Abstract
Participants from Grades 3, 4, 5, and 6 (N = 238) rated selections on an 18-item music listening tape four consecutive times at 2-week intervals, alternately using pictorial and verbal Likert-type response scales. The order of response scale administration was counterbalanced among the classrooms involved. There was no significant difference between preference scores generated by participants using either form of response scale. Test-retest reliabilities ranged from. 65 to. 88 using a 4-week time interval, with higher grade levels associated with higher reliabilities and the pictorial scale generally associated with higher reliabilities. Internal consistency reliabilities measured by coefficient alpha ranged from. 90 to. 93 for the verbal scale and. 92 to. 93 for the pictorial scale. A large and statistically significant majority of participants (84%) expressed a preference for using the pictorial scale to record their music listening preferences.
Original language | English |
---|---|
Pages (from-to) | 425-435 |
Number of pages | 11 |
Journal | Journal of Research in Music Education |
Volume | 46 |
Issue number | 3 |
DOIs | |
Publication status | Published - 1998 |