Ahmed Ziad Benleulmi, Boumediene Ramdani
10 min
This study investigates the psychological drivers behind the public's intention to adopt fully autonomous vehicles (AVs). While previous research has focused heavily on the functional utility of such vehicles, this paper applies the Instrument-Symbolic-Affective (ISA) model—traditionally used for material possessions—to gain a more holistic understanding of user motives. By surveying 240 U.S. residents, the authors analyze which factors influence consumer readiness to adopt Level 5 (fully autonomous) vehicles.
The researchers employed structural equation modeling (SEM) to test a research model grounded in three categories:
Participants watched a brief introductory video about Level 5 automation before completing a survey designed to capture their perceptions across these specific psychological constructs. This approach allowed the team to integrate utilitarian perspectives with deeper emotional and identity-based factors.
The results demonstrate that the intention to use AVs is multifaceted, relying on a complex interplay of motives:
This study highlights that AV adoption is not merely a decision about transportation utility but is deeply rooted in personal identity, enjoyment, and emotional trust. Consequently, the authors suggest that developers and policymakers must look beyond safety metrics to incorporate fun and social status into their marketing and deployment strategies.
Sam: Yes. The authors acknowledge that these motives had appeared in earlier studies. Their framework combines practical and psychological perspectives, which they say urban mobility research has not always considered together. It also accommodates motives particularly relevant to autonomous vehicles, such as concern about surrendering control.
Alex: Make that concrete. What did respondents actually judge, and how did those judgments become evidence?
Sam: Participants first watched a short introductory video and saw a visual aid distinguishing automation levels. They then rated statements adapted from previous research. One practical-benefit item said that using autonomous cars would increase productivity. An intention item asked whether they planned to use the cars frequently once available.
Alex: So they measured several concepts through sets of statements, rather than relying on a single general question. How did they examine the relationships between those concepts?
Sam: They used structural equation modelling, which estimates relationships among concepts measured through survey items. Their version used partial least squares. They checked whether the items reliably measured their intended concepts, then tested how each proposed motive related to intention within the combined model.
Alex: What went into the practical side? Enjoyment sounds emotional, but your opening placed it alongside usefulness.
Sam: That is how this study classifies it. The instrumental group includes expected benefits, expected ease of use, and hedonic motivation—the enjoyment or pleasure anticipated from using the vehicle. The symbolic group includes willingness to try autonomous vehicles and perceived encouragement from important people in one’s social circle.
Alex: And the emotional group presumably captures whether someone feels comfortable handing over the journey?
Sam: It includes desire to retain control, trust, performance risk, and privacy risk. Performance risk means concern that the vehicle could malfunction or fail to deliver its promised benefits. Privacy risk concerns losing control over personal information. Trust involves willingness to become vulnerable to the vehicle.
Alex: With those motives tested together, which relationships survived? I’m especially interested in whether the control and privacy concerns actually predicted intentions.
Sam: The researchers tested nine proposed relationships. Expected benefits, enjoyment, willingness to try the technology, social influence, and trust were positively associated with intention. Concern about malfunction was negatively associated.
Alex: Those results cross all the categories. Did the remaining concerns disappear, or simply fail the study’s statistical test?
Sam: They failed the test in this sample and model. Six relationships were statistically significant. Expected ease of use, desire to retain control, and privacy risk were not. That does not establish that those issues are irrelevant to autonomous-vehicle adoption generally.
Alex: Statistical significance does not tell us whether these are substantial predictors. How large were the relationships?
Sam: The authors classify all the significant effect sizes as small. Enjoyment had the largest reported effect size, at point zero nine. Their benchmark for a medium effect was point one five. So the strongest predictor here was still below their medium-effect benchmark.
Alex: That puts the opening in perspective. Enjoyment stood out within this model, but it was not an overwhelming explanation of willingness.
Sam: That is the right scope. The result supports considering enjoyment alongside usefulness, rather than claiming enjoyment determines adoption. The authors suggest expected ease might have been insignificant because participants lacked experience with autonomous vehicles. That is a possible explanation, not something this design established.
Alex: Experience seems central here. When were the opinions collected, and how representative were the respondents?
Sam: The online survey began in September 2017 and ended in January 2018. It used convenience sampling, not a representative population sample. Most participants had driving experience, and most owned a car or had owned one previously.
Alex: So these were largely people familiar with conventional driving, answering about an unfamiliar form of transport. What is the most consequential limitation?
Sam: Their lack of actual autonomous-vehicle experience. The authors explicitly identify that limitation and suggest studying people who have taken test rides. A video can introduce the concept, but these answers still describe anticipated use rather than responses grounded in actual journeys.
Alex: There is also the direction of the relationship. Could someone already keen on driverless cars simply expect them to be more enjoyable and trustworthy?
Sam: The study cannot resolve that. Its data are cross-sectional: motives and intentions were measured at one point. The authors acknowledge the causal limitation. We should read the results as associations, not evidence that increasing trust or enjoyment will cause greater use.
Alex: Yet the paper offers recommendations for providers and government campaigns. How should listeners separate those recommendations from what was tested?
Sam: The authors propose improving perceived performance and enjoyment, building trust, and lowering perceived malfunction risk. They also suggest targeting people willing to try autonomous vehicles. Those proposals follow from the associations; the study did not test whether a campaign or design change increased adoption.
Alex: So the framework is useful for choosing what to investigate, not a validated recipe for changing behaviour. Who should read the full document, and where should they start?
Sam: Researchers studying transport adoption or technology acceptance should start with Research model and hypotheses. That shows how the authors classify the motives and connect them to earlier work. Then read Structural model assessment and the limitations, keeping the small effects and hypothetical use in view.
Alex: And what should developers, policymakers, or researchers outside this area carry away?
Sam: Developers and policymakers should read the recommendations alongside the limitations, not separately. For everyone else, remember this: usefulness is only part of stated willingness to use driverless cars; enjoyment, social motives, trust, and malfunction worries belong in the same conversation.
Alex: Keep that distinction between willingness and use in mind.