Figure 1
The model shows three exogenous variables on the left, one mediating variable in the center, and three endogenous variables on the right. On the far left, a hexagonal node labeled “Sector” is positioned at the top left. A rectangular indicator labeled “Homecare” points rightward to “Sector” with a loading value of 1.00. From “Sector”, a diagonal right-pointing arrow leads to the central latent variable “T B M”, with a path coefficient of 0.11. Below “Sector”, an oval node labeled “I L” is positioned at the center left. Four arrows extend leftward from “I L” to four vertically arranged rectangular indicator boxes labeled “var44x underscore 2021”, “var46x underscore 2021”, “var49x underscore 2021”, and “var52x underscore 2021”. The corresponding loading values are 0.97, 0.91, 0.91, and 0.75. A right-pointing dashed arrow from “I L” to “T B M” is labeled 0.38 double asterisk. At the bottom left, a hexagonal node labeled “Unit size” is shown. A rectangular indicator labeled “Unit underscore size underscore 2021” points rightward to “Unit size” with a loading value of 1.00. A diagonal upward right-pointing arrow from “Unit size” to “T B M” is labeled negative 0.11. The oval node labeled “T B M” is shown with an inner value of R-squared equals 0.18. From “T B M”, three arrows point upward to three rectangular indicator boxes labeled “var92 underscore 2023”, “var99 underscore 2023”, and “var100 underscore 2023”. The corresponding loading values are 0.86, 0.65, and 0.69. From “T B M”, three dashed arrows extend rightward to the endogenous variables. At the top right, a hexagonal node labeled “O U P” is shown with an inner value of R-squared equals 0.40. A dashed arrow from “T B M” to “O U P” is labeled 0.63 triple asterisk. From “O U P”, three arrows point rightward to indicator boxes labeled “var86 underscore 2023”, “var88 underscore 2023”, and “var89 underscore 2023”, each with a loading value of 0.36. In the middle right, an oval node labeled “T I” is shown with an inner value of R-squared equals 0.31. A dashed arrow from “T B M” to “T I” is labeled negative 0.56 triple asterisk. From “T I”, three arrows point rightward to indicator boxes labeled “Var117rec underscore 2023”, “Var119rec underscore 2023”, and “Var121rec underscore 2023”. The corresponding loading values are 0.93, 0.94, and 0.87. At the bottom right, an oval node labeled “O C B” is shown with an inner value of R-squared equals 0.37. A dashed arrow from “T B M” to “O C B” is labeled 0.61 triple asterisk. From “O C B”, three arrows point rightward to indicator boxes labeled “var75 subscript 2023”, “var76 subscript 2023”, and “var78 subscript 2023”. The corresponding loading values are 0.80, 0.92, and 0.78.Main results from structural model (N = 52). Note: Standardized path coefficients. **p < 0.01; ***p < 0.001. Latent variables measured with Mode A consistent. HTMT max = 0.59; SRMR = 0.10 < HI95 = 0.14. Source: Author’s own work