Does Retention Really Help?

Today Dr. Barresi held her press conference to dispel any myths about the Reading Sufficiency Act and forced retention.  I have posted previously about this topic and so have several others.  If you would like to learn more about it check out OkEducationTruths, Rob Miller, Jason James, and Claudia Swisher's blogs.

In Oklahoma we have always had a tendency to look at what is happening in Florida, California, & Texas to determine our education policy.  Each of these three states implement a change and Oklahoma follows suit, but not until the other states are ready to abandon it or have determined that it doesn't work.  The world is constantly changing, I get that.  As a result, we should be changing and adapting all the time.  Anyone in education that is worth their salt is constantly reviewing their practice and improving. We look at what others are succeeding with & we incorporate what we can into our own instruction.  What we don't want to do is take what doesn't work in other places and  try to make it work here.

Today, during the press conference, Dr. Barresi made this comment:

#oklaed Barresi: ten years ago, when FL implemented this law, third graders were a grade level below OK students. Now a grade level ahead.
— Rick Cobb (@grendelrick) January 27, 2014

I am having trouble finding research to back up this statement, but it could be because I am looking in the wrong place.  I did find some research examining the relationship between retention and later school performance in Florida put together by a small university called Harvard.  The link below will take you directly to the article:

The Effects of Test-Based Retention on Student Outcomes over Time: Regression Discontinuity Evidence from Florida

When you start looking for reputable research regarding retention (alliteration-I think that is on the test) you discover that there is little support for it.  This study brings out an interesting point.  It says that it does work, but only for a short time.  In fact here is a quote from the abstract:

"The achievement gains from test-based retention fade out over time, however, and are statistically insignificant after six years."

According to this study of the Florida retention policy by Harvard, there is an initial improvement for the retained students, but by the time they reach high school the gains are lost.  

You might also want to know that the "Good Cause Exemptions" for Oklahoma students are identical to Florida's "Good Cause Exemptions."  Can we have some legislation in Oklahoma that is actually written for Oklahoma students and not for Florida students?  If you have done any other reading you would know that it wasn't written for Florida either.  Read this post of Rob Miller's to see what I mean. 

There are very few differences between the Florida policy and Oklahoma policy on retention.  When I look, the only significant difference is that Florida has a five tier scale and Oklahoma has a four tier scale.  The Florida policy retains students that fall into the bottom two tiers and Oklahoma will just retain the bottom tier.  The tiers are Advanced, Proficient, Limited Knowledge, & Unsatisfactory (see the PLDs or Performance Level Descriptors for more detail on each meaning) and are arbitrarily assigned point values to fit the magical A-F Grade calculations.  

Dr. Barresi would lead you to believe that if a student scores unsatisfactory on the 3rd grade OCCT Reading test this spring they are at least two grade levels behind in reading.  That couldn't be further from the truth.  We actually have tests we use to determine the reading level of a student & this isn't one of them.  This test is designed to measure how well a 3rd grade student has mastered the 3rd grade PASS soon to be Oklahoma Academic Standards (OAS).  A student could read on a 5th grade level and still not have mastered the 3rd grade standards. 

The test is broken down as follows:

Vocabulary---24%

Comprehension/Critical Literacy---48%

Literature---16%

Research and Information---12% 

See the test Blueprint for a more detailed breakdown.

According to the Test Item Specifications 20-25% of the test is strictly recall and reproduction.  This includes using a dictionary, referencing details in a text and identifying figurative language.  65-70% of the test are skills and concepts questions.  This would be predicting outcomes, using context clues, and identifying and summarizing major events in a story or passage.  The final 5-15% of the test is strategic thinking.  These questions require students to analyze and synthesize information from multiple sources, examine and explain alternate perspectives, describe and illustrate how common themes are found across texts from different cultures, determine the author's purpose and determine how it affects the interpretation of a reading selection, summarize information from multiple sources to address a specific topic, and lastly analyze and describe the characteristics of various types of literature.  

You tell me, does that sound like a way to determine if a student is reading on a third grade level?  

Also of note, every passage on the test is 200-600 words in length followed by 3-5 questions.  Some passages are paired together for compare/contrast purposes.  Those paired passages have 3-5 questions each, plus another 3-5 questions to compare/contrast the passages.  

The only mention of 3rd grade level reading material is in the fact that each passage is supposed to be written on grade reading level which really should be near 3.7 by testing time.  Again, a student can read on grade level and still not pass this test if they have not mastered the skills mentioned above.  

Just because the passages are written on grade level does not mean the test can be used to determine if a student reads on grade level.  It should never be argued that a student is being retained because they performed two grade levels below on a test that was never designed to determine grade level reading ability, but was only designed to determine content mastery.  Nor should a test of this nature ever be used as the sole determining factor in a child being retained.