The School of Architecture's faculty posed the methodological problem: to construct a brief scale to evaluate instruction. Guidelines were: (a) brevity, i.e., 10 items or fewer, (b ) high reliability, (c) minimum word redundancy (single psychological elements), and (d ) applicability to different levels . of technical and artistic instruction. Subjects were 3227 students enrolled in 196 courses offered over seven consecutive semesters (Fall 1978 through Fall 1981). The architecture evaluation scale had 28 5-point ratings (scaled strongly disagree vs strongly agree) applicable to design theory and studio experiences. Scores were intercorrelated and yielded a positive R matrix; average r = .43 (9 < .001) for 378 r-pairs. A single eigenvalue was greater than 1.0; matrix was factored for one iterative principal axis component without rotation. The 28 items ranged in loading from .86 to .27, and the factor mas labeled, Instructor/Course Impact. The first 10 ordered item markers and loadings were: (a) all things considered, the instructor did a good job (.86); (b ) all things considered, the course was worthwhile ( .82 ) ; ( c ) instructor created a desire to learn through interesting and challenging development of subject or project (.79); ( d ) instructor prepared and organized the course effectively (.79); (e) instructor presented class material clearly and concisely (.77) ; ( f ) instructor presented the course with enthusiasm (.74) ; (g ) instructor used class time effectively and constructively ( .73 ) ; ( h ) instructor demonstrated genuine interest in students and their progress (.72); ( i ) instructor was innovative; open to new ideas, techniques and approaches (.71); and ( j ) instructor has good rapport with students; easy to talk to (.69). Average correlation between items and corrected split-half reliability were computed for scales of different lengths: 2 items (.72, .84); 4 items (.67, .89) ; 6 items (.66, .91) ; 8 items (.64, .92) ; 10 items (.60, 9 4 ) . Adding items to the single-factor scale was associated with a decrease in the average r between items and an incredse in S-B reliability. A minimum required reliability of .90 suggests that the scale have at least six items ( 2 ) . During the 198283 academic year, the School of Architecture will use these marker items and expand the evaluation of design studio experiences ( 1 ) .
Read full abstract