The Acquisitional Value of Recasts in Instructed Second Language Speech Learning: Teaching the Perception and Production of English /ɹ/ to Adult Japanese Learners

ERCT Criteria Breakdown

• Level 1 Criteria

  • C

    Class-level RCT

    • The paper only clearly documents random assignment of individual students to classes, not random assignment of classes to the three treatment conditions, and the author self-labels the design as quasi-experimental.
    • "After taking the pretests, all students were randomly assigned to eight classes of about six students each and then categorized into three groups..." (p. 507)
    • Relevant Quotes: 1) "Forty-five learners were randomly assigned to three groups—FFI recasts, FFI only, and Control—and exposed to four hours of communicatively oriented lessons." (p. 499, Abstract) 2) "The present study investigated the acquisitional value of recasts and FFI in L2 speech learning in a quasi-experimental, pre- and posttest design." (p. 506) 3) "After taking the pretests, all students were randomly assigned to eight classes of about six students each and then categorized into three groups: (a) FFI-recast (three classes, n = 17), (b) FFI-only (three classes, n = 18), and (c) Control (two classes, n = 10)." (p. 507) 4) "Although I originally planned to have nine classes (i.e., three classes for each experimental and control group), one Control class was cancelled for practical reasons such as teacher and classroom availability." (p. 526, Note 3) Detailed Analysis: The abstract states that individual learners were randomly assigned to three groups, while the Method section instead describes a two-step process: students were first randomly assigned to eight small classes, and these classes were then "categorized" — not explicitly stated as randomly allocated — into the three treatment conditions (3 FFI-recast, 3 FFI-only, originally 3 Control classes). Note 3 further reveals that the number of Control classes was reduced from three to two "for practical reasons such as teacher and classroom availability," indicating that at least some class-to-condition allocation was governed by logistics rather than randomization. While each class as a whole received a single treatment condition (avoiding contamination between conditions within the same class), there is no unambiguous quote confirming that the assignment of specific classes to specific conditions was itself random, and the author explicitly labels the overall design "quasi-experimental." Given this ambiguity and the explicit quasi-experimental self-classification, criterion C is not met.
  • E

    Exam-based Assessment

    • The study used researcher-designed perception and production tests targeting the /ɹ/-/l/ contrast, not a standardised, widely recognised exam.
    • "Perception was measured using a two-alternative forced choice identification task, while pronunciation performance was elicited using controlled and spontaneous production tests..." (p. 499)
    • Relevant Quotes: 1) "Perception was measured using a two-alternative forced choice identification task, while pronunciation performance was elicited using controlled and spontaneous production tests and assessed by 10 naïve native-speaking listeners." (p. 499) 2) "The impact of instruction on learners' perception abilities of /ɹ/ was measured via a two-alternative forced-choice identification task: They listened to a total of 70 minimally paired words (e.g., 'read' vs. 'lead')..." (p. 510) 3) "Two production tests (word reading and timed picture description) were designed to measure the learners' pronunciation performance of /ɹ/ at two different levels of processing: controlled vs. spontaneous speech." (p. 511) Detailed Analysis: All three outcome measures (the forced-choice perception task, the controlled word-reading production test, and the spontaneous picture- description production test) were bespoke laboratory- style instruments constructed specifically for this study to target the English /ɹ/-/l/ contrast. None of these are standardised, widely recognised exams such as a national curriculum test or state-wide standardised assessment. Therefore criterion E is not met.
  • T

    Term Duration

    • The intervention consisted of only four hours of instruction, with outcomes measured just two weeks later, far short of one academic term.
    • "A posttest session took place two weeks after instruction." (p. 506)
    • Relevant Quotes: 1) "two experimental groups received four hours of instruction that included an FFI treatment on pronunciation of English /ɹ/..." (p. 506) 2) "A posttest session took place two weeks after instruction." (p. 506) 3) "This testing interval was determined based on psycholinguistic research evidence that the integration of novel phoneme sequences into the mental lexicon takes around a week or so rather than immediately after exposure... It could be considered as 'short-delayed' rather than an 'immediate' interval according to the FFI research standards in L2 grammar studies." (p. 523, Note 2) Detailed Analysis: The instructional treatment totaled only four hours, and the primary outcome measurement (the posttest) was collected merely two weeks after instruction ended. The author even explicitly frames this interval as "short-delayed" rather than long-term. This interval is far shorter than the one full academic term (~3-4 months) required by the ERCT standard, so criterion T is not met.
  • D

    Documented Control Group

    • The control group's size, demographics, and exact instructional content are clearly documented.
    • "The students in the control group received instruction on English argumentative skills which did not include any FFI component on /ɹ/ but was identical in all other aspects." (p. 512)
    • Relevant Quotes: 1) "a control group received comparable instruction in terms of length and nonpronunciation (argumentative skills) content, without any focus on form regarding pronunciation." (p. 499) 2) "The students in the control group received instruction on English argumentative skills which did not include any FFI component on /ɹ/ but was identical in all other aspects. That is, all target words in the instructional materials were replaced with comparable words ('rain' → 'snow,' 'run' → 'jog'). For the warmup games, they did other purely communicative activities with no emphasis on pronunciation or listening practice." (p. 512) 3) "Control (n = 10): Age M = 29.7 (SD = 6.7); Gender 1 male, 9 females; LOR in months M = 7.2 (SD = 11.3)." (Table 1, p. 507) Detailed Analysis: The control group's sample size, age, gender composition, and length of residence abroad are reported in Table 1, and the narrative text specifies precisely what the control group received (identical argumentative-skills lessons of the same length, with target words substituted by non-target lexical items, and no pronunciation focus). This level of detail satisfies the documentation requirement, so criterion D is met.

• Level 2 Criteria

  • S

    School-level RCT

    • The study was conducted within a single language institute, not across multiple randomised schools.
    • "The instruction phase took place at a private language institute in Osaka, Japan." (p. 506)
    • Relevant Quotes: 1) "The instruction phase took place at a private language institute in Osaka, Japan. The participants were adult Japanese learners in an EFL setting." (p. 506) 2) "For recruitment purposes, the researcher distributed ads to students who belonged to the language institute and its branch schools as well as university-level schools near the research site." (p. 506) Detailed Analysis: Randomisation occurred at the level of individual students being placed into small classes at a single language institute, not at the level of independent schools being randomly assigned to conditions. There is no school-level randomisation described anywhere in the paper, so criterion S is not met.
  • I

    Independent Conduct

    • The same researcher who designed the intervention also trained the instructor, observed all sessions, and personally conducted the data analysis, with no independent evaluator involved.
    • "the researcher always sat at the back of the classroom as an observer, and all classes were video-recorded for post hoc analyses." (p. 506)
    • Relevant Quotes: 1) "In order to ensure consistency between different instructional treatments, the researcher always sat at the back of the classroom as an observer, and all classes were video-recorded for post hoc analyses." (p. 506) 2) "The instructor received four 1-hour training sessions over 2 days. After the researcher carefully explained the purpose and procedure of each FFI task and the recast treatment with an actual teaching demonstration, the instructor practiced until he felt comfortable with the procedure." (p. 510) 3) "To examine how often students received recasts and repaired their utterances, the researcher watched 12 hours of video recordings of all the FFI-recast classes (4 hours × 3 classes), and analyzed the instructor's recasts and students' repairs." (p. 509) Detailed Analysis: This is a single-author paper in which the same researcher designed the FFI/recast treatment, trained and coached the instructor on how to deliver it, personally observed every class session, coded the video recordings, and carried out the statistical analysis. No external or independent evaluator, agency, or blinded assessor is mentioned at any stage of implementation or analysis. Criterion I is therefore not met.
  • Y

    Year Duration

    • Since criterion T (Term Duration) is not met, criterion Y is automatically not met; the study tracked outcomes for only two weeks after a four-hour intervention.
    • "A posttest session took place two weeks after instruction." (p. 506)
    • Relevant Quotes: 1) "two experimental groups received four hours of instruction..." (p. 506) 2) "A posttest session took place two weeks after instruction." (p. 506) Detailed Analysis: Per the ERCT specific instruction for criterion Y, if criterion T is not met, Y is not met as well. In addition, the actual duration here (four hours of instruction, two weeks to posttest) is nowhere close to 75% of an academic year. Criterion Y is not met.
  • B

    Balanced Control Group

    • The control group received the same amount of instructional time from the same instructor, differing only in the absence of the FFI/recast pronunciation focus.
    • "a control group received comparable instruction in terms of length and nonpronunciation (argumentative skills) content, without any focus on form regarding pronunciation." (p. 499)
    • Relevant Quotes: 1) "a control group received comparable instruction in terms of length and nonpronunciation (argumentative skills) content, without any focus on form regarding pronunciation." (p. 499) 2) "The same instructor taught all three groups and provided recasts only to the FFI-recast group..." (p. 506) 3) "The students in the control group received instruction on English argumentative skills which did not include any FFI component on /ɹ/ but was identical in all other aspects. That is, all target words in the instructional materials were replaced with comparable words ('rain' → 'snow,' 'run' → 'jog'). For the warmup games, they did other purely communicative activities with no emphasis on pronunciation or listening practice." (p. 512) Detailed Analysis: Applying the Criterion B decision tree: no extra time or budget was given to the intervention groups relative to the control group — all three groups received the same four hours of instruction from the same instructor. The only difference was the content and focus of that instruction (pronunciation-focused FFI/recast material for the experimental groups vs. parallel argumentative- skills material with target words swapped for non-target equivalents for the control group), not the quantity of time, materials, or adult attention provided. Since no additional resources (time, budget, or materials) were present in the intervention condition, the balance requirement is trivially satisfied under the decision tree's first branch, independent of any integral- resource argument. Criterion B is met.

• Level 3 Criteria

  • R

    Reproduced

    • No independent replication by a different research team is mentioned in the paper or found via subsequent literature search; the study extends the author's own earlier work.
    • "Saito (2013) and Saito and Lyster (2012) set out to investigate how a range of FFI techniques, including recasts, can promote Japanese learners' abilities to produce /ɹ/..." (p. 504)
    • Relevant Quotes: 1) "Given the significant absence of research attention in instructed L2 speech learning and the tremendous difficulty in acquiring the English /ɹ/, Saito (2013) and Saito and Lyster (2012) set out to investigate how a range of FFI techniques, including recasts, can promote Japanese learners' abilities to produce /ɹ/..." (p. 504) 2) "Some acoustic information of the production data by the FFI-recast group (i.e., F3) has already been partially reported in another study (Saito, 2013)." (p. 507) Detailed Analysis: The paper's precursor studies (Saito, 2013; Saito & Lyster, 2012) were conducted by the same author, so they do not constitute independent replication. No quote in the text references a subsequent independent replication of this specific study by a different research team. Internet Search Verification (2026-07-30): A citation- graph search of papers citing this article (via Semantic Scholar, ~65+ citing works reviewed, spanning 2013-2025, including recent recasts/L2 speech studies such as Zhang & Liao (2023) "The role of auditory processing in L2 vowel learning: evidence from recasts," Li (2023) "A Comparison of Perception-Based and Production-Based Training Approaches," and Kang & Staples (2025) "Transdisciplinary Intersections in Second Language Pronunciation Learning and Teaching") found no paper that identifies itself as an independent replication of this specific Saito (2013) FFI-recast design (three-group /ɹ/ perception and production study with adult Japanese learners at an Osaka language institute) conducted by a different research team in a different context. Citing papers reference this study's findings and methodology but do not attempt to reproduce its specific experimental comparison. Criterion R is not met because no independent replication of this specific study was found either in the paper itself or via subsequent literature search.
  • A

    All-subject Exams

    • Only the single phonetic feature /ɹ/ was assessed, not all main subjects, and criterion E (a prerequisite) is also not met.
    • "This article reports a quasi-experimental study which investigated the efficacy of recasts together with form-focused instruction (FFI) on the development of the speech perception and production of /ɹ/..." (p. 502)
    • Relevant Quotes: 1) "This article reports a quasi-experimental study which investigated the efficacy of recasts together with form-focused instruction (FFI) on the development of the speech perception and production of /ɹ/ by Japanese learners of English..." (p. 502) 2) No quote in the paper indicates assessment of any subject or skill beyond the single phonetic contrast /ɹ/-/l/. Detailed Analysis: Per the ERCT specific instruction, criterion A requires criterion E to be met as a prerequisite; since E is not met (no standardised exam was used), A cannot be met either. Independently, the outcome measures target only one narrow phonetic feature (production/perception of /ɹ/), not a broad set of main subjects. Criterion A is not met.
  • G

    Graduation Tracking

    • Since criterion Y is not met, criterion G is automatically not met; there is no tracking toward graduation, only a two-week posttest, and no follow-up publication tracking this cohort was found.
    • "A posttest session took place two weeks after instruction." (p. 506)
    • Relevant Quotes: 1) "A posttest session took place two weeks after instruction." (p. 506) 2) No follow-up beyond the two-week posttest, and no mention of any subsequent follow-up publication tracking the same cohort, is found anywhere in the paper. Detailed Analysis: Per the ERCT specific instruction, if criterion Y is not met, criterion G is not met. In addition, there is no description of any tracking of participants toward graduation or any longer-term follow-up; the study concludes with a single posttest two weeks after instruction. Internet Search Verification (2026-07-30): A search of the citing-literature graph for this paper (via Semantic Scholar) found no subsequent paper by the same author (Kazuya Saito) or others that tracks this specific cohort of 45 adult EFL learners toward any long-term or graduation-level outcome. Given that participants were adult learners at a private language institute rather than students in a K-12 or degree program, "graduation tracking" in the ERCT sense has no natural endpoint here either. Criterion G is not met.
  • P

    Pre-Registered

    • No mention of a pre-registered protocol, registry ID, or registration date appears anywhere in the paper, and no such registration was located via search.
    • Relevant Quotes: No quotes referencing pre-registration, a study registry, or a published protocol were found anywhere in the paper, including the Method, Notes, and References sections. Detailed Analysis: This 2013 paper contains no statement about pre- registering hypotheses, methods, or planned analyses before data collection. Internet Search Verification (2026-07-30): Pre- registration of study protocols was not standard practice in second language acquisition/applied linguistics research in 2013, and no search results (via the paper's citation record and metadata) surfaced any pre-registration entry associated with this study. Criterion P is not met.