I recall my surprise when I first saw the different reading levels each of my Joe Sherlock books had been assigned. Surprised mostly because I felt they were all about the same reading level. I wrote them, I should know, right?
Some of the Sherlock books were rated at an early third-grade level, while others were deemed to be right for a student with a high fourth-grade and even an early fifth-grade reading level.
Huh? Was I unaware of my difficulty level while writing these books? Did I inadvertantly make some books more complex and difficult than others? Was I having a three-month senior moment during the creation of each of these mysteries? So much so that they could fluctuate almost three grade levels? Was I asleep at the keyboard?
I did some digging.
I learned that books are not assigned a reading level by a reading expert, a panel of teachers, a seasoned librarian or even a children's cognition specialist. I discovered that the books are torn apart and the pages are fed through a scanner. The computer converts the scanned image into text, and the computer uses some kind of complex formula (with some fancy-pants algebra, a half-dozen mysterious algorithms and a handful of isoceles triangles thrown in for good measure) to assign a "grade level appropriateness" rating to each book.
Well, I felt a little bit better. Besides, what does HAL know about reading books? (Is there a problem, Dave?) I walked away somewhat mollified, but still somewhat puzzled. Until I stumbled upon this interesting article by "The Numbers Guy" (aka Carl Bialik) in The Wall Street Journal about exactly how reading levels are determined.
See, when you peel back the layers, things usually just keep getting uglier . . . so view assigned reading levels and grade levels with a grain of salt.