An arXiv preprint reports that, across two studies, stronger trust expectations among smart-home users were associated with a stronger sense that their devices' data practices matched what they expected, a response the researchers call confirmation. That confirmation was associated with greater privacy-focused satisfaction, meaning a more favorable view of the balance between a device's benefits and disclosure risks. Satisfaction was also associated with stronger intentions to keep using the devices, but willingness to block data traffic did not show a consistent pattern between the studies.
Two ways of seeing the traffic
The first study started with 152 pre-survey respondents; 35 completed the paired post-survey used in the analysis. Participants viewed an IoT Inspector report generated from traffic from their own smart-home devices and browsed it for at least five minutes before completing the post-survey.
The second study recruited 132 U.S.-based adults through Prolific. They were at least 18 and used at least one smart speaker, smart TV or smart camera. Each saw a simulated, device-customized report and was randomly assigned to a LowAd or HighAd version with different advertising-related traffic. Manufacturer traffic was fixed at 842 KB uploaded and approximately 390 KB downloaded in both versions; advertising traffic was 90 KB uploaded and 110 KB downloaded in LowAd, versus 1.2 MB uploaded and 900 KB downloaded in HighAd.
The research was supported in part by Google through the CyberNYC initiative. The authors state that the sponsor had no influence on the study's design, analysis, interpretation or reporting.
The behavioral picture split
In the field study, trust expectations were positively associated with confirmation, and confirmation with privacy-focused satisfaction. Confirmation itself was not a significant direct predictor of either blocking or continuation intention. Satisfaction was associated with less blocking and a greater intention to continue using the device. An exploratory mediation analysis reported indirect associations through satisfaction for both outcomes.
The online experiment repeated the expectation-confirmation-satisfaction sequence: trust expectations were associated with confirmation, and confirmation with satisfaction. Confirmation was not a significant predictor of blocking or continuation there, while satisfaction was a significant predictor of continuation intention but not of blocking.
The experiment also pointed to other associations. Greater concern about data collection was associated with more willingness to block and less satisfaction, while greater self-efficacy, the perceived ability to use the technology, was associated with more satisfaction and a stronger intention to continue. Only the baseline interactions involving self-efficacy and continuation intention were significant across the advertising conditions; no significant interactions were reported for collection concerns, awareness or satisfaction.
Among the open-ended responses, consistency judgments were similar across conditions: 39 of 65 HighAd responses (60.0%) and 38 of 65 LowAd responses (58.5%) judged the report consistent. Distrust or inconsistency was descriptively more common in HighAd, at 12 of 65 responses (18.5%), versus 7 of 65 (10.8%) in LowAd. These descriptive figures do not establish that the advertising-traffic condition caused distrust.
What the results cannot settle
One important caveat is that Study 1 used single-item measures, which prevented assessment of internal reliability and discriminant validity; the authors therefore treated its path model as exploratory rather than confirmatory. The analysis also rested on the 35 paired completers rather than all 152 pre-survey respondents.
Study 2 used a simulated report, and its behavioral outcomes were stated willingness to block and intention to continue, not observed long-term device use. The two studies also differed in samples, settings and report conditions, so the reason blocking was associated with satisfaction in Study 1 but with collection concerns in Study 2 remains unresolved.
Paper data and sources
Original title: Evaluating Smart Home Device User Responses to their (Un)Confirmed Privacy Expectations
Authors: Tania Khatun, Mahdieh Sheikh Rezaei, Danny Yuxing Huang et al.
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-20
DOI: Not available
Original paper · Full text