Accommodation Request on a Test That Measures Listening Itself
A candidate applying for a Persian-Farsi seat submits audiological documentation of a central auditory processing disorder. The audiologist's letter states that her hearing thresholds are normal and that her deficit is specific to rapid phonetic transitions and speech-in-noise recognition. She requests four accommodations on the DLAB: (1) that any audio item be replayed on request, (2) that a written transcript of the audio items be provided, (3) 50 percent additional time on the whole test, and (4) a quiet, separate room. She has previously completed a two-year college language sequence in another language with a 3.4 GPA and no accommodations, and her documentation notes that she has never been tested with an unaccommodated listening measure. The test's audio parts are paced by a recording, not by the candidate, and the composite is a threshold used to seat soldiers into courses whose first weeks are almost entirely listening and speaking. Seat decisions for this cycle close in three weeks.
1. Sort the four requested accommodations by whether they preserve or destroy what the test is measuring, and say why.
Read the worked answer
Preserve the construct:
Quiet, separate room (#4). The test is delivered by audio. Background noise adds error variance without adding information. For a candidate whose documented deficit is specifically speech-in-noise, a quiet room removes a contaminant. This is a clean grant and arguably good practice for everyone.
Extra time (#3), with a caveat. The audio parts are paced by the recording, so extra time does nothing there - it is neither granted nor denied in any meaningful sense. It would apply only to the visually delivered rule section. Grant it there; note in the record that the audio portion is unaffected either way.
Destroy the construct:
Replay on request (#1). Real language use is not rewindable. The listening parts measure the ability to extract meaning from a single pass of speech under time pressure; that is exactly the demand of the target course. Replay converts the measure into something else.
Written transcript (#2). This is the most damaging of the four. It converts an auditory comprehension test into a reading comprehension test. The candidate would be measured on a skill she is not disabled in, and the score would carry the same label as everyone else's while meaning something entirely different. If a transcript is provided, the resulting composite must not be used for the same threshold decision.
The uncomfortable core: the disability and the construct overlap. That is not a reason to deny accommodation; it is a reason to be honest that this instrument cannot fully do the job for her.
2. Give the decision-maker a recommendation for this cycle.
Read the worked answer
Grant #4 and the visual-section portion of #3. Deny #1 and #2. Document the reasoning, not just the outcome.
Why not just grant everything and move on. Because a composite produced under #1 and #2 would be reported as a DLAB score and compared to a threshold set on unaccommodated scores. The seat would be filled, and the school would discover the mismatch in week three of a course that starts with listening. Granting accommodations that erase the measured skill is not kindness; it is deferred failure.
Why not deny everything. Because #4 removes a documented contaminant, and refusing it would mean scoring her through noise that other candidates did not face.
The individualized step. Ask whether an equivalent, non-audio-gated route exists for establishing readiness - for example, a structured interview with a language instructor, review of her prior language coursework and grades, and a sample lesson with instructor rating. These do not replace the threshold; they inform whether a waiver is defensible. That is a decision the waiver authority owns, not the test administrator.
Timing. Three weeks is enough for the quiet-room arrangement and the interview; it is not enough for a validation study. Say so rather than promising one.
Analytic honesty. Her 3.4 GPA with no accommodations in a college sequence suggests she compensates well. Compensation in a self-selected, self-paced setting is not evidence of compensation in a paced, immersive one. Treat it as a favorable signal, not as proof.
3. The registrar proposes a standing rule: "no accommodations on the audio portion, ever." Critique it.
Read the worked answer
It is simple, uniform, and wrong in at least two ways.
It conflates the audio portion with the accommodations. The audio portion contains several different demands - stress and pitch pattern discrimination, phonetic discrimination, grammar-rule inference delivered by voice. A quiet room is an accommodation on the audio portion and is clearly construct-preserving. A blanket rule denies the good accommodation along with the bad ones. The correct unit of analysis is the specific accommodation, not the delivery medium.
It removes the individualized assessment the framework requires. Accommodation decisions turn on the interaction between a specific functional limitation and a specific test demand. A rule that forecloses the inquiry before it starts will not survive review, and it will not be applied consistently - someone will make an exception for a sympathetic case, and then the rule is gone anyway, but now without a record.
It also misses the useful information. A request like this one tells you something important: some candidates' disabilities overlap the construct, and the standard instrument cannot fully resolve them. The productive response is a documented alternate-evidence route with the waiver authority's name on it - not a wall, and not a rubber stamp.
Better standing rule: accommodate conditions that add error without adding information (environment, amplification, breaks, extra time where pacing permits); decline changes that alter the measured skill (replay, transcripts); require every decision to cite which category it falls in; and route construct-overlap cases to a named individual with authority to weigh alternate evidence.