Table 3. Multilevel regression models with moderators for evaluation of conversational agents.
| Models | B (SE) | 95% CI | Omnibus test QM | P value |
|---|---|---|---|---|
| Model 1: Publication year | 0.36 | .55 | ||
| Intercept | 38.75 (63.72) | −86.15 to 163.64 | ||
| Publication year | −0.02 (0.03) | −0.08 to 0.04 | ||
| Model 2: Age | 0.69 | .41 | ||
| Intercept | 0.80 (0.43) | −0.06 to 1.65 | ||
| Participants’ average age | −0.01 (0.01) | −0.04 to 0.02 | ||
| Model 3: Type of interaction (reference: interacting with designed CAsa) | 0.28 | .87 | ||
| Intercept | 0.44 (0.14) | 0.17 to 0.71 | ||
| Viewing prerecorded materials | −0.07 (0.30) | −0.67 to 0.52 | ||
| Wizard of OZ | 0.13 (0.31) | −0.48 to 0.73 | ||
| Model 4: Health topic (reference: mental health) | 3.08 | .38 | ||
| Intercept | 0.59 (0.16) | 0.28 to 0.90 | ||
| Teletriage | 0.16 (0.48) | −0.77 to 1.10 | ||
| Promote health behavior | −0.46 (0.31) | −1.06 to 0.14 | ||
| Provide medical information and advice | −0.25 (0.25) | −0.73 to 0.23 | ||
| Model 5: Publication outlet (reference: journal publication) | 0.19 | .91 | ||
| Intercept | 0.43 (0.15) | 0.14 to 0.72 | ||
| Conference proceeding | 0.00 (0.25) | −0.49 to 0.49 | ||
| Thesis and dissertation | 0.18 (0.41) | −0.63 to 0.98 | ||
a
CA: conversational agent.