Conclusion
Based on this selective review, there are plenty of good candidates for screening measures. However, if we are to deepen our knowledge about accurate screening, we need to establish that the candidate measures produce satisfactory sensitivity and specificity in classifying students into risk or risk categories. This will require a certain kind of research on the candidate measures. Specifically, we need studies that adopt the following research protocol.
- Identify a criterion measure and the performance level (score) required for satisfactory reading.
- Choose 1-3 candidate screening measures (possibly from those mentioned in this review) that are: correlated with the criterion measure, relatively brief, and can be reliably administered. In choosing these candidate measures, give priority to those that could also be used for subsequent progress monitoring.
- Decide on an acceptable level of false negatives (e.g., 5-10%) or sensitivity (90-95%).
- Administer the candidate measures during the desired screening period to a large sample of representative students.
- Follow the students and administer the criterion measure.
- Identify cut-scores on the candidate measures that net 90-95% (depending on the decision in step 3) of the students that failed to attain the criterion score.
- Choose the screening measure that produces the highest specificity (the fewest false positives).
- Determine if the sensitivity and specificity can be improved by combining the results of more than one candidate measure.
- Cross-validate the cut-scores, sensitivity, and specificity either by (a) repeating the procedure with another group or (b) applying the cut scores to a subsample that you reserved from the larger sample.
Currently research using this protocol is in short supply. In the meantime, practitioners can rely on this review for selecting the measures that hold the most promise. Whether we can settle on one best approach for screening is another matter. Local preferences for criterion measures, criterion performance levels, and tolerance for under- and over-identification rates will lead to different choices for screening. What is critical at the local level, however, is ensuring that screens select all or nearly all students who require secondary intervention and progress monitoring. Only then can "Response to Treatment" be implemented to identify students with learning disabilities.
Previous Page | Next Page
(Beyond Grade 2) | (References)
