Introduction

Most reading researchers have concluded that early identification and intervention beats late identification and intervention. This belief along with the "No Child Left Behind" legislation has focussed public policy on screening and early detection of children likely to encounter reading difficulties. My focus in this paper is on measurements that might be good candidates for screening students requiring secondary intervention (i.e., intervention that is more systematic and intense than generally effective classroom instruction). I approach this task by considering: the school context in which screening occurs: the attributes of an ideal screening system, important ideas to come out of research on screening, and measures that hold promise as screening tools.

Screening in the Context of a Responsive Intervention Model

A Responsive Intervention Model provides timely and correct intervention to every child who requires additional or different instruction from that given in normally effective general education classrooms. Intervention is timely if struggling readers receive help sooner rather than later, and the amount and type of intervention is adjusted when needed. Intervention is correct if its content, delivery, and amount are appropriate for and specific to children's learning needs, and results in improved outcomes (i.e., interventions are effective). Screening is the mechanism for identifying "struggling readers" who lack prerequisite skills or who acquire reading skills at a pace that puts them at risk for unsatisfactory outcomes.

Attributes of an Ideal Screening Mechanism

Ideally, a screening procedure satisfies at least three criteria. First, it must accurately distinguish individuals that require intervention from those who do not. The former are referred to as "at risk" for poor outcomes, the latter as "not at risk" for poor outcomes. In the framework employed by Lichtenstein and Ireton (1984), a screening measure is characterized by its degree of sensitivity and specificity.

Sensitivity refers to the degree a screening mechanism identifies as "at-risk" individuals who in fact perform unsatisfactorily on a future criterion measure (i.e., "true positives"). Specificity refers to the degree a screening mechanism identifies as "not-at-risk" individuals who later performs satisfactorily on a future criterion measure (i.e., "true negatives").

Sensitivity increases as one type of prediction error (false negatives) decrease. False negatives are individuals that a screening measure classifies not at risk, but who perform poorly on the later criterion measure. In a responsive intervention model, avoiding false negative errors is critical; otherwise individuals most in need of assistance do not receive it.

Specificity increases as another type of prediction error (false positives) decrease. False positives are individuals that a screening mechanism classifies as at risk, but who later perform satisfactorily on the criterion outcome. A screening instrument that produces too many false positive errors wastes precious intervention resources on individuals who do not require them.

The second criterion for an ideal screening mechanism is that it is practical. To attain this criterion a screening procedure must be brief as well as simple enough to be implemented reliably on a wide scale under normal circumstances by normal people. In addition, school personnel must perceive the screening procedures as reasonable. If they perceive screening procedures as onerous, they will not reliably implement them.

A third criterion for an ideal screening system is that the net effect of its implementation is positive. Messick (1980, 1989) refers to this characteristic as consequential validity. Attaining consequential validity means that the screening procedure does no harm--that is, it avoids inequitable treatment; does not consume resources that could be put to better use; and is linked to effective interventions. In other words, screening procedures should be as simple and cost effective as possible for achieving results.

A fourth criterion worth considering (but not essential) is the degree to which screening assessments are integrated with assessments used in other parts of a responsive intervention model. Ideally, assessments used for screening should be linked not only to future criterion outcomes, but also to assessments used for progress monitoring and formative evaluation. Thus, at risk students' responsiveness (or non-responsiveness) to secondary intervention should be apparent on assessments resembling those that identified the students as at risk in the first place. This criterion is particularly relevant to the focus of this conference (Response to Treatment as a model for Learning Disabilities Identification).

Previous Page | Next Page
(Table of Contents) | (Research Highlights)