Silent reading fluency using underlining: Evidence for an alternative method of assessment