The LD OnLine Forums hold over 20 years of user advice and support. While this content is available to all visitors, the Forums are currently closed to any new posts.
Reading Difficulties
Has anyone peformed an IRI before, and how well do you think it works?? Do you think it is pretty accurate?
An informal reading inventory is ideally drawn from reading materials the student reads in school, from the reading series or something comparable. In the best of situations they read a passage from each reader level orally, answer 8-10 short answer/oral comprehension questions and then ditto for passages read silently. Orally read passages are also scored for oral reading accuracy and type of miscues, etc. Most of the time they are preceded with a word recognition test, lists of words randomly drawn from each reader level (20 per level) to screen sight vocabulary. I use IRI data and prefer it immensely to standardized tests in most cases. It yields more data, usually more test items and I get a better idea of what the child’s actual reading is like. I also use IRI data to write goals/objectives. I write goals like: given a word recognition test from third reader level “X” will correctly read atleast 75% of the words in an untimed presentation. Ditto for the graded passages, I write goals for oral reading accuracy (percentage score), comprehension percentage scores, sometimes for number of words read correctly in a minute (I take three one minute samples), and so forth. I have the Burns and Roe (Houghton Mifflin) which uses vocabulary and selections drawn from Houghton Mifflin basal readers.
If you use an intensive phonics approach early on, controlled readers, you may not get really good and accurate data at the first grade level where the IRI emphasis is on reading less controlled and more natural language. That difference really washes out pretty much by the time you have taught pretty much all the short/long vowel sounds.
An informal reading inventory is ideally drawn from reading materials the student reads in school, from the reading series or something comparable. In the best of situations they read a passage from each reader level orally, answer 8-10 short answer/oral comprehension questions and then ditto for passages read silently. Orally read passages are also scored for oral reading accuracy and type of miscues, etc. Most of the time they are preceded with a word recognition test, lists of words randomly drawn from each reader level (20 per level) to screen sight vocabulary. I use IRI data and prefer it immensely to standardized tests in most cases. It yields more data, usually more test items and I get a better idea of what the child’s actual reading is like. I also use IRI data to write goals/objectives. I write goals like: given a word recognition test from third reader level “X” will correctly read atleast 75% of the words in an untimed presentation. Ditto for the graded passages, I write goals for oral reading accuracy (percentage score), comprehension percentage scores, sometimes for number of words read correctly in a minute (I take three one minute samples), and so forth. I have the Burns and Roe (Houghton Mifflin) which uses vocabulary and selections drawn from Houghton Mifflin basal readers.
If you use an intensive phonics approach early on, controlled readers, you may not get really good and accurate data at the first grade level where the IRI emphasis is on reading less controlled and more natural language. That difference really washes out pretty much by the time you have taught pretty much all the short/long vowel sounds.