Technostress Creators Inventory among Slovak employed respondents

Description

Recent technological advancements in the workplace can produce contradictory effects on employees’ well-being and behaviour. Individuals can improve their task performance, but they may also suffer from negative outcomes such as stress, anxiety, and burnout. This quantitative and cross-sectional study aims to verify and validate the psychometric properties of the Slovak adaptation of the Technostress Creators Inventory (TCI). Six hundred and ninety-two employed respondents (50.7% male, 49.3% female) completed the TCI, the Big Five Questionnaire BFI-2, the Computer Self-efficacy Scale, the Utrecht Work Engagement Scale (UWES), and the Bergen Burnout Inventory (BBI-15). The results of the Exploratory Graphic Analysis (EGA) analysis confirmed the five-factor structure of the Slovak version of the TCI, with a confirmation of the stability of the identified dimensions using bootstrapEGA. A Confirmatory Factor Analysis (CFA) indicated that the scale comprised the five subscales found in the English-version of the scale and that no items had been removed. The reliability Cronbach’s alpha for all factors was above 0.80 and showed the measure had appropriate internal consistency. The convergent validity of the TCI was supported by strong positive relationships with burnout and negative relationships with engagement. Negative relationships of perceived technostressors with the personal dimension of computer self-efficacy were found. Neuroticism exhibited positive relationships with techno-overload, techno-invasion, techno-complexity, and techno-insecurity; the remaining four personality dimensions of the Big Five model showed negative relationships with these technostressors. Techno-uncertainty does not show any correlation with personality. The high level of reliability and stability of the dimensions of the TCI scale suggests that the TCI can be used for scientific research and for the measurement and prevention of technological stress in work settings where employees use ICT for their work goals and tasks. FTS - Technostress Creators Inventory BFI - BFI-2, Big Five Questionnaire UWES - Dutch Utrecht Work Engagement Scale V - Bergen Burnout Inventory CSE - Computer Self-Efficacy

Steps to reproduce

Exploratory graph analysis In the first stage, an EGA was performed to determine the functional factor structure. A subsample of 364 participants was randomly selected from the 692 participants using the randomization function of SPSS. An EGA was then conducted on this subsample to determine the factor structure of the twenty-three questionnaire items to measure any facilitators of technostress. We used R Studio 2022.02.2 for this statistical analysis. We then used the Bootstrap Exploratory Graph Analysis (bootEGA) to estimate the stability of the dimensions identified by the EGA. Confirmatory factor analysis Using JASP, a CFA was conducted on the remaining 364 participants in the total sample to determine whether the factor structure required an adjustment. The CFA was used to confirm the hypothesized model using a separate sample of participants. The CFA is a form of structural equation modelling used to determine the goodness of fit between the hypothesized factor structure and the sample data. In the CFA, the factor loading of one indicator variable for each latent variable was fixed at 1.0. This fixed the metric for each latent variable. Correlations were allowed between pairs of latent variables in the model because theoretically different types of causes of technostress should be correlated. Correlations between the other variables were fixed at 0.0. Our assessment of the model fit was guided by five indices: the chi-square/df ratio (χ2/df), the non-normalized fit index (NNFI; Tucker & Lewis, 1973), the comparative fit index (CFI; Bentler, 1990), the incremental fit index (Bentler & Bonnet, 1980), and the root mean square error of approximation (RMSEA; Browne & Cudeck, 1993). Values of 0.90 for NNFI and IFI (Byrne, 2001) indicate a good agreement. Although the recommended values for the CFI range from 0.90 to 0.95, values close to or approaching 0.95 are generally accepted as an indicator of a good agreement (Hu & Bentler, 1999). The χ2/df ratio provides an estimate of model fit that is less sensitive to sample size than the conventional chi-square index. Although there is no clear guideline for the χ2/df ratio, values between 2 and 5 are recommended as appropriate cut-off values (Ullman, 2007). The RMSEA takes the approximation error in the population into account and tests how well the model would fit the population covariance matrix if it were available (Byrne, 2001). Values lower than 0.08 indicate a reasonable fit (Browne & Cudeck, 1993) and values lower than 0.05 indicate a good fit (Stieger, 1990). Values higher than 1.0 should lead to model rejection (Browne & Cudeck, 1993). Reliability and convergent validity analysis Internal consistency was assessed by calculating the Cronbach's alpha coefficients for individual facilitators of technostress. To assess convergent validity, we then correlated the causes of technostress with other constructs which they should theoretically be related to.

Institutions

Categories

Funding

Licence