Is public engagement with science deliberative and evidence-based? The public is often perceived as underprepared to use data and susceptible to partisan and emotional manipulation. Consequently, educational efforts focus on the ability to identify reliable information. We posit that effective engagement with science goes beyond this and hinges on data literacy. We leveraged the unique circumstances of COVID-19, where diverse people inundated with pandemic-related data representations in the media needed to make consequential decisions, to examine whether people use data and what factors affect such use. In a survey of a representative Israeli adult sample, participants reported their information habits and beliefs before and during COVID-19. On being presented with graphs and datasets, they answered data literacy and COVID-19-related functional reasoning assessments (e.g., would you travel abroad?). Data literacy distinguished those who incorporated data from those who did not. Yet, participants incorporated moral, social, and economic considerations at all data literacy levels, suggesting that people may be deliberative even when they do not attend to quantified data. Moreover, participants' trust in science and data interpretation competence were key factors mediating the relationship between self-efficacy in data interpretation skills and the incorporation of data in reasoning. The findings extend beyond COVID-19 to a broader understanding of the factors influencing public engagement with quantitative representations. Rather than focusing solely on remediating data interpretation, we suggest that educative efforts work on multiple fronts and that cultivating trust in science is key to a broader, more deliberative engagement with science.