After a baseline reading rate was established, the intervention was as follows:
- student reads passage at his/her instructional level and graphs the rate
- teacher models the passage
- echo reading of the passage
- two additional practice readings
- Student reads the passage a final time and the results are graphed.
One concern I have over the results is that there is an anticipated rereading effect. Only very rarely will a student read a passage 5 times and listen to it once will there not be an increase in reading rate. There is a reason we ask kids to practice anything- it improves performance. The increase in initial reading rate from the beginning of the study to eight weeks later only had a general trend up for the struggling reader. The general ed student had a small average increase. The other students had the expected mountainous profile that began and ended in the same place. It would be interesting to have the transfer data to see what impact was made outside of the study parameters.
This study does support the idea that different instruction may be useful for students with reading difficulties. Repeated reading may be more useful for them than for the general population.
In watching people implement fluency interventions, I think one thing that is often missing is the modeling of fluent reading. Elementary students are often read to in a fluent manner, but when it comes to repeated reading, they are often sent off to that alone. Without having the passage modeled correctly, how do we expect students to improve? Just reading your passage three times may not be enough if you do not know some words or have not heard it read so that the punctuation informs phrasing. This sort of modeling seems critical for improvements, but it is often skipped in favor of independent reading. Yes, I think students should reread, listen to passages they read being read and echo read, but skipping the listening and modeling seems likely to be doomed or at least less effective.