were ICT characteristics positively appraised by students during remote digital learning during the pandemic and what might the impact of this appraisal be on student learning?

We leverage person-environment fit theory to study these questions, examining the influence of common ICT features that are often considered as demand stressors, namely usefulness and reliability of technology on student remote learning outcomes [16]. The learning outcomes we study in this research are digital learning environment-enabled productivity and innovation [1, 17]. To the best of our knowledge, this paper is the first to attempt to theorize the indirect role of ICT characteristics on student learning outcomes. Our conclusions will help us to understand how to foster positive appraisals during digital learning experiences and thereby better understand how human systems can be engaged as educational institutions invest in digital transformation initiatives. In the following sections, we elaborate on our theoretical framework and hypothesis development.

2 Background literature and theory

Tarafdar, Cooper, and Stich (2019) provide an excellent review of past research on techno distress situations [15]. Therefore, due to space constraints, we will not elaborate on the negative aspects of increasing IT demands here. In this paper, we focus on how and when challenge stress perceptions, also known as techno eustress perceptions, may be experienced by remote learners. In particular, we address the call to identify the technological features amenable to a positive appraisal [15]. Positive appraisals are humanistic intermediary outcomes that can lead to instrumental learning goals. Past research has looked at the relationship between the technological characteristics of ICTs and stressors. Ayyagiri, Grover, and Purvis (2011) has identified three categories of ICT characteristics viewed as antecedents for techno distress perception among end-users: usability features, dynamic features, and intrusive features [16]. Usability features are those characteristics related to the use and adoption of ICTs, including usefulness, reliability, and complexity. Dynamic features are those characteristics related to the pace of change and refer to the way in which ICTs are constantly evolving. Intrusive features refer to the invasiveness of ICTs, and include anonymity and presenteeism. Technology characteristics can influence the person-technology gap by manipulating the individual’s ability to deal with the demands of ICTs. They can also generate supply attributes that contradict the individual’s values, preferences, needs, or expectations. Usability features enhance an individual’s capabilities for dealing with technological challenges and results in reducing strain [11].

Based on person-environment fit theory, this paper argues that usability features, unlike dynamic and intrusive features, can contribute to creating a better fit between the learner and the digital learning environment, as they are viewed as demand-enhancing stressors. Thus, in this study we have incorporated two of the demand-enhancing stressors from the usability features, namely the usefulness characteristics of the ICT tools and the reliability characteristics of the ICT tools as antecedents to techno eustress perceptions.

We will now elaborate on our selected theoretical framework, the Person-Environment (P-E) Fit model. This model is widely used in stress research [18–20]. The P-E Fit model [21–23] defines stress as “a negative discrepancy between an individual’s perceived situation and desired state, provided that this imbalance is significant for the individual” [20] p. 242. The central hypothesis of this model is that a misfit between a person and their environment can produce psychological, physiological, and behavioral strains, e.g. irritation, dissatisfaction, boredom, anxiety, depression, whereas fitness perception should lead to positive cognitive and behavioral outcomes.

This model is based on the principle of a balance between people and their environment. When this balance is upset, it results in strain. The P-E Fit model explains the state of stress as the result of a mismatch between the demands and rewards of the work environment and the expectations and abilities of the people. P-E Fit can take two forms: either the environment provides the rewards and resources to meet the individual’s needs and preferences, or the individual’s skills, abilities, and aptitudes are adapted to the requirements and constraints of the job or position [23]. According to this model, a person-environment imbalance threatens well-being. In response to this tension-inducing imbalance, the individual can implement a range of adjustment strategies. If these adjustment behaviours are ineffective, the additive effects of stress can lead to problems in the long term, such as work dissatisfaction and disturbances in physical and mental health.

Selye has identified two types of responses to stress namely distress and eustress [24]. Distress is the negative outcome of all forms of stressors, physical or psychological, that exceed an individual’s capabilities. Eustress is the positive reactions an individual adopts when exposed to a stressor. Recently, by introducing challenge techno-stressors and challenge appraisal into the technostress literature, researchers have claimed that an individual’s use of ICTs can enhance and improve their skills by pushing them to learn how to use them and by improving their flexibility in different contexts [25, 26].

Further, ICT characteristics can introduce changes that can be perceived as a challenge, i.e. the opportunity to enhance skills, tasks, and work-life activities leading to positive outcomes, or as a threat, subjecting the individual to pressure and an inability to carry out the task leading to negative outcomes [15]. We can therefore perceive certain ICT characteristics as an opportunity in some contexts to enhance flexibility and productivity by enabling users to access and process information [26, 27]. The organizational literature suggests that types of stressors such as time pressure and job control are perceived as challenge stressors and they give rise to challenge appraisal, which is then linked to daily creativity and proactive behaviour [28]. Using a similar logic, we posit that certain ICT characteristics can be perceived as challenge stressors giving rise to challenge appraisals.

In this section, we explain how digital learning environments and their usability features can reduce the misfit described in the Person-Environment Fit model that is shown to cause stress among students. Reducing the gaps will lessen stress levels and will produce greater productivity and innovation in learning. We chose to focus on two learning outcomes –productivity and innovation in learning –as they offer a holistic examination of learning. Our research examines both task-focused (ICT-enabled productivity) and learning-focused outcomes (ICT-enabled innovation in learning process), which allows us to measure them from a learner’s perspective [58]. These two outcomes are also salient student outcomes when it comes to the use of technology for non-routine tasks [17, 59]. Figure 1 presents our research model with corresponding hypotheses, which will be further developed in this section.

OPEN IN VIEWER

Fig. 1

Research Model.

The technostress literature associates three ICT characteristics with ICT use and adoption: usefulness, complexity, and reliability. A lack of choice in the use and adoption of technologies at work [60] is common and is intrinsically linked to the characteristics and requirements of the job position [40, 61]. Despite having a low perception of their usability features, individuals would still have to use ICTs due to the requirements of their particular job. This contributes to a feeling of having to work harder [62]. One study suggests that using technologies for compliance reasons rather than through voluntary adoption creates additional conflicts, which result in more stress [63]. However, reliability and usability features increase the demand perceptions of users rather than contributing to supply-value incongruence [16].

Information systems are often evaluated according to their reliability and on how dependable and consistent the system is [64–66]. However, ICTs are becoming increasingly more complex and their reliability is difficult to ensure [67]. Reliability problems include software errors, quality problems, and failures [68–70]. This unreliability creates frustration and strain among users [62]. Therefore, from a P-E Fit model perspective, we suggest that reliable technologies would reduce the perceived workload and instead create demand ability fit perceptions; thereby triggering techno eustress, whereas unreliable systems will lead to strain due to the additional efforts required should the system break down. Thus, we hypothesize that:

H1: Reliability is positively associated with techno eustress perceptions

According to the P-E Fit model [16], the usability characteristics of ICTs enhance users’ ability to work better and lower the perceived demand ability gap. The more useful users perceive technologies to be, the more they will be able to use them to improve their performance. The individual will therefore perceive the use of technologies as a means of increasing their ability to cope with their workload. Reducing the perceived workload will reduce the misfit and lead to techno eustress. A recent study has also shown that usefulness perception reduces perceptions of technology spatial intrusion that end-users may feel in a remote working situation [10]. Thus, we hypothesize that:

H2: Usefulness is positively associated with techno eustress perceptions

When faced with technostress creators, the first step is to appraise whether the use of ICTs is useful, challenging, and reliable to users –in our case, students. The second step is to devise a suitable response by assessing the coping resources, options, and efforts needed to manage the technostress creators. In this context, individuals use ICTs to reduce their workload and their stress. Usability features will enable users to work better by reducing their perception of their workload and increasing their ability to cope with new technological challenges. We suggest it will impact positively on their productivity.

H3: Techno eustress is positively associated with students’ productivity

Following the same line of argument for the three prior hypotheses, and based on challenge appraisal, we posit that the use of ICTs will appear as challenging, as individuals will perceive them as a means of increasing their abilities and improving their competences to become more flexible, autonomous, and productive. The perception of ICTs’ usefulness will play a key role in the adoption of digital learning environments by students, thereby increasing their performance in learning tasks and contributing to innovation in learning process. Prior studies have also shown that digital technologies may be appraised as a challenge and contribute to creativity and innovation processes and outcomes [59]. Thus, we hypothesize that:

H4: Techno eustress is positively associated with students’ innovation

We adopted a two-wave survey method to collect data and test our hypotheses. The initial questionnaires were sent to approximately 500 students, who were students from a French business school. Respondents were contacted via their academic email addresses and through the school’s Facebook group. The first survey was sent in early March 2020 and stayed open for two weeks; the second was sent in late April 2020 and also stayed open for two weeks. During this period of time, students were exclusively taught via remote learning and were not allowed to visit the campus. The first phase consisted of collecting data on ICT characteristics, while the second aimed to collect data on techno eustress and the dependent variables of student innovation and productivity. After screening both sets of data collected during the two phases, we found a total of 82 usable responses, 75% of which came from female students with 25% from male students. The distribution regarding the year of study was: 49% of first to third-year students, 22% of fourth-year students, and 29% of fifth-year students. Working students made up 25% of the respondents. Students on average had three years of experience with ICTs and on average used ICTs for six to eight hours per day. We also incorporated suitable control variables, as shown in the research model (Fig. 1). For example, self-efficacy is known to contribute to learning outcomes that occur via ICT tools [1], hence we have incorporated it in our research model to isolate the potential influence of independent and mediator variables on our learning outcomes. Table 2 provides details of the scales used in this study.

As the sample size of this study was 82, we checked the smallest acceptable sample size to detect the effect of regression analysis at p < 0.05 using G*Power [75]. The desired power level is typically 0.80, but the researcher performing power analysis can specify a higher level, such as 0.90. Our study’s power analysis results revealed a smallest acceptable sample size of 74 for a power level of 0.95, so we can be reasonably sure, with a 95% probability, of not committing a type II error in our research with the given sample size. As seen in Fig. 2, the G power plot gives us the confidence that our sample size is sufficient to offer significant results.

OPEN IN VIEWER

Fig. 2

Power Analysis.

In this section we provide the implications stemming from our work both for theory and practice respectively.

7 Limitations and future research directions

We acknowledge that our study has several limitations. However, we believe that these limitations provide interesting avenues for future research. First, the COVID-19 pandemic has emphasized the relevance of the IT and digital transformation in all areas of business as well as in our personal lives. However, our study focuses only on the education sector, where there is an ongoing debate about how to increase 21st-century students’ productivity and innovation in learning through IT [95–97]. Future studies could theorize and examine other settings undergoing massive digital transformation because of COVID-19-induced changes. Replication and extension of the current model in non-academic settings that involve digital workforce transformation would help to increase the generalizability of our research findings.

Second, our study uses a two-wave survey method. While collecting the data in two phases mitigates methodological biases, researchers could consider using an experimental method to produce conclusive evidence for the causal impact of perceptions of usability features on learning outcomes. In a similar vein, qualitative studies on the subject could help us to better understand the nuanced processes through which usability features influence learning outcomes.

Third, we tested the theorized model via a unique data set collected during the first pandemic-driven lockdown period (from March to May 2020) in France. However, our sample size was relatively small. This was primarily because the study was designed as a two-wave matched pair survey conducted during a strict COVID-19-induced lockdown period in France. At that time, many students were forced to make an abrupt first time switch to different online learning modes. However, a G Power plot analysis shows that our study had sufficient power to establish conclusive results. Future researchers could conduct similar studies with a larger number of respondents, perhaps from different cultures and countries.

Fourth, in this study, we examined two generic usability features of the digital learning tools, i.e. usefulness and reliability. Examining the role of complexity characterized as a usability feature is a fertile avenue for future research. In addition, issues related to student engagement, mental health conditions (such as depression), and other academic performance outcomes have gained salience, especially among college students [98]. Examining these novel outcome variables could provide interesting directions for future work.

8 Conclusion

The need to better understand the impact of the imposition of digital learning technologies on students during and since the pandemic of 2020/21 is of utmost importance to researchers, pedagogical engineers, practitioners, and education policy makers. Research has made great strides in understanding broader technology design in terms of the selection of collaborative tools such as Zoom and Moodle platforms to quickly navigate the switch to online learning that became necessary during the COVID-19 pandemic. Studies report both negative and positive impacts of digital learning environments on teachers and learners. None of the studies has tried to undertake a nuanced investigation of the impact of technology features perception by learners using various digital learning environments. The theorization undertaken in this study has enabled us to understand under which conditions learners adopt digital learning platforms. Moreover, we are able to better understand how to design tools that generate positive learning outcomes. Since the forced transition has put the onus of remote learning on the student, it is important to study their person-technology fitness perceptions so we can understand how they cope with digital learning tool features that cause demand ability stressors so that they can be productive and creative during the learning process.

Our research provides a novel contribution to theory regarding ICT characteristics as an antecedent to techno eustress perceptions. Our study is one of the first to link ICT characteristics to learners’ learning outcomes via techno eustress perceptions. The ICT characteristic that causes the best person-environment fitness perceptions is the “usability” feature. This study demonstrates that certain usability characteristics, i.e. usefulness of the digital tool, are viewed positively by learners, which then leads to techno eustress perceptions. These techno eustress perceptions propel them to use the digital learning tool effectively for both productive and creative learning processes. Communicating the usefulness of digital learning tools to learners can provide humanistic intermediator outcomes of techno eustress perceptions and can lead to instrumental goals, such as higher productivity in learning and more innovative learning processes. We encourage practitioners and policy makers to incorporate our findings in their design and communication strategies.