Understanding how to interpret and manipulate large data sets is increasingly important today; however, this experience has been slow to trickle down to the typical undergraduate student. Here, we describe the implementation of a project-based learning experience that uses portable air sensors for the real-time measurement of carbon dioxide, ozone, and particulate matter, providing students with data sets that include thousands of measurements. These projects allow students to design their own research question and then independently carry out relevant air sampling. Data visualization was used as a tool to identify trends and relationships among analytes and was emphasized as a way to effectively present these findings to an audience. We have implemented the projects over two academic years with diverse student populations, from high school summer research students to senior undergraduate and graduate students in an environmental analytical chemistry course. The extent of mentoring, and the students' competencies in atmosphere chemistry and spreadsheet and graphing software, were not equivalent between these groups, but all were able to execute successful projects. The projects often focused on the indoor environment as concentrations are not well characterized and tend to vary with human activity, which lend themselves to the development of testable research questions. The paucity of data on indoor concentrations means that, in addition to a valuable experiential learning opportunity, students were engaged in a legitimate citizen science exercise as they set about characterizing a diverse set of indoor environments.