Psychometric assessment of organizational readiness scale for digital innovations and antecedents of organizational readiness

2.1 Organizational readiness theory

Organizational readiness theory is based on: change valence, change efficacy, and factors related to the context [2, 16]. Change valence refers to employees’ commitment to change. Change efficacy is how healthcare organizations perceive their ability to change [2, 17]. Contextual factors are related to the organizational atmosphere in the healthcare organization facilitating innovation-related change. Organizational readiness theory considers organizational readiness a shared psychological state where the employees feel committed to implementing organizational change and have the required collective skills and abilities [16]. Organizational readiness for change is a multi-level and multi-faceted construct, where readiness is evident depending on the nature of change at the individual, group, unit, department, or organizational level [16]. Thus, the extent these are crucial for health organizations to increase their e-health readiness refers to how healthcare organizations are prepared to anticipate change through programs related to ICT [1].

Moreover, health organizational readiness is related to three levels; macro-level (e.g., infrastructure, development policies), meso-level (organizational aspects), and micro-level (e.g., professionals) [18]. Digital readiness refers to a particular change where digital technologies play a crucial role. Yet, this may be interpreted based on context-specific organizations related to tech-specific barriers, attitudes, and capabilities [13]. The lack of readiness for digitalization in healthcare organizations is due to the weak health system at the national level, the lack of capability building, infrastructure for coordination, and increasing collaborations and commitment of policymakers [19], as well as the resistance of health professionals and their lack of readiness towards technology due to their lack of skills [4, 20]. In this regard, an investigation shows Kosovo health professionals show low readiness and trust in the digitalization of healthcare services [57].

2.2 Hypothesis development

Integration of information technology in healthcare organizations improves efficiency and the quality of services [21], which process is influenced by the motivation of human resources and capabilities to implement organizational change [3]. Therefore, the adaptability of human resources for technical and organizational change is essential in delivering digital innovation [2]. Studies show the importance of HR adaptation to technology [22–24], which is influenced by technological literacy [25], the personality of human resources [13, 67], the role of managers [13, 27] the ability of benefit from technology depends on the extent HR use their technology as well [21, 29]. Besides, HR that focuses on flexibility and cognitive readiness focuses on extended knowledge are found to be conducive factors for digital innovation [2, 14]. Subsequently, we propose the following hypotheses:

Organizational readiness for digital innovation is influenced by resource readiness, IT readiness, combining existing technologies with new ones, and planning for new telehealth and e-health. Resource readiness refers to the organization’s flexibility to facilitate its digital innovation needs [14]. At the same time, IT readiness refers to the ability of healthcare organizations to foster digital innovation through IT infrastructure [2]. As for combining existing and new technologies, we define the increase in the speed and efficiency of healthcare organizations’ readiness for digital innovation. In addition, planning for new telehealth and e-health, ranging from simple to sophisticated bio-monitoring systems, refers to providing an additional information source about patients and better services [28].

Studies on resource readiness show relationships between resource readiness and organizational readiness for digital innovations. Lokuge [2, 14] argue that resource readiness influences organizational readiness for digital innovations. In another study, Hussain [17] supported the hypothesis that resource readiness affects digital financial innovation through its change efficacy. While studies on IT readiness suggest that IT readiness positively influences digital financial innovation [17], and implementing hardware and software positively influences e-health systems [11]. Other studies show that the barrier to IT readiness is the IT infrastructure [18, 31], while other studies show that the IT skills of health professionals may hinder IT readiness [11, 18], which then influences organizational readiness for digital innovation.

Planning new telehealth and e-health (PNTH) is crucial for health organizations which may lead to organizational readiness for digital innovation. Telehealth is essential to increase the monitoring of patients using devices that capture vital signs and other related healthcare information transmitted via technology [28]. Scholars maintain that the decision of managers to plan and adopt new telehealth technologies and e-health in which the adaptation of human resources is crucially essential in implementing telehealth [28]. This includes well-trained employees in healthcare information technology in increasing the ability to benefit from it [21]. The extent that human resources adapt to technology and the role that managers play in health organizations influence the extent that organization combines existing technologies with new ones. For example, if employees in health organizations adopt old technology, managers need to build a strategy to prepare the readiness of human resources for this. Thus, shared beliefs on digital readiness in management and employees are the source to guide and implement organizational change [13] by combining existing technologies with new ones that will affect organizational readiness for digital innovation. Based on the discussion above, we recommend the following hypothesis:

Partnership readiness, cultural readiness, and effective implementation of innovations are also among the factors that affect the organizational readiness of healthcare organizations for digital innovation. Partnership readiness is the affiliation of stakeholders outside the organization who support digital innovation [2]. Studies acknowledge the importance of building partnerships and influencing organizational readiness for digital innovation [30, 32]. For example, Hussain’s [17] study shows that digital financial innovation is affected positively by partnership readiness. However, according to Lokuge [14], creating and preserving innovation readies are challenging for health organizations and external partners. To respond to such a challenge, before focusing on digital innovation, it is critical to ensure the acceptance of patterns and their engagement in ensuring e-health readiness [33]. In addition, studies show that mutual trust between client consultants helps improve project outputs and the knowledge transfer process [34]. Similarly, another study shows how organizations improve their knowledge pool through knowledge transfer using clients to improve IT performance and build organizational readiness [32].

Cultural readiness can be defined as health organizations’ advantages on their central values that lead to digital innovation [2, 65]. Scholars argue that organizational readiness is higher among organizations when all members have the culture to focus on implementing organizational change and the confidence to do so [3, 16]. It is related to the extent to which organization members share beliefs on digital empowerment and involvement, which is influenced by how managers act as role models to lead the organization toward digital readiness [13]. Thus, concerning change built within the organization and collective involvement, coordinated behavior change is found to have a crucial role [16]. The extent managers have decentralized the decision-making influences organizational readiness for digital innovation [14]. Studies show that cultural readiness positively influences digital financial innovation [17].

There is a significant correlation between organizational culture and digital capabilities and innovation [35], and organizational culture is crucial to developing the absorptive capacity that influences the implementation of new technologies [36]. The organizational readiness for digital innovation depends on the effectiveness of health organizations in implementing innovations, as 90% of ideas never come into implementation face due to the lack of readiness [14]. Increasing the effectiveness of organizational readiness is crucial that health organizations managers understand contextual factors within the organization [37], such as; resources, human resources, culture, decision-making, communication, and reward system [14]. Also, to increase healthcare organizations’ effectiveness, it is crucial to focus on developing knowledge management, coordinating and collaborating at inter-functional and inter-organizational levels, and building innovation infrastructure [38]. Based on the discussion above, we recommend the following hypothesis:

4.1 Demographic characteristics of participants

As a result of the research, frequency, and percentage analysis were applied to the data, and findings related to the demographic characteristics of the participants were obtained. Table 1 shows that most participants were men (63.3). Approximately two-thirds of the participants declared they had completed at least undergraduate education. While 50.6% of the participants have a monthly income of 500 Euros or less, 38.9% have an income of 500–1000 Euros. As for the working experience, 38.7% were 0–5 years, and 18.1% had 21 years or more of work experience. As for occupation, 16.1% are non-nurse health personnel, 19.5% are doctors, 51% are nurses, and 13.4% are other personnel. 28.7% are between the ages of 17–26.

The skewness and kurtosis coefficients of all expressions in the scale were examined to determine whether the variables included in the study showed a normal distribution. It was observed that the skewness and kurtosis coefficients of the data set were within the expected values (skewness <2, Kurtosis <7) [39, 40] (see Table 2). Accordingly, when the results of the normality test of the data are examined, it can be said that the data show a normal distribution.

Cronbach Alpha is a widely used method to measure the reliability and internal consistency of the construct [76, 78]. Özdamar [41] and Gliem [42] state the criterion values for the reliability coefficient as follows; the scale is unreliable in the range of 0.00 < α< 0.40, low reliable in the range of 0.41 < α< 0.60, average reliable in the range of 0.61 < α< 0.80, 0.81 < α< 1.00 range is quite reliable. The Cronbach Alpha values of the variables in the structure range from 0.831 to 0.903 (see Table 2). According to the relevant authors’ scale, the dimensions’ reliability is relatively high [77, 79]. According to the reliability analysis results, no statement was deleted because the level of reliability did not increase when the statements were deleted. Whether or not this instrument will yield the same findings each time it is used on the same subject in the same context is the subject of the dependability question. Science-related instruments are typically dependable if they yield consistent results regardless of how they are used or who administers them [75].

After the construct validity test, convergent and divergent (discriminant) validity should also be tested [43, 80]. To measure convergent validity, composite reliability (CR) and mean variance AVE (Average Variance Extracted) tests were performed [81]. For convergent validity, all CR values for the scale are expected to be greater than the AVE values and the AVE value to be greater than 0.5. In addition, while the CR value should exceed 0.70 for each construct, for convergent validity, the AVE value exceeding 0.50 is a condition for validity [55]. In the measurement model, the CR value was between 0.823 and 0.904, and the AVE value was between 0.562 and 0.724. These results revealed the reliability and internal consistency of the scale [78]. It is seen that the goodness of fit values provided for each structure are in the range of acceptable or excellent values.

To provide divergent validity, the condition that the square root of the AVE is greater than the correlation between the factors must be met [43]. When the correlation values of each structure with the other structures are examined, it is seen that the divergent validity condition is met for each structure, as it is lower than the said value (see Table 3). The convergent and discriminant validity of the scales used in the study were also evaluated. Convergent validity means that items under the same factor are related to each other and to the factor they belong to, while discriminant validity means that items under a factor are less related to other factors than their factor. Besides the CR coefficient, the AVE value should also be calculated for convergent validity. The AVE value expresses the average squares of the factor loads of the items in a factor. AVE > 0.500, CR > 0.700, and CR > AVE conditions must be met to be able to say that a factor has convergent validity [44, 45]. For discriminant validity, besides the AVE value, the correlation coefficients between the factors, the square of the maximum shared variance (MSV), and the mean of the average shared square variance (ASV) values should be calculated. The MSV value represents the square of the highest correlation coefficient between a factor and other factors. The ASV value represents the average square of the correlation coefficients between a factor and other factors. To say that the factors have discriminant validity, the conditions for the correlation coefficients between the factors MSV < AVE, ASV < AVE, and Square root AVE > must be met [45–47] (see Table 2).

Structural Equation Modeling (SEM) is a statistical technique to test and evaluate theoretical models of complex relationship relationships among variables. The assumptions of SEM are as follows: i) Linearity: The relationships between the variables in the model are assumed to be linear. In other words, the changes in one variable are assumed to produce a proportional change in another variable. Ii) Normality: The data distribution is assumed to be normal or approximately normal. This assumption is essential because SEM relies on maximum likelihood estimation, assuming the data are normally distributed. Iii) Independence: The observations used in SEM are assumed to be independent. That is, the values of one observation should not be related to the values of another observation in the sample [86]. Iv) Absence of multicollinearity: Multicollinearity is the presence of high correlations between two or more independent variables in a model. This assumption is important because it can cause problems estimating the model parameters. V) Sample size: The sample size used in SEM should be sufficient to produce stable estimates of the model parameters. The exact sample size required depends on several factors, including the complexity of the model and the number of variables in the model [87]. Vi) Missing data: The presence of missing data can be a problem for SEM. The assumption is that the data are missing at random, which means the missing values are unrelated to the values of other variables in the model. Vii) Model identification: The model should be identified, which means that the number of parameters in the model should be less than the number of observations. If the model is over-identified, it may not be possible to estimate the parameters of the model. Viii) Non-negative variance: The variances of the latent variables should be non-negative. This is because negative variances are not meaningful in SEM [88]. The estimation method followed in this article is the Maximum Likelihood (ML) model. One of the most widely used methods for estimation is the Maximum Likelihood (ML) estimation technique, which assumes multivariate normal data [89–91]. Structural equation modeling is also used to examine correlations between latent (unobserved) variables or constructs that are measured by (a number of) manifest (observable) variables or indicators [85]. Second-level confirmatory factor analysis was performed in this article to reveal latent relationships. SEM is presented in Fig. 1.

OPEN IN VIEWER

Fig. 1

Secondary confirmatory factor analysis.

The Confirmatory factor analysis (CFA) is used to test the suitability of the factors determined by explanatory factor analysis to the factor structures defined by the hypothesis [48, 49]. While testing the fit between the model and the data, some goodness-of-fit tests may be used, or all may be preferred [50]. The general indicators for model fit are CMIN/DF = x2/df, GFI, CFI, IFI, AGFI, RMR, and RMSEA. The reported values may vary according to the importance the researcher wants to draw attention to. Many goodness-of-fit measures are used in structural equation model evaluations in the literature. It isn’t easy to meet all of these criteria in a model. However, meeting some essential criteria is mandatory for model validity [51, 52]. Therefore, secondary confirmatory factor analysis was performed for the model, and the analysis results are presented in Fig. 1. The required goodness of fit values for the model and the obtained values are given in Table 4.

Although binary outcomes may also be employed, linear regression models frequently examine the relationship between a continuous outcome and independent variables. As arbitrary, biased outcome adjustments are typically unneeded, linear regression models are relatively resilient to breaches of the normality assumption in scenarios with high sample sizes. Instead, researchers should identify model miss-specifications, which may distort results regardless of sample size. They include outlier values, excessive leverage, heteroscedasticity, correlated errors, nonlinearity, and interactions [92]. The regression model is found to be significant (F = 354,042; p = 0.000). Simultaneously, H1-AHR (β= 0.076; t 3,561; p = 0.00), H2-COR (β= 0.072; t 3,158; p = 0.00), H3- PNTH (β= 0.109; t 4,523; p = 0.00), H5-ITR (β= 0.098; t 4,275; p = 0.00), H6-RR (β= 0.112; t 4,668; p = 0.00), H7-PR (β= 0.195; t = 8,57; p = 0.00), H8-CUR (β= 0.066; t 2,635; p = 0.01), H9-IIE (β= 0.354; t 15,169; p = 0.00), and variables have a positive and significant effect on of organizational readiness for digital innovation (β= 0.17; t = 4.662; p = 0.04). IOT (β= –0.063; t –2,6; p = 0.01) variable negatively and significantly affects organizational readiness for digital innovation. The model explains 72,5 percent of the changes in dependent variables organizational readiness for digital innovation (R2 = 0.725). Analysis results are presented in Table 5.

5.1 Managerial and policy implications

This research’s managerial and policy implications can be evaluated under five headings. Firstly, the tool was tested on organizational readiness for digital innovations, which can help organizations determine their readiness for new technological change [2]. The fact that the survey can be easily applied among all organizations and their stakeholders and takes a short time makes it a valuable tool for organizations. At the very least, instead of getting off to a bad start by applying the tool in readiness for digital innovations, organizations can identify their shortcomings through the tool and make the necessary improvements. Secondly, organizations could identify the readiness of different stakeholders through organizational readiness for digital innovation [2, 98]. In this way, they can prepare organizations and increase innovations’ success by creating mechanisms that will enable all stakeholders to act together. Thirdly, given the critical role of technological literacy, the challenges healthcare professionals face to adapt to technological changes, and the availability of resources, managers in healthcare organizations need to consider the antecedents that affect organizational readiness for digital innovation early in the process. Fourthly, managers need to motivate their employees by collaborating with professional healthcare organizations experiencing organizational change towards digital innovation, organizing training on technology, and focusing on changing the organizational culture in healthcare institutions. Finally, about policymakers, we recommend increasing investment in healthcare institutions, significantly increasing the capabilities of healthcare services to deploy new technologies, and providing financial support for healthcare professionals to develop their technology-related skills and abilities.

5.2 Limitations and future suggestions

The limitations of the study can be categorized into four main groups. First, Khoja [1] and Lokuge [2] adapted scale has been tested and validated in a country with relatively weak institutions in the transition economy context. Therefore, the scale must be tested and applied in developing and developed countries. Future research can test the validity of the respective scale by repeating similar analyses in developing and developed countries. Second, the cross-sectional data reflect the current situation but do not reveal long-term developments’ impact on organizational readiness for digital innovation. Therefore, future research can longitudinally examine the organizational readiness of healthcare organizations for digital innovation. Third, the Covid-19 pandemic-related factors may have influenced and directed the organizational readiness of health organizations for digital innovation [53]. For this reason, future research can investigate the issue by conducting in-depth interviews with employees and managers with pre-Covid 19 experience [63, 64] with case studies. Fourth, in the study, we did not touch on the managerial and institutional phenomena that affect the organizational readiness of health organizations for digital innovation. Therefore, future research can examine which organizational and institutional factors facilitate or complicate this process. Fifth, the data used in the research were collected from one university hospital. Therefore, future studies can expand the study’s sample by collecting data from different hospitals. Sixth, although the sample size consisted of 1236 questionnaires, studies with larger samples are likely to yield better results [67]. Finally, the organizational readiness structure for digital innovation in this research can be used to assess readiness from multiple stakeholder perspectives [2]. For example, this study collected data from doctors, nurses, health personnel other than nurses, and other personnel working in a hospital. However, this does not provide information on the readiness of different stakeholders. Therefore, future research can compare stakeholders’ views on organizational readiness for digital innovation by collecting data from samples, including healthcare providers, non-governmental organizations, and policymakers.