Ensuring Test Access for Students With Disabilities: A Special Issue of Assessment for Effective Intervention

The Imperative of Access to Tests

Students in public schools today take a wide variety of tests. Brief, standardized measures of basic academic skills are used in elementary school for universal screening and for monitoring students’ response to instruction and intervention. Teacher-made classroom tests assess students’ subject-area knowledge and skills in all grades. And particularly in the past 20 years or so, large-scale state assessments have been used for school accountability purposes, raising the “stakes” of test outcomes. The range and consequential nature of testing makes it even more important for administrators to ensure that students with disabilities have access to the tests. Disability conditions often involve deficits in test access skills (cf. Ketterlin-Geller, 2005). For instance, a student with a reading disability may have excellent science knowledge while also having difficulty decoding some of the words on a science test. Generally speaking, such a student must still take the same high-stakes tests as their peers; legal regulations require such participation, in part to make sure that school districts are not neglecting the academic achievement goals for students with disabilities. An accessibility feature, such as a text-to-speech function on a computerized science test that allows any student to click on a word and hear the computer read the word, may allow this student to fully access the test. The student may also need targeted disability accommodations, such as additional testing time or an audio version of the full test items. One dimension of fairness in testing is examinees having access to the constructs as the test measures them (American Educational Research Association et al., 2014), and it is paramount that tests be fair for students with disabilities. Moreover, when decisions about access are made poorly, this can compromise equity of educational opportunity for students from different backgrounds (Lovett, 2021).

Assessment for Effective Intervention has historically been a leader in publishing scholarship on this topic. In fact, nearly 20 years ago, AEI published a special issue (Volume 31, Issue 1; Fall 2005) on “Testing Accommodations and Inclusive Assessment Practices.” In their introduction to that special issue, the guest editors observed that “Researchers are continuing to learn about the use, effects, and consequences of testing accommodations, and much of what they are learning can guide Individualized Education Program team decision making and assessment practices for students with disabilities” (Niebling & Elliott, 2005, p. 5). Our predecessors’ comment remains true today, and we are pleased to share updated research findings with AEI’s audience.

The Special Issue Articles

In our first special issue article, Witmer and Bouck (2024) report on their investigation of eighth-grade students’ use of accessibility tools on the 2017 National Assessment of Educational Progress (NAEP) mathematics test. The 2017 test is the first NAEP version to be administered on computers, and the computers recorded rich “process data” that show every action that students engaged in, including use of accessibility tools such as crossing out an answer option or activating a text-to-speech function. National Assessment of Educational Progress data generate what is often known as the “Nation’s Report Card”; the data are invariably reported by the media and are used to evaluate education policies and practices, and so if students cannot access the test, poor policy decisions may well follow. In addition, NAEP items resemble those found on other common standardized tests (e.g., passages for reading comprehension, word problems for measuring mathematics skills). At the same time, NAEP is a low-stakes test for individual students; NAEP scores do not affect students’ grades, and so as Witmer and Bouck point out, students’ motivation to do well on the test may be suboptimal, which could affect use of accessibility features. Witmer and Bouck found that students with disabilities used such features relatively rarely, even though use was associated with higher performance on the test. School staff may therefore find it helpful to teach or encourage students to use such features where appropriate.

In the next article, Dembitzer and Kettler (2024) examine “universally designed accommodations”—technically, universal design accessibility features—focusing on additional time and audio presentation of text. High school students took two versions of a measure based on the NAEP reading test. Dembitzer and Kettler separated their sample by actual levels of functional impairment (based on curriculum-based measurement reading probes) rather than using disability labels as a poor proxy. In addition to examining the relationship between functional impairment and use of these “accommodations,” the investigators added a mixed-methods element to their research design, soliciting qualitative, open-ended feedback about students’ experiences of the accessibility features. Of particular note, most students found having the accessibility features available helpful, even though almost none of the students used the features. This finding shows the distinction between what might be called the “social validity” of accessibility features (Wolf, 1978) and their actual utility. There is good reason for school staff to care about social validity, but a student’s subjective positive experience of having an accessibility feature—or an accommodation—should not be taken as evidence of its appropriateness.

In the next article, Liu et al. (2024) reports on the performance of students with learning difficulties in mathematics on a computer-based assessment in geometry. This assessment drew items from an actual state exam used for education accountability, making it realistic and ecologically valid. The assessment design contained four “designed instructional conditions”—essentially, accessibility features: the ability to annotate diagrams on the computer screen, the ability to rotate figures to assist with spatial reasoning, a pop-up glossary feature, and labels on the geometry diagrams that cued the student to aspects of the instructions for an item. The authors investigated patterns of use of these accessibility features as well as relationships with performance on the test items. Interestingly, the most commonly used accessibility feature was the pop-up glossary, suggesting that students with learning difficulties in mathematics may benefit from targeted instruction in factual verbal knowledge about mathematics concepts—knowledge that is not often tested deliberately on math tests but is important subject-specific knowledge, and that is needed to access math test items nonetheless.

Finally, Lovett and Fienup (2024) contributed a conceptual review article in which they use applied behavior analysis as a framework for developing interventions that can be provided in addition to—or in lieu of—assessment accommodations. Since accommodations are given to students with deficits in access skills, an alternative strategy, when possible, would be to increase those access skills through some kind of intervention. Lovett and Fienup first conceptualize common accommodations in terms of alterations to the various stimulus and response elements of test-taking, and then discuss related evidence-based interventions that can be used to teach students with disabilities to make responses of the form that nondisabled students are expected to make, to the stimuli found in standard administrations of tests. To play on the title of the journal, their article might be thought of as presenting “interventions for effective assessment!”

The Present and Future of Accessibility Research

The articles in this special issue suggest important and interesting trends that were not evident in AEI’s prior special issue on the same topic, or much of the research published since then. The three empirical articles all examine, in some way, the actual use of accessibility features by students, rather than simply the provision of the features. Much of the extant literature on accessibility has looked at the relationship between receipt of accommodations and subsequent test performance without examining the critical mediator of actual use. The availability of process data from computer-based tests has made use data easier to collect (Lee & Haberman, 2016), and we are hopeful that this trend will continue. Meanwhile, Lovett and Fienup’s article links accessibility to interventions, a connection missing from the vast majority of the accommodations literature, despite the increased emphasis in recent years on intervention-heavy roles for school psychologists and the refinement of a library of effective interventions for common academic and behavior problems (cf. Merrell et al., 2022).

Although we believe that the articles in the present issue are timely and valuable, they also illustrate continuing challenges and suggest areas of remaining research need. The three empirical articles all involve measures with low stakes for students, and so while the results may generalize well to other low-stakes tests (of which there are many), additional research is needed on exams that students are typically highly motivated to do as well as possible on, such as classroom tests, graduation competency exams, and college admissions tests. Beyond the issue of consequences for students, some of the articles involve data with likely ceiling effects, where accessibility features may not have been used because they were not needed, whereas in other contexts, the same features would be used far more often. In addition, while Lovett and Fienup’s article is provocative, we are not aware of much published research examining empirically what happens when interventions are added to—or used to replace—testing accommodations. Although Harrison et al. (2020) ran a randomized controlled trial of interventions versus accommodations for attention-deficit hyperactivity disorder (ADHD), the accommodations related more to instruction than testing; a similar study on testing accommodations would be very useful to have.

There is a final challenge suggested by the present articles, and it is not new. In the 2005 AEI special issue on accessibility and accommodations, the guest editors were disappointed that “decisions regarding the selection and implementation of testing accommodations often are driven more by policy and teacher judgment than by research findings” (Niebling & Elliott, 2005, p. 5). In our experience, this continues to be a problem, but since 2005, major advances have been made in implementation science—the study of how to increase the likelihood that evidence-based practices will be implemented in real-world settings (e.g., Odom et al., 2020; Sanetti & Collier-Meek, 2019). We have a better understanding of the barriers to implementation—and the corresponding facilitators—than we did 20 years ago. It is our hope that the articles in this special issue have suggested practices that are ripe for implementation: measuring students’ subjective experience of accommodations along with objective data on accommodation use and effects; inspecting use of accessibility features to better understand deficits in particular access skills; and attempting interventions before rushing to accommodate access skill deficits.