Game on: Computerized Training Promotes Second Language Stress–Suffix Associations
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsPlease see the attached file.
Comments for author File: 
Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for Authors- Summary
This study investigates the impact of digital game-based training on L2 Spanish learners’ ability to associate stress patterns with verb suffixes. To this end, it examines the effects of various linguistic (lexical stress, syllabic structure, phonotactic probability) and learner variables (L2 proficiency, L2 use, working memory), all of which are considered to influence participants’ ability to detect stress-suffix associations. The study was carried out to beginner L2 Spanish learners, who were native speakers of English. They participated in a game-based activity, in which they had to choose the right verb suffix based on the stress condition of the first syllable of the word. According to the Spanish verb system, paroxytone stress was expected to be associated with present tense suffixes and oxytone with past tense suffixes. Additionally, they participated in a series of pre-tests and post-tests, which measured the linguistic and learner factors. The data were analyzed with generalized linear mixed models and generalized additive mixed models in R, to account for individual variation.
The main contribution of this paper is that it addresses the following key research questions: (a) whether game-based training facilitates the detection of stress-tense suffix associations and (b) whether working memory (WM) facilitates participants’ performance in the game. The data analysis revealed that digital training improved lexical processing in terms of reaction times (RTs) across all conditions. In contrast, higher accuracy was observed only in specific conditions (CV and oxytone verb). Regarding WM, higher WM learners had higher accuracy and slower RTs than lower WM learners in all conditions. These findings can be significant, as they indicate that L2 instruction of suprasegmental cues -often given lower priority in L2 instruction- can play a crucial role in lexical processing.
The study is interesting and innovative, as it integrates game-based training with the acquisition of stress-suffix association. In parallel, it examines the interaction of various linguistic and learner variables. Finally, it provides insights into L2 suprasegmental acquisition and supports previous findings showing that beginner L2 learners struggle to acquire linguistic forms that are present in both their L1 (Spanish) and L2 (English) but function differently, such as lexical stress.
- General concept comments
Article: It’s a very interesting work with a robust and well-conducted methodology. However, I have some concerns about the number of participants (N=20). Recognizing the practical issues of recruiting L2 participants in an educational setting and particularly for a repetitive training task, I believe that a larger sample of participants (N>30) would increase statistical power and probably would give higher effect sizes.
In the methodology section, the authors report an eye-tracking pre-test and post-test. However, its relationship to the linguistic and learner factors examined in the study, if any, is unclear. The authors briefly mention that “Participants completed an eye-tracking pre-test and post-test to determine whether the learners were able to apply what they had learned in the game to new instances. The results are reported elsewhere” (line 442-445)”. Providing more details on the aims of this task and its connection to the main research questions would enhance clarity.
According to the results, faster RTs were observed across all conditions as learners progressed in the game. However, higher accuracy with increased gameplay was only observed in the CV verb and oxytone conditions. In addition, in some conditions (e.g. paroxytone and CVC verbs) accuracy decreased with increased gameplay, but overall RTs improved. The authors attribute this contradictory finding to the low proficiency level and a possible subsequent inability of the participants to “attend to the predictive prosodic cues both accurately and quickly” (line 821-824). However, we cannot exclude the possibility of random effects due to external factors, such as lack of concentration or the emergence of systematic bias in stress assignment. This risk could be mitigated if the game incorporated techniques to periodically check participants’ attention, particularly in the final levels, which included more items. In other types of experimental tasks (e.g. priming tasks) these techniques involve the use of fillers (e.g. trisyllabic verbs or nouns) to prevent the development of patterns in the stress assignment of two-syllable verbs. However, since the task does not include such attention markers, and given the number of participants, a correlation analysis between accuracy and RTs could help determine whether the observed effects are due to random variation. If such a test has been conducted, it should be included in the results.
In the first paragraph of the Descriptive Statistics section (l. 540-554), overall results are given but there is no information on the significance of the results. In the same section, a figure presenting overall improvement in accuracy and RTs across all conditions could be introduced before the specific item type figures to enhance interpretability.
Review: The authors conduct a thorough review of previous research on linguistic prediction on L2 processing. Furthermore, they present research findings that support the selection of the linguistic and learner factors they examine in the current study. They also highlight the increasing attention on digital games as tools for language learning and identify a gap in the literature concerning game-based training for L2 predictive processing skills. At this point, given that this field is rather unexplored, it would be helpful to distinguish between training and game-based training and provide references, if any, to the impact of training on predictive skills in other contexts (e.g. instructional intervention, experimental tasks). This addition could further support the study’s importance.
Finally, I believe that a more detailed description of the Spanish lexical system would enhance clarity, particularly in the Introduction section or Linguistic Variables section. A brief mention of the stress system from a phonological perspective would be quite helpful for readers who are not familiar with the language. This discussion could include information on (a) possible stress positions across morphological classes (b) stress patterns in the verbal system (c) stress position in dissyllabic nouns (d) stress variation in verbs based on morphological factors. verbs which vary according to different aspects of verb morphology. This information would further support the study’s focus on verbal stress.
- Specific comments
- Line 20-22: I suggest rephrasing this sentence, in terms of clarity.
- Line 48: The focus on disyllabic verbs should be made clear earlier in the text. It first appears in the Material
- Line 165-168: In this section, adding a few examples would better illustrate the difference in the stress load between English and Spanish.
- Line 174-175: Is the paroxytone stress the more frequent pattern in verbs or across the whole lexicon?
- Line 183: Consider using the term “unmarked”, which is more linguistically informative, instead of “canonical”.
- Line 396: Were the participants informed about the aim and procedure of the study before providing consent to participate? This information should be included in the Participants
- Line 409: The Procedure section could provide information on the following aspects: (a) Did the procedure take place in an educational setting or in the participants’ own environment? (b) Were the participants supervised by researchers or instructors during the procedure? (c) Were they provided with written or oral instruction before the procedure?
- Line 795-797: A reference from an experimental or theoretical study would strengthen this claim.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsThe revised manuscript presents a clearer and more coherent articulation of the research background, research questions, methodology, results, and discussion. Additionally, the inclusion of pedagogical implications and practical suggestions for language instructors adds valuable insight and relevance to the study.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
