Issues and arguments in the measurement of second language pronunciation
Bibliographic record
Abstract
This thesis examines systematic sources of variance in raters' judgments of second (L2) language speech, including rater cognitive and experience variables, rating scale properties, and characteristics of the speech, in order to better understand influences on raters' scoring decisions. The thesis culminates in the development of an empirically-based L2 comprehensibility scale that describes, with greater precision, the quality of speech that is characteristic at different comprehensibility levels. Study 1 examines the effect of individual differences in raters' cognitive abilities on their ratings of L2 speech. Thirty music majors and 30 non-music majors rated 40 L2 speech samples for comprehensibility, accentedness, and fluency and were additionally assessed for musical ability, phonological memory, and attention control. Results showed that music majors assigned significantly lower ratings than non-music majors solely for accentedness, particularly for low ability learners. However, phonological memory and attention control did not influence their ratings.Study 2 examines the effects of two additional sources of variance—rating scale length and rater experience—on raters' judgments of L2 comprehensibility, accentedness, and fluency. Twenty experienced and 20 novice raters judged 38 L2 speech samples using 5-point or 9-point numerical rating scales. In addition, raters' perceptions of the rating process were elicited through verbal protocols and interviews. Results showed that experienced and novice raters achieved high consensus about the highest and lowest scoring L2 speakers but had difficulty differentiating between scale levels in the absence of guidance from the rating instrument. Finally, the goal of Study 3 was to construct an L2 comprehensibility scale rooted in raters' perspectives of influences on their judgments, and characteristics of the L2 speech. To this end, 19 speech measures used to analyze 40 L2 speech samples were examined in relation to 60 raters' mean L2 comprehensibility ratings and three ESL teachers' indications of their most salient scoring criteria. Overall, a wide range of measures contributed to listeners' comprehensibility judgments, with vocabulary and fluency measures distinguishing between low-level learners, grammatical and discourse-level measures distinguishing between high-level learners, and word stress distinguishing between all levels. Taken together, these papers advance our understanding of raters' perspectives in L2 pronunciation assessment.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.423 | 0.717 |
| Meta-epidemiology (narrow) | 0.002 | 0.002 |
| Meta-epidemiology (broad) | 0.002 | 0.003 |
| Bibliometrics | 0.006 | 0.004 |
| Science and technology studies | 0.004 | 0.040 |
| Scholarly communication | 0.012 | 0.013 |
| Open science | 0.010 | 0.007 |
| Research integrity | 0.008 | 0.013 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".