Over the last decade the algebra-based first course in statistics has received a tremendous amount of interest by both teachers in the classroom and educational researchers. As a result of improved teaching in the first course, as well as demands for a more quantitatively literate citizenry, increasing numbers of students from diverse backgrounds are enrolling in a second course in statistics. As such, our goal has been to design a second course that will be of interest and relevance to students with only an algebra prerequisite. In this paper we present concrete examples, based on genuine research studies, that teachers can use in their classrooms to make multivariable thinking accessible to these students, through data visualization and a focus on the statistical investigation process, rather than mathematical formulas. Instead of teaching separate analysis methods of regression and anova, the ideas of explaining variation, building models, adjusting for confounding variables, and interaction are presented in a unified framework, in order to build conceptual and intuitive understanding.