Adoption of digital health apps
Healthcare / mHealth
An extension of the Technology Acceptance Model (TAM) to digital health apps. Alongside the classic TAM variables Perceived Ease of Use and Perceived Usefulness, we model Health Consciousness as a domain-specific motivator and Perceived Risk as a negative privacy path — a rarely so clearly delimited f² effect.
- N
- 380
- LVs
- 6
- Effect
- PR → BI negative
Cloning requires a free OpenPLS account.
Theoretical background
Davis (1989) established TAM, which to this day remains an extremely popular acceptance model with two core determinants: perceptions of usefulness (PU) and ease of use (PEOU). These shape attitude, which in turn drives intention.
For mHealth applications, the classic TAM is not enough: Sun et al. (2013) integrated Health Consciousness (positive) and Perceived Risk (negative) as domain-specific extensions. Privacy is particularly relevant in healthcare because sensitive data is shared.
The model is methodologically interesting because it contains a clearly negative path (PR → BI). Such effects are valuable in PLS-SEM teaching: they show that sign, significance, and f² effect sizes need to be interpreted separately.
Structural model
Classic TAM chain (PEOU → PU → Attitude → Intention) plus two domain-specific antecedents: Health Consciousness positive, Perceived Risk negative.
Perceived Ease of Use
Perceived ease of operating the app.
Perceived Usefulness
Expected health-related usefulness.
Health Consciousness
Health awareness and personal responsibility.
Perceived Risk
Perceived privacy and data-misuse risk.
Attitude
General attitude toward the app.
Behavioral Intention
Intention to use the app regularly.
Hypotheses
| H1 | PEOU → PU | + | Ease of use raises perceived usefulness. |
| H2 | PU → ATT | + | Usefulness raises attitude toward the app. |
| H3 | HC → ATT | + | Health consciousness strengthens the positive attitude. |
| H4 | PR → ATT | - | Perceived data risk undermines attitude. |
| H5 | PU → BI | + | Usefulness directly raises usage intention. |
| H6 | ATT → BI | + | A positive attitude leads to usage intention. |
Methodology & data
N = 380 synthetic responses on 20 reflective indicators (seven-point Likert). Effect sizes are calibrated so the negative PR path is significant but smaller than the positive drivers — typical for real mHealth studies (Sun et al., 2013; Cocosila & Archer, 2010).
Expected results
- R²(BI) ≈ 0.52
- The five determinants explain about 52 % of variance in intention — solid for an extended TAM.
- PR → BI ≈ −0.18
- A clearly negative but moderate path. A teaching example for negative effects: report sign, significance, and effect size separately.
- f²(PR)
- The f² effect size for Perceived Risk typically lies in the small-to-medium range (≈ 0.06). Ideal for discussing Cohen thresholds.
Reproduce in 60 seconds
- 1
Clone the project
A single click in your OpenPLS workspace creates a fully editable copy: model, indicators, and dataset, all linked and ready to go.
- 2
Run the computation
OpenPLS solves outer weights, path coefficients, R², HTMT, SRMR, and bootstrap confidence intervals in a few seconds.
- 3
Compare against expected metrics
The key metrics documented below come from the published original. Your computed values should fall inside the bootstrap 95 % CIs.
References
- Davis, F. D. (1989). Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information Technology. MIS Quarterly, 13(3), 319–340. doi.org/10.2307/249008
- Sun, Y., Wang, N., Guo, X., & Peng, Z. (2013). Understanding the acceptance of mobile health services: A comparison and integration of alternative models. Journal of Electronic Commerce Research, 14(2), 183–200. www.jecr.org/node/313
- Cocosila, M., & Archer, N. (2010). Adoption of mobile ICT for health promotion: an empirical investigation. Electronic Markets, 20(3), 241–250. doi.org/10.1007/s12525-010-0042-y