Volume 3, Issue 3: Autumn 2022. Special Collection: Technology, Work, and Inequality. DOI: 10.1037/tmb0000081
In line with the recent literature, the aim of this article is to adopt a psychological approach to understand how technology is subjectively perceived and experienced at work, where the use of technology is seldom an individual choice, as well as its effects on employee’s well-being. This study aims to adopt a person-centered approach to create clusters of technology appraisal, explain such clusters’ membership through sociodemographics, and use these clusters to predict work-related well-being outcomes. In a sample of 692 Swiss working adults (Mage = 39.56, SD = 12.45, 60% female) active in both private and public sectors, this study first analyzed clusters of technology appraisal taking into account perceived usefulness, ease to use, and limitation of autonomy using a TwoStep cluster analysis. Then, these clusters’ membership was predicted by sociodemographic and individual characteristics (i.e., age, sex, education level, and generalized self-efficacy) using a multinomial logistic regression. Finally, differences in burnout, work engagement, and job boredom between clusters were examined using analyses of variance. Three different clusters of technology appraisal were found: The Tech-Enthusiasts, the Tech-Ambivalents, and the Tech-Detractors. Age, sex, educational level, and self-efficacy predicted clusters’ membership. Differences in burnout and work engagement were found between the clusters. No difference was found in boredom between the clusters. These findings highlight the importance of developing relevant and inclusive interventions to promote well-being and equality at work.
Keywords: technology appraisal, demographics, boredom, burnout, work engagement
Funding: Koorosh Massoudi made their contribution partly within the framework of the National Centre of Competence in Research-LIVES, financed by the Swiss National Science Foundation (Grant 51NF40-185901).Disclosures: The authors have no conflicts of interest to disclose.
Data Availability: The data that support the findings of this study are available under request to Cecilia Toscanelli.
Correspondence concerning this article should be addressed to Cecilia Toscanelli, Faculty of Social and Political Sciences, Institute of Psychology, University of Lausanne, Géopolis, CH-1015 Lausanne, Switzerland [email protected]
Fast and ongoing technological innovations have brought profound changes to the world of work, both at the structural level, by affecting the labor market’s structure and organization, and at the content level, by modifying the nature of the tasks and the demands faced by employees (International Labour Organization [ILO], 2019; Toscanelli et al., 2019). These changes have equivocal and unequal consequences on people’s employment opportunities. On the one hand, technological innovation has created new forms of work, thus leading to flexible and facilitated employment opportunities for independent high-skilled workers (Eurofound, 2015). On the other hand, the authors have also argued that unequal access to technology may undermine access to decent working conditions for vulnerable groups and hinder their efforts to attain job stability and security (Blustein et al., 2018; ILO, 2019; Masdonati et al., 2019; Massoudi et al., 2018).
These changes have also ambivalent effects on employees’ work-related well-being. While some authors recognize the beneficial effects of technological innovation in terms of increased flexibility and communication effectiveness at the workplace (Chesley, 2010), others have highlighted its detrimental effects on work and workers (ILO, 2019). For example, technological applications have led to the automation of a large proportion of tasks, exposing certain employees to insufficient workload and understimulation, with adverse effects in terms of chronic boredom at the workplace, perceived lack of meaning, and work disengagement (Glaser, 2014). For others, on the contrary, these changes have led to increased workload and pace, frequent interruptions, and constant unpredictability, associated with high levels of stress and burnout (Ter Hoeven et al., 2016; Thomas et al., 2006). It appears thus that technological applications at the workplace can either represent a resource—by facilitating accessibility and efficiency—or a hindrance—by increasing mental demands or decreasing opportunities for growth and self-accomplishment (Ter Hoeven et al., 2016). Indeed, as brought by Brangier and Hammes (2007), technology at the workplace can act in two different ways: As a symbiote, resulting in mutual benefits for workers and organizations, or as a parasite, contributing to the human alienation of workers. Although a few studies have examined the effects of exposure to technology, more research is needed to disentangle such ambivalent relations between the use of technology and well-being at the workplace. In order to do so, the focus of new studies must move from mere exposure to technology to the way it is appraised by employees (Salanova & Schaufeli, 2000) to better understand the underlying psychological processes.
A large body of literature on the relationship between humans and technology highlights the conditions that lead people to the deliberate choice of adopting and using technology. In this regard, several authors refer to the technology acceptance model (TAM; Davis et al., 1989)—derived from the theory of reasoned action (Fishbein & Ajzen, 1977)—and its consequent models (e.g., Venkatesh & Bala, 2008). This approach aims at understanding the behavioral intention of using a technological application based on its perceived usefulness and ease of use (see King & He, 2006). Needless to say, both these choice criteria—perceived usefulness and perceived ease of use—may vary according to diverse factors, namely individual differences like subjective norms, computer self-efficacy, or anxiety level (Venkatesh & Bala, 2008), contextual variables such as task-technology fit (Goodhue & Thompson, 1995) or job relevance (Venkatesh & Bala, 2008), as well as sociodemographic characteristics, such as gender or age (King & He, 2006; Venkatesh & Bala, 2008).
Considering that “new technology is often so powerful that organizations cannot afford to ignore or not to buy and use them” (Burke & Ng, 2006, p. 90), the presence of technology in the work setting is often largely driven by organizational or market-based constraints, and way beyond employees’ individual choices. In this regard, Brangier and Hammes (2007) propose to approach the relationship between people and technology as based on a strong and constant contiguity, maintained by a mutual contribution. Accordingly, the relation between employees and technology could be of a symbiotic nature, in which technology contributes to or facilitates work or, in contrary, of a parasite nature, where technology represents a hindrance to employees’ autonomy and adds to discomfort and difficulties (ILO, 2019). This implies that workers benefit more or less from the introduction of a new technology, since their ability to update their skills and adapt to novelty may vary (Burke & Ng, 2006). Consequently, their inability—or unwillingness—to adapt to technological changes may put certain employees at risk of adverse outcomes in terms of reduced employability, high job insecurity, and low work-related well-being.
Since modern work is characterized by fast and frequent technological changes and applications, it appears important to take into account the subjective appraisal of technology and its major role in predicting well-being outcomes (Salanova & Schaufeli, 2000). In this study, elements from the TAM, namely perceived usefulness and perceived ease to use, have been retained to evaluate technology appraisal. Moreover, since the literature also highlights that technological applications, through automation of tasks and processes, may reduce opportunities to freely choose one’s work tasks or methods (e.g., ILO, 2019; Toscanelli et al., 2019), this study will also take into account the perceived impact of technology on employees’ autonomy.
To study the potential links between technology appraisal and work-related well-being, this study will draw on the job demands–resources (JD-R) model (Demerouti et al., 2001) and its integration within the technology appraisal context (Day et al., 2010). The JD-R model (Demerouti et al., 2001) posits that the core characteristics of a given work environment can be resumed in terms of job demands and resources. Demands are defined as “those physical, social, or organizational aspects of job that require sustained physical or mental effort and are therefore associated with certain physiological and psychological costs” (Demerouti et al., 2001, p. 501), while resources refer to those physical, social, or organizational aspects of job that may do any of the following: (a) be functional in achieving work goals; (b) reduce job demands at the associated physiological and psychological costs; (c) stimulate personal growth and development (Demerouti et al., 2001, p. 501). Drawing on this framework and its application to workplace technology (Day et al., 2010), it is pertinent to consider that technology, when perceived as difficult to use or useless for task performance, may represent a demand for employees, whereas perceived reduction of autonomy may position technology as a threat to their resources. On the opposite, a positive appraisal may translate employees’ perception of technology as a resource, hence leading to positive outcomes.
Based on the above-mentioned variables and adopting an exploratory person-centered approach, we thus seek to identify different profiles of employees regarding their appraisal of work-based technology. For example, we expect some employees to present a positive appraisal of each variable, perceiving technology as easy to use, useful, and contributing to free them from difficult or cumbersome tasks. At the opposite, it seems also logical to imagine that other employees may experience technology as difficult, devoid of utility, and somewhat imposed to them, and thus a threat to their autonomy. Moreover, we can also assume to find at least a mixed or average cluster, presenting more moderate combinations of these variables. Finally, given the documented positive associations between perceived ease to use and perceived usefulness (Venkatesh & Bala, 2008), we do not expect to find a combination in which these two variables are opposed to each other. Despite these a priori expectations, we formulate our first hypothesis in an exploratory fashion as follows:
Hypothesis 1: Distinct patterns (i.e., clusters) of perceived ease of use, usefulness, and impact on autonomy should emerge denoting a differing degree of positive or negative technology appraisal.
The current literature shows sociodemographic differences in individual appraisal and attitudes toward technology (e.g., Rojas-Méndez et al., 2017). For example, examining differences in the attitudes toward the Internet, Porter and Donthu (2006) found that age was negatively related to perceived ease to use Internet and positively related to perceived barriers, whereas educational level was positively related to perceived ease to use the Internet. Previous research has shown in this regard that, when compared to their younger counterparts, older employees report less experience and more anxiety in the use of technology (Porter & Donthu, 2006). In the framework of technology readiness, Rojas-Méndez et al. (2017) found similar results, confirming that younger people with higher education have a more positive attitude toward technology.
These differences can be explained by the idea that older participants could appraise technology more negatively than their younger counterparts because they may experience higher effort expectancy when facing technological innovations (Venkatesh et al., 2003). Indeed, since age “has been shown to be associated with difficulty in processing complex stimuli” (Venkatesh et al., 2003, p. 450), older employees could face more difficulties when aiming to acquire and update their digital skills. Furthermore, it also appears that the observed differences may in fact stem from generational, rather than age-related characteristics. Indeed, the authors highlight the timing effect of the exposure to and training in technology proficiency, since learning is more efficient when it occurs in early stages of life, as it can be the case for younger generations (Apella et al., 2020), and thus positively impact further performance and appraisals (Hurwitz & Schmitt, 2020). Finally, the existing literature points out that age alone is not a comprehensive predictor of technology appraisal, since other variables such as education and gender may be more determinant (Dodel, 2021).
Similarly, previous results on educational level have been explained by the associations between lower education and lower technology competency and knowledge as well as lower perceived ease to use (see Porter & Donthu, 2006). These results can be interpreted based on the analysis offered by Dodel (2021), who explains that “inequalities in digital skills tend to arise when their development is left to incidental learning” (p. 8). Therefore, it is possible that when compared with people with lower levels of education, those with higher levels of education had access to more life-long opportunities to learn new skills, namely those necessary to master technological tools and applications.
Rojas-Méndez et al. (2017) also examined the role of gender-related differences in attitudes toward technology, finding that, compared to men, women perceived more discomfort—defined as “a perceived lack of control over technology and a feeling of being overwhelmed by it” (Rojas-Méndez et al., 2017, p. 21)—as well as more insecurity—defined as “distrust of technology and skepticism about its ability to work properly” (Rojas-Méndez et al., 2017, p. 21). These results are in line with previous research, some studies reporting for example that, when confronted to new technologies, men experience less discomfort and insecurity (Tsikriktsis, 2004) and feel more self-confident (Elliott & Hall, 2005) than women. Other authors (e.g., Aguirre-Urreta & Marakas, 2010) have highlighted other gender-related attitudinal differences which may affect the relationship to technology, with women reporting lower levels of computer self-efficacy and risk-taking intentions, and higher levels of computer anxiety. In order to understand the gender-related differences in technology appraisal, it is important to consider gender within a societal context, referring to “the characteristics of women, men, girls and boys that are socially constructed” (World Health Organization, 2022). In this regard, the authors suggest that differences in technology appraisal are grounded in a strong psychological basis stemming from male-dominant systems (Venkatesh et al., 2000, 2003), in which systemic norms and barriers lead to several differences, such as a higher susceptibility of women to experience anxiety when facing technology or higher susceptibility to learned helplessness (Venkatesh et al., 2000), which may affect perceived behavioral control and self-efficacy of women, and thus negatively affect their attitudes toward technology. This is consistent with the idea that “gender affects perceptions about digital competence more than the competence itself” (Dodel, 2021, p. 6), research showing that women tend to underestimate their self-perceived skills as compared to their observed skills (Hargittai & Shafer, 2006). Accordingly, in our study, gender-based differences and disparities in technology appraisal will be approached as the consequences of social inequality mechanisms and role stereotypes, rather than deriving from traits or dispositions inherent to women.
In addition to the above-mentioned interpretations, it is also important to remind the role of the occupational positions and work environments occupied by more vulnerable employees (women, older, or less educated employees). In fact, disparities in the labor market or at the workplace could also contribute to differences in terms of digital skills. As highlighted by Day et al. (2010), variables such as organizational support can facilitate updating and developing one’s skills, and thus counter the potential difficulties posed by technology at work. On the contrary, in an unsupportive or unfavorable work environment, some employees may experience more difficulties to keep up with technological developments at work. Therefore, one could assume that a negative technology appraisal reported by female, older, or less educated employees, could derive from an accumulation of vulnerabilities due to their limited access to job resources—in terms of opportunities for skill learning and career development when compared to male, younger, or highly educated employees (Eurostat, 2018; Federal Statistical Office [FSO], 2020; ILO, 2019; Krajnakova & Vojtovic, 2017; Masdonati et al., 2019; Roberson et al., 2020; Secrétariat d’Etat à l’économie [SECO], 2019; Slack & Jensen, 2008).
Following this reasoning, an individual characteristic that seems particularly important in understanding the appraisal of technology, and which emerged several times in relation to the previously mentioned variables, is self-efficacy. Generalized self-efficacy has been defined as “beliefs in one’s capabilities to mobilize the motivation, cognitive resources, and courses of action needed to meet given situational demands” (Wood & Bandura, 1989, p. 407). This individual characteristic has been widely studied in the field of work and organizational psychology and studies have highlighted its role as protector against adverse well-being outcomes as well as facilitating positive outcomes such as performance or job satisfaction (see Schyns & Von Collani, 2002). Moreover, generalized self-efficacy has been shown to play a role concerning the perception of control over difficulties, and hence play a moderator role between the stressors and their outcomes (Jex & Bliese, 1999). Generalized self-efficacy can therefore be considered in the framework of TAM (Venkatesh & Bala, 2008) as an individual characteristic which could be useful in promoting better technology appraisal especially for vulnerable groups since it “can influence individuals’ perceptions of perceived usefulness and perceived ease of use” (Venkatesh & Bala, 2008, p. 276). For this reason, generalized self-efficacy will be taken into account in this study as well as its interactions with demographic variables.
Based on these findings, we thus expect our participants to differ in their technology appraisal based on their sex, age, level of education, and generalized self-efficacy. We thus formulate our second hypothesis as follows:
Hypothesis 2a: Men will have a higher chance of falling into the positive appraisal cluster, while women will have a higher chance of falling into the negative technology appraisal cluster.
Hypothesis 2b: Younger workers will have a higher chance of falling into the positive appraisal cluster while older workers will have a higher chance of falling into the negative technology appraisal cluster.
Hypothesis 2c: Workers with higher educational level will have a higher chance of falling into the positive appraisal cluster while workers with lower educational level will have a higher chance of falling into the negative technology appraisal cluster.
Hypothesis 2d: Generalized self-efficacy will interact with demographic variables increasing opportunities for women, older workers, and workers with a low educational level of falling into a positive cluster instead of a negative one.
As previously discussed, the link between technology appraisal and well-being can be theorized by taking into account the JD-R model (Day et al., 2010; Demerouti et al., 2001). Job demands and resources, comprising of factors that are work related (e.g., workload, autonomy, social support) or individual related (e.g., sense of coherence, optimism), lead to processes that negatively (i.e., health impairment processes) or positively (i.e., motivational processes) affect work-related health and well-being. For our study, three outcomes were taken into account: boredom at work, burnout, and engagement.
To study the potential outcomes of technology appraisal, we will retain three indicators that are particularly representative of well-being at the workplace, namely job boredom, burnout, and work engagement (see Schaufeli & Salanova, 2014). These three measures empirically represent three distinct constructs (Reijseger et al., 2013). Indeed, engagement is a well-established positive indicator of work-related well-being (Leiter & Bakker, 2010; Schaufeli & Salanova, 2014). On the opposite pole, burnout and boredom represent negative experiences in response to unfavorable working conditions, in terms of overstimulation (i.e., burnout) and understimulation (i.e., boredom; Harju et al., 2014; Schaufeli & Salanova, 2014).
Boredom at work can be defined as a state of ill-being, characterized by low arousal and displeasure, occurring in an understimulating work context (Loukidou et al., 2009; Schaufeli & Salanova, 2014). The literature points to the potential role of technology in the experience of boredom, since it can thwart employees’ needs for autonomy, stimulation, and skill utilization through the automation and routinization of work tasks and procedures. For example, Loukidou et al. (2009) explain that “the use of technology to routinize working practices has meant that the skills of workers, even in many white-collar jobs, exceed the requirement of their jobs” (p. 382). However, the impact of technology on boredom can be controversial, since it can also offer possibilities to employees to cope effectively with boredom through activities such as cyberloafing (Pindek et al., 2018). Moreover, Mael and Jex (2015) explain that heavy users of information technology outside the workplace may present high needs for stimulation, and thus be more prone to experiencing boredom at the workplace. It is also useful to remind that, even though job boredom has traditionally been studied as resulting from an understimulating work environment characterized by trivial and underchallenging tasks, it may also occur when employees are faced with overly demanding and complex tasks to perform (Westgate, 2020).
Following this rationale, we propose to retain boredom as a potential adverse outcome of negative technology appraisal for employees who experience technology use as taxing and difficult, especially when they fail to perceive its usefulness, as well as for those who perceive technology as a hindrance to skill utilization or autonomy in their work environment. In contrast, we expect that a positive appraisal of technology would be negatively linked to job boredom.
Burnout is defined as a triadic syndrome conceptualized by three main dimensions (Schaufeli & Salanova, 2014). The first dimension, namely exhaustion, refers to a state of fatigue and emotional depletion. The second dimension, cynicism, represents an indifferent and distant attitude toward one’s work. Finally, the third dimension labeled lack of efficiency represents feelings of lack of capability and achievement. At the opposite pole on the activation–deactivation continuum, work engagement is defined as “a positive, fulfilling, work-related state of mind” (Bakker & Demerouti, 2008, p. 209) characterized by three dimensions: (a) vigor, which represents the energetic and resilient approach of one’s work, (b) dedication, reflecting a sense of meaning, challenge, and pride experienced at work, and (c) absorption which refers to a state of concentration and immersion in one’s work (Schaufeli & Salanova, 2014). Even though it is possible to view these concepts as unidimensional, the authors still recommend measuring their constitutive dimension separately (de Bruin & Henn, 2013; Maslach, 1993; Maslach & Jackson, 1986).
Work engagement has been linked with several positive effects such as creativity, productivity, positive emotions, good health, and ability to mobilize resources (Bakker & Demerouti, 2008), while burnout has been linked to low job satisfaction, poor performance, and turnover intentions (e.g., Fogarty et al., 2000). In a study focusing on the relation between technology and work-related well-being, Ter Hoeven et al. (2016) found that the use of communication technology had a positive impact on employees’ engagement through increased accessibility and efficiency, while contributing to burnout through increased interruptions and unpredictability. These results corroborate the findings of a study conducted in 2012 (ten Brummelhuis et al., 2012) on the detrimental and beneficial outcomes of new ways of working (NWW) enabled by intensive use of information and communication technologies (e.g., remote working). This study showed that in such working modalities, increased flexibility, and control over communications were associated with higher engagement and lower exhaustion, while frequent interruptions were linked with exhaustion. An impact of techno-stressors (i.e., overload induced by technology, invasion of privacy, etc.) on exhaustion has also been highlighted (Maier et al., 2015). Moreover, at a between-person level, a longitudinal study conducted in 2018 highlighted the link between perceived pressure to be attainable through technologies at work (i.e., workplace telepressure) and higher levels of physical and cognitive exhaustion (Santuzzi & Barber, 2018).
Hence, we expect a positive technology appraisal to be linked positively to work engagement and negatively to burnout, and a negative technology appraisal to be linked positively with burnout and negatively to work engagement.
Based on the above-mentioned results and literature, we hypothesize the relations between technology appraisal and well-being as follows:
Hypothesis 3a: Negative technology appraisal patterns entail higher levels of job boredom when compared to positive technology appraisal patterns.
Hypothesis 3b: Negative technology appraisal patterns entail higher levels of burnout when compared to positive technology appraisal patterns.
Hypothesis 3c: Negative technology appraisal patterns entail lower levels of work engagement when compared to positive technology appraisal patterns.
In the present study, in line with the recent literature, we adopt a psychological approach to the relation between human and technology to better understand the subjective perceptions and experiences of technology by employees, and their effects on their sustained well-being, especially in an evolving work context where recourse to technology is seldom an individual choice. More precisely, in line with the paradigms proposed by Landers and Marin (2021), the present study aims at understanding purely psychological process, thus adopting a technology-as-context approach. Indeed, our study focuses mainly on participants’ appraisal of technology by taking into account technology in the background, while not specifically considering the type of technology itself, its characteristics or potential evolutions. This study will contribute to the literature by (a) adopting a person-centered approach to identify different patterns of technology appraisal, (b) explaining such patterns through individual and sociodemographic differences, and (c) using these patterns to predict outcomes in terms of work-related well-being. Figure 1 illustrates the conceptual model of the present study.
The sample was composed of participants working in different work sectors. A first subsample was collected in early 2019 in two public institutions in the French-speaking part of Switzerland, in collaboration with the respective Human Resource (HR) departments (n = 203). Data on a second subsample, comprised of participants working in public and private sectors, were collected in late 2019 (n = 198) and late 2020 (n = 234) by students enrolled in a three-credit methodology course at the University of Lausanne. Finally, a smaller third subsample was collected using a snowball sampling technique (n = 57). As a result, the total sample for this study was composed of 692 Swiss working adults (Mage = 39.56, SD = 12.45, 60% female).1 Concerning their activity domain, 17.6% of the respondents were executives, 20.2% were academic and liberal professionals, 27.5% were administrative personnel, 11.7% were active in intermediate professions,2 6.5% were sales personnel, and the remaining participants were blue-collar workers, such as craftsperson, machine operators, and unskilled workers. Nine percent of participants did not report their domain of activity. The average tenure of the sample was around 7 years (Mtenure = 6.75, SD = 7.67), with 14.2% not reporting this information. Finally, by rating the item on the presence of work-related technology (i.e., “technology is very present in my work” on a 5-point Likert scale going from strongly disagree to strongly agree), 70.7% of the sample agreed (rather or strongly) that technology was present in their work. Our study complied with American Psychological Association (APA) ethical standards.
Participants filled out an online questionnaire transmitted with a clickable link. At the beginning of the survey, the questionnaire thanked for agreeing to participate in this study and stated that responses would be treated as strictly anonymous, confidential, and in accordance with the ethical rules of the Swiss Psychological Society.
Several demographic and sociodemographic data were collected. The age of the participants was measured with an item, asking to indicate their age in a blank space. Concerning sex, participants had the option to choose the male or female answer (female = 1; male = 2). The educational level has been assessed with a multiple choice of 14 options (from 1 = compulsory school to 14 = PhD). Then, for this study, the education level was categorized into three different categories (i.e., 1 = compulsory school and lower secondary education; 2 = upper secondary education; 3 = tertiary education).
The level of generalized self-efficacy was measured using the General Self-Efficacy Scale (GSES; Schwarzer & Jerusalem, 1995), which consisted of 10 items that participants were asked to answer on a Likert scale ranging from 1 (not at all true) to 4 (completely true). The scale scores showed good reliability (Cronbach’s α = .88).
Based on the previously cited theoretical framework, technology appraisal has been measured with three items developed by the authors referring to different aspects (perceived ease of use, usefulness, and impact on autonomy) of technology in the workplace, with a 5-point Likert scale going from 1 (strongly disagree) to 5 (totally agree). The first item referred to technology in the workplace in relation to task-related perceived usefulness (“technology has improved the execution of my daily tasks”). The second item concerned the ease of use (“it is easy for me to use the technology”). Item 3 referred to the general impact on autonomy (“technology, in the context of my work, has limited my autonomy”).
Model Fit Summary of the Cluster Analysis for Technology Appraisal | ||||||
Model | No. of clusters | BIC | SC | Sm % | LS ratio | ItemDiff |
---|---|---|---|---|---|---|
4 | 3 | 992.537 | 0.4 | 20.1 | 2.39 | Yes |
3 | 4 | 905.865 | 0.5 | 10.7% | 4.36 | No |
2 | 5 | 848.618 | 0.4 | 9.5% | 3.02 | No |
1 | 6 | 798.956 | 0.4 | 7.9% | 3.09 | No |
Note. ANOVA = analysis of variance; BIC = Bayesian information criterion; SC = silhouette coefficient; Sm % = percentage smaller cluster; LS ratio = ratio between the effective of the largest and the smallest cluster; ItemDiff = mean differentiation of each item in each cluster based on ANOVA, post hoc Bonferroni. |
Job boredom was measured with the DuTCH Boredom Scale (DUBS; Reijseger et al., 2013) in its French version (Toscanelli et al., 2022). This instrument is composed of six items (e.g., “I feel bored at my job”). Participants answered to this questionnaire using a 5-point Likert-type scale ranging from 1 (never) to 5 (always). The scale scores showed good reliability (Cronbach’s α = .87).
Burnout and its subdimensions were measured with the Maslach Burnout Inventory—General Survey (MBI-GS; Schaufeli et al., 1996). This measure includes the subscale of exhaustion with five items (e.g., “I feel emotionally drained from my work”), the subscale of cynicism with five items (e.g., “I have become more cynical about whether my work contributes anything”), and the subscale of lack of efficiency with six items (e.g., “I can effectively solve problems that arise in my work”) to be rated on a 7-point Likert-type scale ranging from 1 (never) to 7 (always). In the present study, Cronbach’s α for the three subscales indicated good reliability (respectively α = .87; α = .83; α = .84).
Engagement and its subdimensions were measured with the Utrecht Work Engagement Scale (UWES; Schaufeli et al., 2006) in its French validation (Zecca et al., 2015). This scale includes nine total items. Three items measure vigor (e.g., “At my job, I feel strong and vigorous”), three items measure dedication (e.g., “I am proud on the work that I do”), and three items measure absorption (e.g., “I am immersed in my work”), to be rated on a 7-point Likert scale going from 1 (never) to 7 (always). In the present study, Cronbach’s αs for the three subscales indicated good reliability (respectively α = .80; α = .90; α = .80).
The analyses were conducted in three steps. First, a TwoStep cluster analysis was carried out using IBM SPSS software Version 26. The log-likelihood was used. This analysis was based on the assumption that all items were independent and not highly correlated (Bacher et al., 2004; Tkaczynski, 2017) and that items used followed a normal distribution (Bacher et al., 2004; Tkaczynski, 2017). Normality was assessed using skewness and kurtosis analyses. As we used a sample larger than 300 participants, either absolute skew values larger than two or absolute kurtosis values larger than seven were used as reference values to determine non normality (Kim, 2013). The best-fitting model was selected and validated based on several statistical fit criteria: Bayesian information criterion (BIC), the silhouette measure of cohesion and separation that should be equal or above 0.2 (Tkaczynski, 2017), the size of the smaller cluster, and the ratio of cluster sizes (larger cluster to smaller cluster) that should be below or around 2.0 (Tkaczynski, 2017). In order to validate our clusters, we have verified that no item had low ratings, and we ran an analysis of variance (ANOVA) to ensure that all items within a cluster were significantly different. Finally, we randomly separated the sample into two samples, ran again the cluster analysis and compared the news results with our previous cluster solution in order to validate our initial cluster solution (Tkaczynski, 2017).
In the second step, we ran a series of multinomial logistic regressions to test Hypothesis H2. Multinomial logistic regression models describe who is most likely to fall into a particular cluster rather than another. Hence, sex, age, education level, and generalized self-efficacy were used as exploratory factors. In order to obtain a complete and nuanced view of the association between these variables and the different clusters, we alternated the reference group and all the clusters were compared. This resulted in a model in which generalized self-efficacy, all sociodemographics, and their interaction with self-efficacy were included as predictors of cluster membership.
In the last step, we ran a series of one-way ANOVAs to test Hypothesis H3, namely the associations between, on the one hand, technology appraisal clusters and, on the other, burnout, engagement, and job boredom. All the groups were compared. ANOVAs were followed by multiple comparisons using Bonferroni post hoc t tests. Moreover, homogeneity of variances was assessed using Levene’s test (homogeneity is determined with a nonsignificant p value of above .05) and normality was assessed using skewness and kurtosis analyses. Again, as we used a sample larger than 300 participants, either absolute skew values larger than two or absolute kurtosis values larger than seven were used as reference values to determine non normality (Kim, 2013).
Results for the normality test indicated a normal distribution for each item. Table 1 indicates results concerning clusters indicators and criterion of validation. By letting the statistical software free to determine the number of clusters, we obtained a six clusters solution of technology appraisal. Even though presenting a fair silhouette coefficient of 0.4, this solution was problematic due to the poor interpretability of the clusters, as well as the small number of participants in some of them, with a ratio of sizes (larger cluster to smaller cluster) of 3.09. Hence, we proceeded to investigate reduced cluster solutions. A three-cluster solution, with a silhouette coefficient of 0.4 and a ratio of sizes of 2.39 was retained in order to (a) obtain interpretable clusters and (b) obtain sufficient participants per cluster to allow subsequent analyses based on significant differentiation between item means in each cluster. The importance of predictors in this three profiles clustering was (a) Item 3 (i.e., “technology, in the context of my work, has limited my autonomy”; importance = 1), (b) Item 2 (“it is easy for me to use the technology,” importance = 0.89), and (c) Item 1 (i.e., “technology has improved the execution of my daily tasks,” importance = 0.66).
The three-profile solution is composed as follows. The largest profile (48%) presents high scores on Item 1 (M = 4.30, SD = .56) and Item 2 (M = 4.47, SD = .54) and low scores on Item 3 (M = 1.51, SD = .50). Hence, this group is composed by workers for whom technology is useful, easy to use, without reducing their autonomy. This cluster was labeled Tech-Enthusiasts. The second cluster (31.9%) includes high-average scores on Item 1 (M = 3.59, SD = .88), high scores on Item 2 (M = 4.30, SD = .53), and high-average scores on Item 3 (M = 3.20, SD = .76). This group is characterized by workers who perceive technology as limiting their autonomy, but still improving the execution of their tasks and easy to use. This cluster was labeled Tech-Ambivalents. The last and smallest group, Cluster 3 (20.1%), includes low-average scores on Item 1 (M = 2.65, SD = 1.14), low-average scores on Item 2 (M = 2.86, SD = 1.00), and low scores on Item 3 (M = 2.20, SD = 1.04). This third group is characterized by people reporting that technology, even though not a threat to their autonomy, is not very easy to use, nor really helpful. This cluster was labeled Tech-Detractors.
Means, standard deviations, and Pearson’s bivariate correlations are reported in Table 2. The results of the correlation analysis showed small correlations between demographics, small correlations between demographics and items of technology appraisal, as well as small correlations between demographics and clusters, in line with the theoretical framework.
Bivariate Pearson’s Correlations of All Variables | |||||||||||||||||
Variable | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1. Age | — | ||||||||||||||||
2. Sex | .00 | — | |||||||||||||||
3. Edu. level | .01 | .12** | — | ||||||||||||||
4. Self-efficacy | .07 | .13** | .16** | — | |||||||||||||
5. Technology 1 | .00 | .12** | −.05 | .03 | — | ||||||||||||
6. Technology 2 | −.23** | .13** | .07 | .25** | .31** | — | |||||||||||
7. Technology 3 | .01 | −.01 | −.12** | −.14** | −.08* | −.18** | — | ||||||||||
8. Cluster 1 | −.03 | .09* | .13** | .15** | .52** | .40** | −.63** | — | |||||||||
9. Cluster 2 | −.08* | −.02 | −.10* | −.08* | −.10** | .16** | .67** | −.66** | — | ||||||||
10. Cluster 3 | .13** | −.09* | −.05 | −.10** | −.53** | −.68** | .01 | −.48** | −.34** | — | |||||||
11. Job boredom | −.28** | .05 | .02 | −.19** | −.07 | .03 | .07 | −.06 | .05 | .02 | — | ||||||
12. Exhaustion | −.07 | −.06 | .02 | −.22** | −.14** | −.12** | .20** | −.18** | .14** | .05 | .32** | — | |||||
13. Cynicism | −.02 | .00 | .02 | −.25** | −.20** | −.12** | .27** | −.24** | .16** | .12** | .44** | .59** | — | ||||
14. Lack of efficiency | −.09* | −.08* | −.06 | −.46** | −.10** | −.18** | .18** | −.20** | .09* | .15** | 33** | .26** | .46** | — | |||
15. Vigor | .16** | −.03 | −.06 | .35** | .11** | .08* | −.10** | .12** | −.08* | −.06 | −.53** | −.48** | −.53** | −.51** | — | ||
16. Dedication | .09* | .02 | .04 | .30** | .16** | .08* | −.16** | .17** | −.12** | −.08* | −.48** | −.36** | −.57** | −.57** | .76** | — | |
17. Absorption | .13** | −.00 | −.00 | .22** | .16** | .06 | −.02 | .07 | −.00 | −.09* | −.50** | −.20** | −.37** | −.45** | .68** | .72** | — |
M | 39.56 | 1.40 | 2.22 | 3.22 | 3.74 | 4.09 | 2.19 | 0.48 | 0.32 | 0.20 | 2.40 | 3.20 | 2.91 | 2.94 | 4.98 | 5.22 | 5.08 |
SD | 12.45 | 0.49 | 0.92 | 0.47 | 1.03 | 0.90 | 1.03 | 0.50 | 0.47 | 0.40 | 1.01 | 1.26 | 1.25 | 0.90 | 1.02 | 1.25 | 1.09 |
Note. Sex: 1 = female; 2 = male; Technology 1 = “technology has improved the execution of my daily tasks”; Technology 2 = “it is easy for me to use the technology”; Technology 3 = “technology, in the context of my work, has limited my autonomy”; Cluster 1 = Tech-Enthusiasts; Cluster 2 = Tech-Ambivalents; Cluster 3 = Tech-Detractors; N = 692; M = mean; SD= standard deviation. * p < .05. ** p < .01. |
Concerning their correlation with demographics, Tech-Enthusiasts positively correlated with sex, educational level, and generalized self-efficacy, and negatively correlated with the three dimensions of burnout and positively with two dimensions of engagement, namely vigor and dedication. The cluster Tech-Ambivalents correlated negatively with age, educational level, and generalized self-efficacy, and positively with the three dimensions of burnout and negatively with two dimensions of engagement (vigor and dedication). Finally, the third cluster, Tech-Detractors, correlated positively with age and negatively with sex and generalized self-efficacy, as well as positively with the two dimensions of burnout, namely cynicism and lack of efficiency, and negatively with two dimensions of engagement, namely dedication and absorption. The three clusters were negatively correlated with each other.
Table 3 shows the results for the logistic regression. Concerning age, the significant odd ratio below 1 for Tech-Detractors compared to Tech-Enthusiasts, indicated that younger participants are more likely to belong to Tech-Enthusiasts than to Tech-Detractors. Concerning the comparison to Tech-Ambivalents, odds ratios below 1 indicated that younger participants are more likely to belong to Tech-Ambivalents than to Tech-Detractors. Concerning sex, the comparison between Tech-Detractors and Tech-Enthusiasts showed an odds ratio above 1, indicating that women are more likely to belong to Tech-Detractors, than to Tech-Enthusiasts. Concerning educational level, the comparison of Tech-Enthusiasts and Tech-Ambivalents to Tech-Detractors showed an odds ratio above 1 for the lowest educational level, indicating that participants with the lowest educational level are more likely to belong to Tech-Ambivalents and to Tech-Detractors, than to Tech-Enthusiasts. Moreover, results showed that people with an average level of education are more likely to belong to Tech-Ambivalents than to Tech-Detractors. Finally, the interaction between sex and generalized self-efficacy was significant when comparing Tech-Ambivalents to Tech-Enthusiasts. Figure 2 enables us to graphically interpret this result, which indicates that middle and higher levels of generalized self-efficacy in women increase the probability to belong to Tech-Enthusiasts rather than to Tech-Ambivalents.
Multinomial Logistic Regression Analyses of the Type of Technology Appraisal Clusters: Odds Ratios | ||||
Predictor variables | Coef. | Compared clusters | Odds ratios | 95% CI |
---|---|---|---|---|
Age | .027 | 3 versus 1 | 1.03** | [1.009, 1.045] |
Sex (female) | .524 | 3 versus 1 | 1.69* | [1.078, 2.648] |
Educational level | ||||
1 | .510 | 2 versus 1 | 1.66** | [1.127, 2.459] |
2 | −.866 | 3 versus 2 | .42* | [.178, .992] |
3 (ref.) | ||||
Self-efficacy | −.536 | 2 versus 1 | .59** | [.396, .865] |
Self-efficacy × Sex | .301 | 2 versus 1 | 1.35** | [1.071, 1.704] |
Note. CI = confidence interval; Cluster 1 = Tech-Enthusiasts; Cluster 2 = Tech-Ambivalents; Cluster 3 = Tech-Detractors; reference cluster is indicated in bold. Only significant results are summarized, based on higher odds of belonging to a certain cluster over the reference cluster. Nonsignificant results can be obtained by writing to the corresponding author. N = 692. Coef. = logit coefficient, lower bound and upper bound of the 95% confidence interval are indicated in brackets as follows (lower bound; upper bound); Self-efficacy × Sex = interaction between self-efficacy and sex. * p < .05. ** p < .01. |
Results for the normality test indicated a normal distribution of the data, and variances were homogeneous. Results of the ANOVAs are reported in Table 4. Concerning the three dimensions of burnout, namely exhaustion, cynicism, and lack of efficiency, results showed a significant difference between the scores obtained by participants in Cluster 1 and those in the other two clusters. Indeed, Cluster 1 presents significantly lower scores on all three dimensions of burnout when compared to Clusters 2 and 3. Concerning work engagement, participants in Cluster 1 showed significantly higher scores of vigor and dedication compared to scores reported by participants in Clusters 2 and 3. Concerning absorption, participants in Cluster 1 showed significantly higher scores compared to those reported by participants in Cluster 3. Finally, concerning job boredom, no significant difference between the three groups was found.
ANOVA for Each Cluster for Burnout, Engagement, and Job Boredom | ||||||||||||
Job boredom | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Cluster | Mean | N | F | Difference | ||||||||
Cluster 1 | 2.335 | 85 | 1.317 | ns | ||||||||
Cluster 2 | 2.468 | 388 | ||||||||||
Cluster 3 | 2.444 | 219 | ||||||||||
Burnout | ||||||||||||
Exhaustion | Cynicism | Lack of efficiency | ||||||||||
Mean (SD) | N | F | Difference | Mean (SD) | N | F | Difference | Mean (SD) | N | F | Difference | |
Cluster 1 | 2.969 | 85 | 11.719*** | C1 < C2*** | 2.592 | 85 | 21.836*** | C1 < C2*** | 2.758 | 85 | 15.205*** | C1 < C2*** |
Cluster 2 | 3.466 | 388 | 3.205 | 388 | 3.054 | 388 | ||||||
Cluster 3 | 3.337 | 219 | C3 > C1** | 3.194 | 219 | C3 > C1*** | 3.201 | 219 | C3 > C1*** | |||
Work engagement | ||||||||||||
Vigor | Dedication | Absorption | ||||||||||
Mean (SD) | N | F | Difference | Mean (SD) | N | F | Difference | Mean (SD) | N | F | Difference | |
Cluster 1 | 5.103 | 85 | 5.030** | C1 > C2* | 5.443 | 85 | 10.092*** | C1 > C2*** | 5.168 | 85 | 3.132** | C1 > C3* |
Cluster 2 | 4.858 | 388 | 5.012 | 388 | 5.080 | 388 | ||||||
Cluster 3 | 4.861 | 219 | C3 < C1* | 5.034 | 219 | C3 < C1*** | 4.892 | 219 | ||||
Note. df for each ANOVA were (2, 689). n = 692. ns = not significant. In the Difference column, only significant differences are reported; ANOVA = analysis of variance. |
The first aim of this study was to adopt a psychological and person-centered approach to identify different patterns of technology appraisal. Our first exploratory hypothesis was supported by our results, allowing us to identify three clusters that represent sensibly different perceptions of technology at the workplace. The first cluster, labeled Tech-Enthusiasts, gathers employees who present a general positive appraisal, by appraising technology as user-friendly, useful, and not threatening their autonomy. Furthermore, two other profiles, presenting moderate levels and mixed combinations of our variables were found. A second cluster, labeled Tech-Ambivalents, regrouped employees reporting somewhat mixed feelings about technology, perceiving its limiting impact on their autonomy, but still recognizing its ease of use and its contributions to facilitate their work. Finally, a third cluster named Tech-Detractors brought together employees who did not perceive technology as hindering their autonomy, but still doubted their capacities to use it and its useful contribution to their professional tasks.
The second aim of this study was to predict and explain membership in these clusters based on participants’ sociodemographic characteristics. As formulated by our second hypothesis, we expected that men, younger employees, and employees with higher educational levels and higher generalized self-efficacy would have higher chances of falling into a positive appraisal cluster, while women, older participants and, participants with lower educational levels would more likely fall into the negative technology appraisal clusters. Results confirmed our hypothesis. Concerning age, when compared to older employees, youngest ones were indeed more likely to belong to the Tech-Enthusiasts than the Tech-Detractors cluster, and also had higher chances of being Tech-Ambivalents rather than Tech-Detractors. The middle-aged group had a higher chance to fall into the Tech-Ambivalents than the Tech-Detractors cluster. In sum and as indicated by previous research (Porter & Donthu, 2006; Rojas-Méndez et al., 2017), age does seem to matter when it comes to the employees’ relations to technology and their appraisal of its practicality and purpose. Concerning sex, results also confirmed our hypothesis, women being more likely to belong to Tech-Detractors than to Tech-Enthusiasts, indicating a less positive appraisal of technology and more concerns regarding the abilities to efficiently use it to pursue their work goals. Concerning educational level, workers with a low educational level were more likely to belong to Tech-Ambivalents and Tech-Detractors than to Tech-Enthusiasts, compared to highly qualified workers, while workers with average educational level (i.e., upper secondary education) were more likely to be Tech-Ambivalents than Tech-Detractors. Finally, the interaction between sex and self-efficacy was significant when comparing Tech-Ambivalents to Tech-Enthusiasts, indicating that women with medium and higher levels of generalized self-efficacy are more likely to belong to the positive appraisal cluster than to Tech-Ambivalents, while for men the effect of self-efficacy was not found (i.e., men having higher chances to fall into the positive cluster regardless of their level of self-efficacy). To sum, and in line with our expectations, it appears that younger, male, most educated employees, and interestingly women with medium and high self-efficacy are indeed those who are most likely to experience a symbiotic relation (Brangier & Hammes, 2007) with technology, feeling confident in their ability to use it to optimize their work, without feeling restrained or controlled by it. Our findings may also point to the fact that, due to certain sociodemographic characteristics (i.e., being a woman, older of age, or less educated), certain employees may experience difficulties to keep up with technological developments at work. Hence, as it will be explained later, it is important to develop appropriate interventions in order to help vulnerable employees in acquiring the necessary resources to cope with difficulties linked with technology in the workplace.
Beyond identifying these differences in technology appraisal at the workplace, the third aim of our study was to investigate their associations with well-being outcomes for employees. In this regard, we expected people with a positive appraisal of technology to report less job boredom, lower levels of burnout, and higher levels of engagement than their counterparts with a more mitigated or negative appraisal of technology. Results partially confirmed our hypothesis. First, when compared to both Tech-Ambivalents and Tech-Detractors, Tech-Enthusiasts reported lower levels on all three dimensions of burnout, as well as higher levels on two dimensions of engagement, namely vigor and dedication. Moreover, Tech-Enthusiasts showed higher levels of absorption than Tech-Detractors. In line with the existing literature (Ter Hoeven et al., 2016), we can interpret these results through the lens of the double role of technology as a symbiote—a mean for positive performance and engagement for workers equipped with the necessary skills to most benefit from it—or a parasite—adding to the stress and demands experienced by less comfortable workers while hindering their autonomy. Moreover, it is important to stress that the introduction of novel technological solutions and applications (e.g., Smart Technology, Artificial intelligence Robotics and Algorithm [STARA]) can lead certain employees, especially whose with a negative appraisal of technology, to experience feelings of distress, insecurity, reduced self-efficacy and sense of purpose, and to develop distant and cynical attitudes toward their work, thinking “why bother for my job, once the robot is programmed I will be given root” (Oosthuizen, 2019, p. 32). Concerning job boredom, our analyses did not show significant differences between the three clusters. Nevertheless, this inconclusive result can be explained by the fact that none of the identified clusters seem to face working conditions (i.e., low demand but especially low resources) that are recognized as predictors of boredom (Reijseger et al., 2013). For instance, while Tech-Detractors might experience difficulties using technology and making sense of its utility, they still preserve their autonomy, which is a core resource to cope with adverse working conditions (Mikkelsen et al., 1999) and especially job boredom (Harju et al., 2014; Reijseger et al., 2013). On the other hand, although threatened in their autonomy, Tech-Ambivalents still report abilities to efficiently use technology to pursue their work goals, which may in turn satisfy their need for stimulation and accomplishment and preserve them from boredom (van Hooft & van Hooff, 2018). Job boredom could hence characterize a totally negative appraisal pattern, in which neither autonomy nor technology-relevant skills are present.
These findings lead us to several considerations. First, from a methodological point of view, this study contributes to the literature on the TAM model—which has mainly been studied following a variable-centered approach—using a person-centered approach to propose a group-based point of view on technology appraisal. Then, by highlighting different patterns of technology appraisal, our results point to different needs of employees, which may in turn contribute to the identification and implementation of different types of actions and interventions that could help them cope with technological change and innovation. Third, results may point to barriers and inequalities faced by some employees—namely older, less qualified, or women in the workplace. As mentioned in the theoretical part of this study, the work environment occupied by more vulnerable populations (women, older employees, and less educated employees) and the work disparities that may result from it could also contribute to the development of disparities in terms of digital skills. These results confirm other findings on precariousness and vulnerabilities encountered by older, less qualified, and female workers in the labor market. Indeed, previous studies have reported higher risks of underemployment an unemployment for older workers (above 55 years) when compared to their middle-aged counterparts (Slack & Jensen, 2008). Moreover, research has also highlighted the need for policies to support older workforce to maintain durable inclusion and develop sustainable careers (Krajnakova & Vojtovic, 2017), since they experience more difficulties to find a new job when losing theirs and thus are more at risk of long unemployment (SECO, 2019). Concerning gender inequalities, it is also well documented that women face specific risks and barriers in the labor market. For instance, women were found to have an inferior mean salary compared to men (20% less in 2019), while facing higher risks of violence and harassment in the workplace (ILO, 2019), and more difficulties to develop their careers and gain access to leadership positions (Roberson et al., 2020). Finally, it also appears that lower level of education and qualification is associated with lower employment rates and higher risks of unemployment (Eurostat, 2018; FSO, 2020). Even though these inequalities are well documented in the literature, our findings contribute to the latter in several ways. First, by linking cluster membership to sociodemographic characteristics, results suggest that part of the inequalities at work may be reinforced by the way employees appraise technology at the workplace. Second, by also considering well-being indicators, results also show that different appraisal patterns could correspond to differences in terms of work-related well-being.
In line with ILO’s (2019) recommendations, the practical implications of this study point to tailored interventions to improve the appraisal, acceptance, and incorporation of work-based technology. At an organizational level, it seems indeed important to take into account employees’ needs, skills, and perceptions when introducing technological applications, in order to enhance their contribution potential and minimize their hindering effects. The appraisal patterns identified in our study could guide such interventions. For instance, when dealing with Tech-Ambivalents, it appears more important to anticipate and alleviate possible limiting impacts on their autonomy, rather than put emphasis on training. In this sense, interventions aiming at autonomy improvement (e.g., job crafting and autonomy support, see Slemp et al., 2015) might be useful. On the other hand, Tech-Detractors could need to reinforce their skills to adapt to technological innovations, but also better understand its potential purpose. Accordingly, organizations should provide the necessary resources (awareness-raising measures, continuous training, gender, and age-equitable management) to help employees overcome disadvantages and develop the necessary skills to cope with technological demands.
This study highlights the importance of adopting a psychological perspective to better understand the relationship between workers and technology and identify its antecedents and consequences. Moreover, this study highlights the need for a more sensible and supportive introduction of technology developments in the workplace, in order to avoid technology-related inequalities and their consequent adverse outcomes on well-being and performance.
Our study has several limitations. First, the unequal distribution of participants across sociodemographic categories did not allow to test the interactions between these categories. Future research could, therefore, use larger samples to test interactions between sociodemographic variables for logistic regression. Then, the nonrandom nature of our sample implies precautions in the interpretation of the results, since the effects found for the different demographic factors may be partly due to the specific occupations/organizations studied. Moreover, the items used to measure technology appraisal have been developed by the authors and hence have not been validated. Future research should use existing validated scales or better investigate psychometric properties of these items. In addition, it should be reminded that our study is based on a cross-sectional design and evaluates the differences in outcomes through ANOVAs. Hence, causality cannot be established. Following this reasoning, it could be possible that employees who experience burnout tend to appraise technology negatively (i.e., emotional exhaustion and cynicism lead to negative attitudes), rather than the reverse. In order to better establish the causal links between clusters and outcomes, a longitudinal design is recommended. Moreover, our study considered demographics and generalized self-efficacy as predictors of clusters. Based on the TAM theory, future studies should integrate other predictors of perceived ease to use and usefulness at an individual level, such as computer anxiety or computer self-efficacy, in order to better understand technology appraisal. Then, based on the analysis of Landers and Marin (2021), the framework applied in this study (i.e., technology-as-context) does not allow for identifying tailored solutions to address issues and needs rising from the utilization of specific types of technology. Future research should then adopt a technology-as-designed paradigm to study different technological applications and tools “in terms of their specific design characteristics, users, intended users, and how each of these might change in the future as the technology is redesigned and redeveloped over time” (Landers & Marin, 2021, p. 241). Such an approach will contribute to a better understanding of the way technology may impact and influence employees’ daily experiences at work, by offering “accuracy in describing and predicting real-world phenomena” (Landers & Marin, 2021, p. 254). Finally, our study was conducted before the COVID-19 pandemics and during October 2020. Further research should investigate more thoroughly the impact of the pandemic in terms of the fast and massive technological mobilization required from employees.