By Dr. Evelyn Johnson
Boise State University
This past week, I worked closely with an elementary school to develop a screening process for quickly identifying students who may be at risk for not meeting performance standards in reading, math, and behavioral expectations. A lot of interesting discussions came up regarding the role of assessment, the requirements for developing and implementing an effective screening process, and the need to consider the costs and logistics of implementation against the benefit of improving the accuracy of screening decisions.
So, with this entry I’m going to digress from secondary issues for just a minute to discuss more general issues related to screening in an RTI framework. Screening is one of the essential components of an RTI framework -- kids are screened for academic and behavior issues about three times a year to identify as early as possible those who may need support to meet grade-level performance standards.
Screening presents many challenges for schools. The goal is to identify all or nearly all of the kids at risk while minimizing the number of students who are over-identified. Screening is meant to be a brief assessment that results in a highly accurate identification of an initial risk pool of students. Screening is not meant to be diagnostic or in-depth -- the goal is to quickly determine who may be at-risk and then follow-up with more in-depth procedures to determine the extent and nature of the student’s difficulties.
Some of the challenges of developing a quality screening program include finding measures that are:
- highly predictive of the construct of interest (e.g., reading or math or behavior)
- group administered and efficiently scored
- accurate in sorting at-risk from not-at risk students
- high in face validity -- do teachers understand what is being measured by the screen and does it appear to be relevant to the construct of interest?
Fortunately, new tools are available to help practitioners locate good screening measures for grades K-12. An example is the Screening Tools Chart found at the National RTI Center’s Web site.
The specific measures used for screening are one part of the process, but the approach or model under which they are used is also important. Different schools operate under different screening approaches, each with its own advantages and disadvantages. In general, screening approaches can be categorized under the models listed below. These are explained by Jenkins and Johnson (2009) in more detail in an article on the NCLD’s RTI Action Network Web site.
Direct Route Models – In direct route models of screening, students who are identified as at-risk on the screening measure are placed directly into a Tier 2 intervention with no further testing. This type of a model requires a screening process to be highly accurate, identifying nearly all of the students who are at-risk while minimizing the number of students who are over-identified. Few available single screening measures are accurate enough to be used with great confidence in a direct route model -- too many at-risk students are missed, or too many students are falsely identified as at-risk, and valuable, limited intervention resources are unnecessarily provided.
Progress Monitoring Models – In progress monitoring (PM) models of screening, students initially identified as at-risk are monitored for a series of 5-6 weeks to determine if they "self-correct." Do they respond to Tier 1 instruction sufficiently so that they no longer appear at-risk on the screening measure? PM models have been shown to greatly improve accuracy, but they can be logistically difficult to manage. In a recent article published in TEACHING Exceptional Children, Davis, Lindo, and Compton (2007) provide an in-depth description of a PM model for screening (CEC members-only content).
Multiple Measure Models – In this approach to screening, more than one measure is used. The idea is that accuracy can be improved by administering several screening measures at the same time. Research has shown that such approaches result in higher accuracy levels than direct route approaches, but the interpretation of multiple cut scores, as well as the time and resources required to administer multiple measures to all students, make this approach less feasible for schools to implement.
Multiple Gating Models – Perhaps the easiest way to explain this model is by analogy. Multiple gating models are similar to the process of an eye exam. The eye chart is the initial screen that all of us take. Depending on how well we read the lines on the eye chart, we may be referred for further examination. The goal in this phase is to determine if the initial screening result was accurate and, if so, to what extent our vision is impaired, in what way (e.g., are we nearsighted or farsighted?), and finally to determine appropriate courses of action (think of the optometrist changing the lenses and asking “Is this better or worse?”). Multiple gating models generally are more accurate than direct route models but can be resource-intensive to implement.
In a secondary setting, the goal of screening procedures is to identify those students who are at-risk for not meeting the academic and behavior requirements of school early on. In high school, screening procedures can also help identify students at-risk for dropping out. In my next entry we’ll look at some great resources developed by the National High School Center to assist high schools in screening for students who are at-risk.
Thanks for reading!