**5. Research results**

### **5.1. Model reliability and validity test**

In every dimension of the physicians and specialist assistant category, advertising value equivalency (AVE) values were all greater than the standard value of 0.5; CR values in all cases were greater than the standard value of 0.7. The overall model regression analysis *R*<sup>2</sup> ‐ value was 0.68, and Cronbach's α falls above 0.84, exceeding Fornell and Larcker's [12] rec‐ ommended high reliability index value (>0.7). This suggests that this study's structural model has good reliability and validity (**Table 1**).


**Table 1.** Reliability and validity.

### **5.2. Path analysis**

This study utilized partial least squares (PLSs) to evaluate the study. Using covariance struc‐ ture as a basis, structural equation modeling (SEM) recommended that sample size must be at least 200 or above or 10 times the parameter [13]. Second, using covariance structure as a basis, SEM's predicted value must comply with the norm; otherwise, a disassociation with reality effect will ensue [14, 15]. In addition, a covariance‐based SEM can only process reflec‐ tive indicators and is unable to process formative indicators. As this study was conducted with a sample size of 195, which did not meet the standard 10 times of the parameter, PLS was selected as the method for analysis (**Figure 1**, **Table 2**).

This study found that online clinical guideline system usage had a significant positive impact on the system's user satisfaction (*β* = 0.228, *p* < 0.05). Additionally, regarding the relationship between individual knowledge management behavior and individual net benefit, only a small portion had a significant positive impact. Knowledge internalization has a significant impact on individual net benefit (*β* = 0.183, *p* < 0.05), while knowledge socialization reveals no impact on individual net benefit (*β* = 0.05, *p* > 0.1). The online clinical guideline system's user satisfac‐ tion had a significant positive impact on both individual knowledge management behavior (user satisfaction→knowledge socialization, *β* = 0.191, *p* < 0.1; user satisfaction→knowledge internalization, *β* = 0.306, *p* <0.05) and individual net benefit (*β* = 0.549, *p* < 0.05), and indi‐ vidual knowledge management behavior had a partial significant positive impact on individ‐ ual net benefit (knowledge socialization→individual net benefit, *β* = 0.05, *p* > 0.1; knowledge internalization→individual net benefit, *β* = 0.183, *p* < 0.05).

**Figure 1.** PLS model.

**AVE CR** *R***<sup>2</sup> Cronbach's α**

‐

0.64 0.91 0.89

H3: An online clinical guideline system use has a significant positive impact on individual

H4: An online clinical guideline system's user satisfaction has a significant positive impact on

H5: An online clinical guideline system's user satisfaction has a significant positive impact on

H6: Individual knowledge management behavior has a significant positive impact on indi‐

This study investigated a medical center in Tainan, primarily focusing on some of the center's resident physicians, part‐time physicians, practicing physicians, medical interns, and special‐ ist assistants, and conducted questionnaire‐based surveys and investigation. It investigates the users, including 120 physicians and 80 physician assistants (PAs) who used the online clinical guidelines in a medical center of southern Taiwan. A total of 195 of the 200 question‐ naires surveyed including 111 physicians and 70 PA are effective, leading to 97.5% effective

In every dimension of the physicians and specialist assistant category, advertising value equivalency (AVE) values were all greater than the standard value of 0.5; CR values in all cases were greater than the standard value of 0.7. The overall model regression analysis *R*<sup>2</sup>

value was 0.68, and Cronbach's α falls above 0.84, exceeding Fornell and Larcker's [12] rec‐ ommended high reliability index value (>0.7). This suggests that this study's structural model

0.69 0.91 0.88

User intention 0.81 0.92 0.88 Actual system use 0.68 0.89 0.84 User satisfaction 0.80 0.96 0.95

Individual net benefit 0.76 0.96 0.68 0.96

Knowledge socialization

net benefit.

individual net benefit.

vidual net benefit.

response rate.

**5. Research results**

**5.1. Model reliability and validity test**

has good reliability and validity (**Table 1**).

**4.2. Research design**

individual knowledge management behavior.

210 Knowledge Management Strategies and Applications

Knowledge internalization

**Table 1.** Reliability and validity.


Sample size: 500 \**p* < 0.05; \*\**p* < 0.01.

**Table 2.** Structural method's path coefficient testing result.

Thus, hypotheses H1, H4, and H5 are supported. However, system use did not significantly influence knowledge internalization, user intention did not significantly influence individual net benefits, and individual knowledge socialization did not have a significant impact on indi‐ vidual net benefit. As a result, hypotheses H2, H3, and H6 are partially supported.
