Praxis ESOL common mistakes
Praxis ESOL · Understand the traps and how to avoid them.
Labels any 'smallest unit' in a stem as a phoneme, even when the item is about written letters or spelling patterns - for example, counting the letters in 'though' and reporting four phonemes.
Phoneme and grapheme are both introduced as 'the smallest unit of sound/symbol,' so candidates skim past the qualifier that tells them which one the item is testing.
What to do instead
Before choosing, say the word aloud and count what you actually hear; 'though' is two phonemes (/ð/, /oʊ/) but four graphemes. If the count changes when you say it, it is a phoneme item; if it only changes when you spell it, it is a grapheme item.
Treats every affix as derivational, or claims that adding -s to a noun or -ed to a verb 'makes a new word.'
Inflectional and derivational morphemes get memorized as vocabulary lists rather than as a two-part decision rule, so any affix item feels like the same item.
What to do instead
Apply a two-question test: does the affix change the part of speech or the core meaning (derivational: un-, -ness, -tion, -ly, -ful), or does it only mark grammatical information such as number, tense, possession, or comparison (inflectional: plural -s, -ed, -er/-est, possessive -'s)? The word class stays the same for inflectional affixes.
On a vignette describing a student who chats easily with peers and jokes at lunch, selects 'exit the student from EL services' or 'no language support needed.'
Conversational fluency is the most visible and most rewarded behavior in a classroom, and the exam vignette deliberately makes it salient to see whether the candidate separates social from academic language.
What to do instead
Anchor BICS at roughly 1–3 years and CALP at roughly 5–7 years in the fixed phrase you carry into the test, then ask what the task demands: a lunchroom conversation is a BICS task, a textbook chapter with content-specific vocabulary and complex syntax is a CALP task. Judge readiness against the academic task, not the student's social ease.
Attributes every learner error to first-language transfer - including overgeneralizations such as 'goed,' 'foots,' or 'mouses.'
Transfer is the term candidates associate with 'interference from L1,' so it is over-applied to any error that looks unusual; the alternative explanation (rule overgeneralization during development) is less familiar.
What to do instead
Run a source check: if the erroneous form occurs in the learner's L1 and in the input, consider transfer; if the form exists in neither the L1 nor the learner's input ('goed' from the regular past rule, 'foots' from the regular plural rule), it is a developmental overgeneralization. Choose the option that names the actual source.
Responds to errors in a fluency-focused speaking activity by selecting 'correct each error as it occurs' or by replacing the activity with a grammar drill.
The stem mentions errors, so an option that directly addresses errors feels responsive; candidates also conflate 'students need feedback' with 'students need immediate explicit correction.'
What to do instead
Match feedback to the purpose of the task. Fluency/speaking-for-meaning tasks take delayed, meaning-focused feedback (recasts, teacher notes, follow-up mini-lesson); accuracy-focused tasks can take immediate correction. Eliminate any option that stops the communicative task in order to correct form, unless the stem says the goal is accuracy.
Plans sheltered instruction by keeping the same text and task demands as the general education lesson and simply speaking more slowly, repeating directions, and defining vocabulary - or, conversely, lowers the content objective itself.
Scaffolding gets miscoded as 'make it easier,' so slowing delivery or simplifying the content both feel like sheltering.
What to do instead
Hold the content objective constant and change only the access: visuals, realia, sentence frames and word banks, modeled examples, chunked text, graphic organizers, L1 supports, and comprehension checks that require student output rather than 'Do you understand?' If an option removes the grade-level content expectation, eliminate it.
Selects a test described as highly consistent or internally reliable as 'best' when the stem asks whether the instrument measures the intended standard or construct.
Reliability and validity are both quality-sounding words, and consistency is easier to spot in an option than construct alignment.
What to do instead
Ask two separate questions. Validity: does this instrument measure the construct it claims to measure, for this purpose, with this population? Reliability: would it give consistent results across occasions and raters? Validity is the priority - a reliable test can reliably measure the wrong thing - so when the stem is about measuring the standard, choose the option that names construct alignment, not consistency.
Confuses the state English language proficiency test with an achievement test, using ELP screening results to draw conclusions about content mastery or to justify a special education referral.
Vignettes contain several test types, and any score looks like evidence of learning; the purpose labels are not reinforced in everyday school talk.
What to do instead
Classify by purpose and by who uses the score: proficiency tests describe what a student can do in English and drive identification, placement, and exit; achievement tests measure mastery of taught curriculum; diagnostic tests pinpoint specific skill strengths and gaps. An ELP score alone cannot establish content mastery or a disability.
Interprets a student's silence, avoided eye contact, or reluctance to challenge a teacher as rudeness, disengagement, or low motivation, and selects a classroom-management consequence as the response.
Participation norms are learned implicitly and feel like universal standards, so culturally patterned behavior is read through the teacher's own frame.
What to do instead
Describe the behavior before evaluating it ('the student does not volunteer or make eye contact') and then select the option that gathers information about family and community norms and uses varied participation structures (small group, written response, choral response). Eliminate options that diagnose attitude from behavior, and keep cultural style distinct from both language proficiency and disability.
Asserts that a specific program model is legally required - for example, that Castañeda v. Pickard mandates transitional bilingual education, or that Lau v. Nichols requires bilingual instruction.
Landmark cases and program models are studied together, so the case names become fused with specific models, and 'bilingual' is often the only model a candidate can name.
What to do instead
Attach each authority to its actual holding: Lau requires meaningful access to the curriculum and does not prescribe a model; Castañeda supplies a three-part test (a research-based educational theory, implementation with adequate resources and staff, and evaluation showing the program works) that any model can satisfy or fail; Plyler requires enrollment regardless of immigration status; Title III requires ELP standards, annual ELP assessment, and accountability. If an option names a mandatory model, eliminate it.