Academia.eduAcademia.edu

Figure 1 – uploaded by Euripidis Loukis

See full PDFdownloadDownload figure
In order to test the abovementioned research hypotheses H1—H9, we estimated the following multivariate model: In the Appendix | are shown the exact definitions of all these variables, whic correspond to the e-Business Survey questions we have used. The depender variable Prop_Cloud is a measure of the propensity to adopt CC, which initially wa a three-level variable assessing how relevant the respondent firm finds CC, havin as possible values: very relevant, partly relevant, or not relevant; but as we can se in Table 3 the relative frequency of the first value is so small (1.3%) that we had t merge the two levels ‘very relevant’ and ‘partly relevant’ to one (having the relativ frequency of 12.4%). So the Prop_Cloud variable we use in the model estimations 1 a binary one, which takes the value | if a firm reports relevance (even partial) of CC for their activities, and 0 if it replies that CC is not relevant for them. Ac indenendent variahles we have inclided cix hinaryv ones (1/0) exictence ao

Table 3 In order to test the abovementioned research hypotheses H1—H9, we estimated the following multivariate model: In the Appendix | are shown the exact definitions of all these variables, whic correspond to the e-Business Survey questions we have used. The depender variable Prop_Cloud is a measure of the propensity to adopt CC, which initially wa a three-level variable assessing how relevant the respondent firm finds CC, havin as possible values: very relevant, partly relevant, or not relevant; but as we can se in Table 3 the relative frequency of the first value is so small (1.3%) that we had t merge the two levels ‘very relevant’ and ‘partly relevant’ to one (having the relativ frequency of 12.4%). So the Prop_Cloud variable we use in the model estimations 1 a binary one, which takes the value | if a firm reports relevance (even partial) of CC for their activities, and 0 if it replies that CC is not relevant for them. Ac indenendent variahles we have inclided cix hinaryv ones (1/0) exictence ao

Related Figures (4)

of previous experience of ICT outsourcing, and firm’s size (see Sects. 3.2, 3.3, 3.4, 3.5, 3.6); the effects of the first five of them on CC adoption have not been investigated in previous CC adoption empirical research. Finally, we identified two characteristics of firm’s external environment that we expect to affect its propensity for CC adoption: the intensity of the price competition and the intensity of quality competition the firm faces (see Sect. 3.7). The above thoughts led to the development of our research model shown in Fig. |. The corresponding research hypotheses H1—H9 are presented in more detail in the next Sects. 3.1, 3.2, 3.3, 3.4, 3.5, 3.6 and 3.7. 3.1 ICT infrastructure sophistication Our first research hypothesis concerns the effect of firm’s ICT infrastructure sophistication on its propensity to adopt CC. In previous CC literature (Marstor et al. 2011; Venders and Whitley 2012; Miiller et al. 2015) we find argument: concerning the high usefulness and value of CC for firms not having sop histicatec ICT infrastructures, as it enables them to easily and rapidly gain access to more ICT capabilities and functionalities, at a low cost without the need for ICT investments So we would expect that firms having weak ICT infrastructures wit h limitec capabilities and sophistication might have a stronger motivation to adopt CC thar the ones having highly sophisticated ICT infrastructures. However, there exist alsc arguments pointing to the opposite direction: CC can be quite useful and valuable also for firms having highly sophisticated ICT infrastructures, as it enables them tc
of previous experience of ICT outsourcing, and firm’s size (see Sects. 3.2, 3.3, 3.4, 3.5, 3.6); the effects of the first five of them on CC adoption have not been investigated in previous CC adoption empirical research. Finally, we identified two characteristics of firm’s external environment that we expect to affect its propensity for CC adoption: the intensity of the price competition and the intensity of quality competition the firm faces (see Sect. 3.7). The above thoughts led to the development of our research model shown in Fig. |. The corresponding research hypotheses H1—H9 are presented in more detail in the next Sects. 3.1, 3.2, 3.3, 3.4, 3.5, 3.6 and 3.7. 3.1 ICT infrastructure sophistication Our first research hypothesis concerns the effect of firm’s ICT infrastructure sophistication on its propensity to adopt CC. In previous CC literature (Marstor et al. 2011; Venders and Whitley 2012; Miiller et al. 2015) we find argument: concerning the high usefulness and value of CC for firms not having sop histicatec ICT infrastructures, as it enables them to easily and rapidly gain access to more ICT capabilities and functionalities, at a low cost without the need for ICT investments So we would expect that firms having weak ICT infrastructures wit h limitec capabilities and sophistication might have a stronger motivation to adopt CC thar the ones having highly sophisticated ICT infrastructures. However, there exist alsc arguments pointing to the opposite direction: CC can be quite useful and valuable also for firms having highly sophisticated ICT infrastructures, as it enables them tc
Table 3 Relative frequencies and correlations Statistically significant correlations are shown in bold
Table 3 Relative frequencies and correlations Statistically significant correlations are shown in bold
Also, the ICT investment reduction strategy variable (ICT_Invest_Red) has a positive and statistically significant effect, and this provides support for research hypothesis H2. Both innovation variables (Prodserv_Inn and Proc_Inn) show positive but statistically insignificant (at the 10%-test level) effects, but their respective standard errors (not shown here) indicate nearness to statistical significance at the 10%-test level. So research hypothesis H3 is not confirmed according to standard statistical criteria, but it is not far away from confirmation. The variables for ICT-personnel (ICT_Pers) and for ICT outsourcing (ICT_Outs) also have positive and statistically significant coefficients. On the contrary, the variable for ICT skills and—rather unexpectedly—the variables for firm size show no statistically significant effect on the CC adoption propensity. Therefore hypotheses H4 and H6 receive empirical support, contrary to the Hypotheses H5 and H7, which are not confirmed by our estimates.
Also, the ICT investment reduction strategy variable (ICT_Invest_Red) has a positive and statistically significant effect, and this provides support for research hypothesis H2. Both innovation variables (Prodserv_Inn and Proc_Inn) show positive but statistically insignificant (at the 10%-test level) effects, but their respective standard errors (not shown here) indicate nearness to statistical significance at the 10%-test level. So research hypothesis H3 is not confirmed according to standard statistical criteria, but it is not far away from confirmation. The variables for ICT-personnel (ICT_Pers) and for ICT outsourcing (ICT_Outs) also have positive and statistically significant coefficients. On the contrary, the variable for ICT skills and—rather unexpectedly—the variables for firm size show no statistically significant effect on the CC adoption propensity. Therefore hypotheses H4 and H6 receive empirical support, contrary to the Hypotheses H5 and H7, which are not confirmed by our estimates.
Table 5 Summary of findings
Table 5 Summary of findings
Related topics:
BusinessInformation SystemsComputer ScienceEconomicsIndustrial OrganizationInformation and Communications Technology
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved