A Closer Look

Testing Accommodations: From the 2019 Admissions Scandal to the Bigger Scandal of Poor Decision-Making

Americans were shocked by the news that broke in March 2019. A number of affluent individuals, including celebrities Lori Loughlin and Felicity Huffman, had allegedly paid William “Rick” Singer to help their children get into selective colleges through fraudulent means. Singer’s strategies for gaming the admissions process were diverse, including faking athletic accomplishments and even lying about students’ ethnic backgrounds, but one of his techniques involved providing fraudulent college admissions test scores. Singer had associates who worked as proctors, and who could arrange for the scores. But logistical problems remained—in particular, Singer’s associates needed time to procure or produce tests with the high-scoring correct answers.

Enter testing accommodations. Singer realized that if students could obtain a 2-day testing appointment, this would provide sufficient time. The trick then became getting disability accommodations for the students; 100% extended testing time accommodations would automatically push the test to a 2-day administration. Remarkably, Singer knew a great deal about how disabilities are diagnosed and how accommodations recommendations are made. He counseled parents that their children merely needed to show discrepancies between different diagnostic test scores, and that any child would have such discrepancies. Leaving nothing to chance, he asked parents to tell their children to “be slow” and “be stupid” when seeing a psychologist for diagnosis, and to make sure to get 100% extended time. (The wire-tapped conversations between Singer and the parents make for chilling reading for a school psychologist!)

In some ways, the admissions scandal was a highly unusual event; certainly, there is no evidence that this kind of admissions fraud is widespread. However, the accommodations practices that Singer depended on are widespread. The available research suggests that schools make accommodation decisions without attending to relevant data or considering the issues that are key to accurate decisions. Teachers tend to overrecommend accommodations, and special education teams tend to make decisions based on students’ affective qualities such as anxiety and self-esteem.

A much sounder model for accommodation decisions was suggested over 20 years ago by Susan Phillips, then an educational measurement professor. She posed several questions that should be asked before an accommodation is given. They address the following issues:

  1. Do test scores obtained with accommodations have the same psychometric properties as those obtained under standard testing conditions?
  2. Does the proposed accommodation still allow the test to measure what it is designed to measure?
  3. Are the benefits of the accommodation specific to students with disabilities?
  4. Are students with disabilities actually unable to adapt to standard testing conditions?
  5. Is the evidence of disability reliable and valid?

If the answer to any of those questions is “no,” school psychologists should hesitate to recommend accommodations, and special education teams should be hesitant to grant them.

Admittedly, the answers to some of Phillips’s questions are complex and only found in large research literatures. Optimally, state and local educational agencies would review that literature to assist decision makers, but this is more an ideal than a reality. However, interested school psychologists can find analyses of the literature for themselves. In practice, the easiest step that psychologists can take to improve decisions is simply requiring that a student have absolute (normative) deficits in the most relevant skills needed to access tests. For extended time accommodations (the most common type), relevant deficits could include a score below the average range on academic fluency measures; for read-aloud accommodations, similar deficits on measures of word identification (decoding) would be expected. Deficits on more artificial diagnostic tasks—such as processing speed or phonological awareness—are generally less relevant for accommodations decisions.

Admittedly, making accommodation decisions based on empirical evidence and rigorous criteria can be politically difficult. Students, parents, teachers, and administrators all have incentives for getting higher test scores by any means. However, school psychologists understand that the real purpose of tests is to yield valid inferences about students’ skill levels, and sometimes higher scores are actually less valid. By educating our coworkers about these issues, we can help to defend evidence-based practices. We may even prevent another admissions scandal!

About the Author

Benjamin J. Lovett, PhD
Benjamin J. Lovett, PhD, is an associate professor of psychology and education at Teachers College, Columbia University, where his research focuses on the diagnosis of learning and attention problems, the provision of testing accommodations, and the nature and management of test anxiety.