Page 213

that orthographic processing is proceeding neither quickly nor completely enough to do its job.

For readers who are progressing normally, it is often not before the middle of second grade that the ability to read with expressive fluency and comprehension emerges reliably (Chall, 1983; Gates, 1947; Gray, 1937; Ilg and Ames, 1950). Clinical (Harris and Sipay, 1975) evidence and laboratory (Stanovich, 1984) evidence concur that children who can read second-grade texts accurately can read and learn from text with reasonable efficiency and productivity on their own, provided the text level is appropriate. One of the most important questions for second- and third-grade teachers is therefore how best to help children reach this level. Given that the goal is to help children learn to read the words and understand them too, a promising tactic would seem to be to engage them in more connected reading of appropriate text.

It has long been appreciated that a critical factor in considering the learning impact of time spent reading is the difficulty of the text relative to the student's ability. Common terms to describe differences among text are the following:

·      The independent reading level is the highest level at which a child can read easily and fluently: without assistance, with few errors in word recognition, and with good comprehension and recall.

·      The instructional level is the highest level at which the child can do satisfactory reading provided that he or she receives preparation and supervision from a teacher: errors in word recognition are not frequent, and comprehension and recall are satisfactory.

·      The frustration level is the level at which the child's reading skills break down: fluency disappears, errors in word recognition are numerous, comprehension is faulty, recall is sketchy, and signs of emotional tension and discomfort become evident (cited in Harris and Sipay, 1975).

Regardless of a child's reading ability, if too many of the words of a text are problematic, both comprehension and reading growth itself are impeded. As a general rule, it has been suggested that error