OSAT ESL (OK177) common mistakes
OSAT ESL (OK177) · Understand the traps and how to avoid them.
Treating BICS and CALP as interchangeable labels, so on scenario items the candidate selects the answer that says a student who chats fluently at lunch is 'ready to exit services' or 'has no language barrier.'
The two terms are memorized as a matched pair of acronyms without the underlying distinction: BICS is context-embedded and cognitively undemanding, CALP is context-reduced and cognitively demanding, and CALP takes years longer to develop. Candidates apply the conversational data point because it is the most visible evidence in the vignette.
What to do instead
Drill the distinction as a two-axis grid (context embedded/reduced x cognitively undemanding/demanding) and place typical classroom tasks on it. For every scenario, ask 'what is the task demand and how much context is supplied?' before judging proficiency. Anchor the timeline: conversational proficiency roughly 1-3 years, academic proficiency roughly 5-7 years, so a fluent-sounding second-year student is rarely ready for exit on academic grounds alone.
Applying an additive bilingual/dual-language rationale to a transitional bilingual program item, or vice versa, and choosing the program model that matches the candidate's own philosophy rather than the described goal.
Program models are learned as a list of names with vague positive associations. The candidate does not hold a clean decision rule separating the goal (monolingual assimilation vs. full bilingualism), the treatment of L1 (exit quickly vs. maintain and develop), and the target population (ELs only vs. ELs plus English speakers).
What to do instead
Build a comparison table with columns for goal, L1 role, duration, and who is served, then cover transitional early-exit, developmental/maintenance, two-way dual immersion, sheltered/structured English immersion, and newcomer programs. Practice sorting stems by the stated program goal first, then eliminate models whose L1 policy contradicts that goal.
On legal and policy items, selecting the option that says a school may waive services or place an EL in a lower grade because the family 'requested it,' or that a parent signature can excuse the district from providing language services.
Candidates reason from general parental-rights intuitions rather than from the specific federal framework (Lau v. Nichols, Castañeda v. Pickard's three-prong test, Title VI, and the EEOA). They also confuse the right to refuse a specific program with the district's non-waivable obligation to serve.
What to do instead
Memorize the Castañeda three prongs (sound educational theory, effective implementation with adequate resources, evaluation and program modification) and the principle that parents may decline a particular program but the district must still provide meaningful access and cannot segregate students unnecessarily. Rehearse the difference between a parent opting out of a bilingual model and a district failing to identify or serve an EL.
Writing lesson objectives that name an activity ('students will do a jigsaw about ecosystems') or a vague aspiration ('students will understand ecosystems') instead of a measurable language objective paired with a content objective.
Candidates conflate the lesson plan's procedures section with its objectives section, and they have not practiced the two-objective format that sheltered instruction requires. The language objective is the part they most often omit entirely.
What to do instead
Use a fixed frame: content objective = observable content verb (identify, explain, compare) plus the specific content; language objective = the language function (describe, sequence, justify) plus the language form or vocabulary students will use, and the modality (speak, write, read). Check every objective by asking 'how would I observe a student meeting this in 30 seconds?'
Choosing an assessment accommodation that changes what is measured, such as reading the reading-comprehension passages aloud to an EL or translating a test of English vocabulary into the student's L1, and defending it as 'making it fair.'
The candidate optimizes for reducing anxiety and equalizing scores rather than for validity. Accommodations and modifications are not separated in their mental model, so any supportive act looks acceptable.
What to do instead
Apply the validity test: does the change alter the construct being measured? Accommodations (extended time, separate setting, glossaries for content-area tests, directions read aloud when reading is not the construct) preserve the construct; modifications (simplifying text, translating the assessed skill, reducing items) change it. For each proposed support, state explicitly what the test is measuring and whether the support touches it.
Assuming a silent newcomer who understands simple commands but produces no speech needs referral for special education or is being defiant, rather than recognizing the silent period as a normal stage of second-language development.
The candidate equates observable verbal output with comprehension and internalization, and has not been given a developmental sequence to place the behavior against. Silence reads as a problem to be fixed.
What to do instead
Learn the typical receptive-before-productive sequence and the expected duration range for a silent period, plus the distinction between a silent period and a genuine communication disorder or a hearing/vision concern. In scenario items, look for evidence of comprehension (following directions, nonverbal responses, L1 literacy) before considering referral, and prefer supports like Total Physical Response, partner work, and alternative response modes over pressure to speak.
Treating oral language as a warm-up rather than as assessed content, so the candidate picks 'give students free conversation time' where the stem asks for structured academic oral development, or omits any oral-language objective from a listening/speaking lesson.
Speaking feels informal and hard to score, so it is planned as an activity rather than as a standard with criteria. The candidate lacks concrete routines (think-pair-share with accountable talk stems, sentence frames, structured academic controversy) to attach to the objective.
What to do instead
Prepare named routines and the language each one scaffolds, plus a simple oral rubric (comprehensibility, vocabulary range, grammatical control, discourse organization). For each routine, be able to state the language function it develops and the evidence you would collect, so items about 'promoting academic talk' have a specific answer rather than a generic one.
Recommending phonics-intensive intervention for an EL whose word-reading is accurate but whose comprehension lags, or recommending comprehension strategy instruction for a student whose decoding is the actual bottleneck.
The candidate does not run a diagnostic decision tree. The label 'struggling reader' triggers whichever intervention was studied most recently, and word recognition versus language comprehension is not separated in their model of reading.
What to do instead
Use the simple view of reading as a decision tree: assess decoding/word recognition and listening comprehension separately, then match the intervention to the weaker component. Know that ELs' comprehension difficulties often reflect vocabulary and background knowledge rather than decoding, and that a decoding problem in an EL is not automatically a language difference. Practice with short profiles until the branching is automatic.
On writing items, marking every nonstandard feature as an error to be corrected, or alternatively accepting all nonstandard forms as 'just a difference,' instead of distinguishing developmental/interlanguage errors, L1 transfer, and dialect features from genuine errors in the target register.
The candidate has no framework for error analysis, so the response defaults to either a deficit lens or an uncritical acceptance lens. Both fail items that ask what feedback is most appropriate.
What to do instead
Classify errors before responding: developmental overgeneralization, L1 transfer (article omission, preposition choice, tense-aspect mapping), dialect features (habitual 'be,' multiple negation), and register-inappropriate but grammatical choices. Match feedback to type and to stage, and prefer focused, prioritized feedback on one or two patterns over comprehensive marking, which research and exam items both disfavor.
Designing content-area instruction that lowers the cognitive demand (simplified worksheets, copying definitions, watered-down texts) in the name of 'sheltering,' or leaving the text and task untouched and calling it 'high expectations.'
The candidate conflates language support with content simplification, and has not internalized the principle that ELs need both access to grade-level content and scaffolded language. Without a planning routine, scaffolding defaults to reduction.
What to do instead
Plan with a routine that holds the content standard constant while varying the supports: visuals, graphic organizers, sentence frames, pre-taught vocabulary, chunked text, native-language previews, and structured peer interaction. For any proposed scaffold, state which language demand it reduces and confirm the content thinking remains intact. Also practice culturally responsive moves such as connecting content to students' funds of knowledge without treating culture as a set of stereotypes.