This study sheds light on the travel behavior of drivers when they encounter fabricated messages in work zones. Using the response of 4302 participants to a stated preference survey, researchers develop a multivariate ordered response model and a structural equation model to study speed change and distraction response behavior. The results of their models for fabricated announcements signify that drivers normally follow the announcement and are affected likewise. The selected socioeconomic and attitudinal variables are shown to have mixed impacts in our speed and distraction models. Some variables are statistically significant for each model, while other variables are only statistically significant for one of the models. For instance, drivers that have seen a fabricated announcement before are less likely to speed up when encountering the message, while drivers who rely on technology for their daily travels are more likely to be distracted. Higher income is shown in our models to signify undesirable behaviors: speeding up and being distracted. Contrastingly, female drivers are less likely to do nothing or be distracted by the announcement. The findings, taken together, have implications for researchers and practitioners. First, they illustrate how cyber attacks can destabilize traffic in work zone and put the life of work zone crew members in jeopardy. Second, they explain the degree of compliance with compromised dynamic message signs.