Student ratings of the overall effectiveness of their instructor were recorded from 75 sections of college algebra freshman classes at California State University, Long Beach. These ratings were appraised in relation to seven independent variables, including number of students per class, number of rows per class, mean student grade, instructor, time of the class, frequency of instruction per week, and whether Web-based instruction was offered. A multiple regression analysis of the data revealed three significant findings: 1) ratings of individual instructors sizably differed; 2) mean student grades positively correlated to their ratings of the teachers; and 3) the number of rows per classroom was negatively associated with student ratings, that is, the classrooms with the most rows tended to yield the lowest student ratings. The last finding provides statistical support for the impression that the proximity to the teacher is important in the learning experience and that the design of the classroom matters.