Second grade students (N = 193) in three urban elementary schools in Pennsylvania received sightsinging instruction for 15 sessions of general music classes, each 25 minutes in length. The children read notation and sang four-note tonal patterns, with one new pattern presented each session and all previously learned patterns practiced at the beginning and end of each session. During each session a new song with an activity was learned by rote and sung. In the 16th session all 15 patterns were reviewed. Independent variables included instructional treatment, school, and sex. Variations in instructional treatment included singing the patterns with solfege or loo, and singing a related or unrelated song. Related songs used the pattern as the first four notes of the song. The following four treatment conditions were randomly assigned to classrooms: (1) solfege/related song; (2) solfege/unrelated song; (3) loo/related song; (4) loo/unrelated song. The children were individually tested at three points in time on their ability to read and sing the patterns: a pretest prior to instruction; a posttest after the 16 sessions; and a retention test that followed an additional eight weeks of no sightsinging instruction. For each test, the children sightsang patterns learned in class (familiar patterns) and 10 patterns which had not been practiced (unfamiliar patterns) to assess the ability to transfer learning. Sightsinging performance was evaluated for pitch and contour accuracy. Also examined were correlations of sightsinging scores with pitch matching, tonal discrimination using the IMMA, school ability using the Otis-Lennon Test, and reading fluency using Dynamic Indicators of Basic Early Literacy Skills (DIBELS).

The instruction resulted in a significant improvement in sightsinging achievement. Nonsignificant differences from posttest to retention test for all treatment groups indicated that students' performance in sightsinging remained statistically stable. Significant improvement in singing unfamiliar patterns indicated that skills transferred. Treatment effectiveness differed according to pattern type. For familiar patterns, contour accuracy scores were significantly higher in the solfege condition. For unfamiliar patterns, however, contour accuracy scores were significantly higher in the loo condition. Learning related songs during instruction had no significant effect on students' ability to sightsing the patterns. Pitch matching correlated highly with sightsinging note accuracy, but moderately with contour accuracy. Sightsinging scores correlated moderately with IMMA and school ability scores.