7 The Effects of Task-Technology Fit (TTF) as Moderation on Use and User Performance
7.2 Task-Technology Fit (TTF) as Moderation
7.5.3 Cross-Product Interaction (Cells 9 to 12)
The path coefficients, t values, p values, significance levels, and confidence intervals of the structural path models estimated to test the interactions in cells 9 to 12 (Figure 7.3), by evaluating the moderating effects of mHealth technology characteristics on the relationship between mobility in CHW tasks and use and user performance, are shown in Table 7.3.
Table 7.3. Structural Path Model Results: Cross-Product Interaction (Cells 9 - 12)
Results in Table 7.3 indicate that in addition to the negative on-diagonal effects observed, two diagonal interactions were significant for CHW performance, and one off-diagonal interaction was significant for use. The first significant off-off-diagonal interaction finding was that interdependence support moderates the effect of mobility in tasks on use (path coefficient = 0.311, t = 1.65, p < 0.10) and user performance (path coefficient = -0.279, t = 1.65, p < 0.10). These moderating effects are however not consistent with Proposition 3 (P3) and Proposition 4 (P4), as they are not in the expected direction.
The structural path model estimated to test TTF moderation effects of interacting mobility and interdependence support is depicted in Figure 7.4.
Figure 7.4. Path Model: Mobility Interdependence Support Fit
The moderating effect of the technology’s interdependence support on the links between mobility task characteristics and tool use, and CHW performance is illustrated in Figures 7.5 and 7.6.
Figure 7.5. Mobility Interdependence Support Fit: Interaction Effects on Use
Figure 7.5 shows that the effect of mobility of tasks on the use of the tool depends on whether the tool has functionality that integrates data from others. It shows that mobility of tasks increases use of the tool when functionality is low, but decreases use when functionality is high.
Figure 7.6. Mobility Interdependence Support Fit: Interaction Effects on User Performance
Similarly, Figure 7.6 shows that as the mobility of tasks increases, performance will decrease with a tool with high support for data integration, but increase with a tool with low support for data integration. When a CHW moves a short distance from location to location, they are less likely to depend on the use of the mHealth tool unless they have a high need to access integrated data functionality of the tool. However, it is very difficult to improve the performance of CHWs who move a lot from location to location, as their use of the tool and their performance does not depend as much on whether it has data integration capabilities.
The second significant off-diagonal interaction was that information dependency support of the tool moderates the effect of mobility in task characteristics on user performance (path coefficient = -0.360, t = 2.29, p < 0.05). However, this moderating effect is not consistent with Proposition 4 (P4) since it is not in the expected direction.
The structural path model estimated to test TTF moderation effects of interacting mobility and information dependency support is depicted in Figure 7.7. The moderating effect is illustrated in Figure 7.8.
Figure 7.7. Path Model: Mobility Information Dependency Support Fit
Figure 7.8. Mobility Information Dependency Support Fit: Interaction Effects on User Performance
In Figure 7.8, a similar pattern exists where, as the mobility of tasks increases, performance will decrease with a tool with high support for information provision, but increase with a tool with low support for data provision. The performance of CHWs who move a lot from location to location does not depend on whether the mHealth tool has data provision capabilities.
The final significant interaction was an on-diagonal interaction of mobility with mobility support, in relation to use (path coefficient = -0.315, t = 4.71, p < 0.01). This on-diagonal interaction was discussed in Chapter 6.
7.5.4 Cross-Product Interaction (Cells 9 to 12)
The path coefficients, t values, p values, significance levels, and confidence intervals of the structural path models estimated to test the interactions in cells 13 to 16 (Figure 7.3), by evaluating the moderating effects of mHealth technology characteristics on the relationship between information dependency in CHW tasks and use and user performance, are shown in Table 7.4.
Table 7.4. Structural Path Model Results: Cross-Product Interaction (Cells 13 - 16)
Cell Interaction Effect Path
Coefficient
t p Sig Level 90% CI
13 Information Dependency x Time Criticality Support (ID * TCS)
Use
-0.057 0.40 0.69 NS [-0.29, 0.18]
Information Dependency x Time Criticality Support (ID * TCS) User Performance
16 Information Dependency x Information Dependency Support (ID * IDS) Use
-0.141 1.74 0.08 * [-0.27, -0.01]
R2 = 0.188, f 2 (ID * IDS) Use = 0.03, Q2 = 0.095, q 2 (ID * IDS) Use = 0.02 Information Dependency x Information Dependency Support
(ID * IDS) User Performance
0.253 2.80 0.01 ** [0.11, 0.40]
R2 = 0.189, f 2 (ID * IDS) User Performance = 0.07, Q2 = 0.117, q 2 (ID * IDS) User Performance = 0.04 NS = Not Significant. *p < 0.10. **p < 0.05. ***p < 0.01.
The significant but negative on-diagonal interaction between information dependency and information dependency support was previously discussed in Chapter 6.
Results in Table 7.4 indicate that only one off-diagonal interaction was significant, between information dependency of tasks and time criticality support of the mHealth tool, in relation to user performance (path coefficient = 0.271, t = 1.90, p < 0.10). Thus Proposition 4 (P4) was partially supported for information dependency.
The structural path model estimated to test TTF moderation effects of interacting information dependency and time criticality support is depicted in Figure 7.9.
Figure 7.9. Path Model Information Dependency Time Criticality Support Fit
Figure 7.10 shows the moderating effect of time criticality support on the link between information dependency of tasks and mHealth use and CHW performance.
Figure 7.10. Information Dependency Time Criticality Support Fit: Interaction Effects on User Performance
Figure 7.10 shows that the effect of information dependency of tasks on the performance of the user depends on whether the tool has functionality that enables time-critical responsiveness. When information dependency is high, performance increases with time criticality support but decreases with lack of support. This is likely because users who need access to information to complete their tasks are likely to perform better when that information is provided quickly.