HtcHnm / Pixabay
Many educators of English Language Learners, including myself, are dissatisfied with the tests various states use to determine English proficiency for English Language Learners.
I recently put this question out on social media:
What is the best research out there discussing state English tests for English Language Learners (WIDA, ELPAC, ELPA21, etc.), comparing them and exploring how valid or not valid they are?
No one had any suggestions for research but, boy, teachers had a lot to say about the tests – all negative!
I didn’t have a lot of extra time, so I took a lazy way out and asked both Google Gemini and ChatGPT the same question. Here are their answers – note, I have not doublechecked their responses or confirmed that their citations exist. I didn’t want to publish a super-lengthy post, so here’s a PDF summary from ChatGPT.
Here’s what Gemini said:
When it comes to the “Big Three” of English Language Proficiency (ELP) assessments—WIDA (ACCESS), ELPAC, and ELPA21—the research is a mix of high-level psychometric data (from the test makers) and critical analysis from independent sociolinguists.
The consensus in the field is that while these tests are technically “reliable” (they produce consistent scores), their “validity” (whether they actually measure a student’s ability to succeed in a classroom) is a subject of intense debate.
1. The Comparison: WIDA vs. ELPA21 vs. ELPAC
Because different states use different tests, “apples-to-apples” comparisons are rare, but the following distinctions are frequently cited in research:
| Assessment | Scope | Key Research Finding |
| WIDA (ACCESS) | 41+ States/Territories | The “Gold Standard”: Generally considered to have the most robust research base and professional development. However, research suggests it is “harder to pass,” often keeping students in EL status longer than other tests. |
| ELPA21 | ~8 States | Innovation vs. Evidence: Designed to align specifically with the “English Language Proficiency Standards.” Independent reviews (e.g., ERIC, 2019) have noted that because it is newer, publicly available validation research is thinner than WIDA’s. |
| ELPAC | California Only | High Stakes: Research from the Migration Policy Institute highlights California’s shift to ELPAC as an attempt to create more “authentic” tasks, but notes that reclassification still varies wildly by district due to “supplemental” criteria. |
2. Is the Research “Valid”? (The Great Validity Debate)
Validity in testing means: Does this test measure what it claims to measure? The Argument for Validity (The Psychometric View):
Annual technical reports (like the WIDA Annual Technical Report, 2024-2025) consistently show high internal consistency. This means the tests are very good at ranking students by their command of “Academic English.” If a student scores a 5.0, they almost certainly have more academic vocabulary than a student who scores a 2.0.
The Argument Against Validity (The Sociolinguistic View):
Independent researchers (notably Jamal Abedi and Kenji Hakuta) argue that these tests often suffer from Construct-Irrelevant Variance. In plain English:
-
A Test of “Testing”: The tests often measure a student’s “test-taking skills” or “cultural background” rather than their actual language ability.
-
The “Moving Goalpost”: As academic standards (like Common Core) get harder, the ELP tests get harder too. Research suggests that even some native English speakers would struggle to pass the “exit” level of these exams.
-
The Literacy Trap: Critics argue that the heavy focus on reading and writing in these tests ignores the “communicative competence” (the ability to actually talk and function in class) that you see in Task-Based Learning.
3. Key Resources to Follow
If you want to read the “best” sources, look for these specific bodies of work:
-
The CRESST Reports (UCLA): They specialize in the validity of ELL assessments. Look for Abedi’s work on “linguistic complexity” in testing.
-
WIDA Research Publications: They are surprisingly transparent. Their 2024-2025 reports discuss how proficiency scores have declined post-pandemic and what that means for test accuracy.
-
The Migration Policy Institute (MPI): Their report, “The Patchy Landscape of State EL Policies under ESSA,” is the definitive guide on how different states use these tests to trap or release students from EL services.
Source link
#State #English #Tests #ELLs #WIDA #ELPAC #Valid #Worth #Heres #Google #Gemini #ChatGPT
