In the area of Second Language Acquisition, great attention has been paid to the possible potential of feedback in written production and its impact on the development of the learner’s interlanguage (Polio, 2012; Van Beuningen, 2010). The argument is that feedback is related to the processes of attention to form, which is crucial for acquisition (Ellis, 2016). Recently, researchers have proposed that model texts, samples of well-written texts that learners compare with their original draft (Martínez Esteban & Roca de Larios, 2010), can be used as an effective written corrective feedback (WCF) technique. Actually, there are a handful of studies in which the effects of models have been analyzed in individual (e.g., García Mayo & Loidi Labandibar, 2017; Hanaoka, 2006a, 2006b, 2007; Kang, 2020; Montealegre Ramón, 2019) and collaborative (e.g., Coyle, Cánovas Guirao & Roca de Larios, 2018; Coyle & Roca de Larios, 2014; Luquin & García Mayo, 2021; Yang & Zhang, 2010) writing or both (Lázaro-Ibarrola, 2021, Martínez Esteban & Roca de Larios, 2010). Overall, the results have revealed that models are effective for directing attention to content, lexical features, and alternative expressions and ideas, and that collaboration furthers the learning potential of the model. However, studies on collaborative writing and models as a form of WCF have been conducted with adult participants, but research with children is scarce despite the growth of early EFL learning in school settings over the past twenty years. In addition, more work is needed to elucidate whether the benefits reported in these studies only lead to greater precision in L2 writing or to language development in the long term (Polio, 2012).
Besides the internal cognitive factors that intervene in the learner’s attention and response to feedback, affective variables can also mediate in how students engage with corrections. In this sense, the notion of student commitment has received more and more attention from researchers who have analyzed cognitive responses and affective behaviors shown by students towards different types of feedback (Wigglesworth, & Storch, 2012; Tocalli-Beller & Swain, 2005). However, very few studies have explored children’s motivation toward model texts (see Lázaro-Ibarrola & Villarreal (2021) and Villarreal and Lázaro-Ibarrola (?)). Certainly, writing is a complex and time-consuming activity which requires concentration, even more so for YLs who are still developing cognitively, physically, and socially. Therefore, considering such factors when conducting research with children seems of paramount importance if we want to draw a complete and more accurate picture of what is going on during the acquisition of an L2.
The research presented here represents an attempt to bridge these gaps in the field by examining the short- and long-term effects of collaborative writing and model texts on the L2 learning of young learners, measuring their motivation along the way.
The participants were 60 11- to 12-year-old Spanish children forming a total of 30 pairs from three EFL classes randomly assigned to a control group (CG), a treatment group (TG), and a long-term treatment group (LTG). The groups were engaged in two four-stage collaborative writing cycles of 3 weeks each separated by four months. The four-stage task involved (a) noticing of linguistic problems while writing a picture-based story (Stage 1), comparison of their texts with a native-speaker model (Stage 2)/self-correction of the students’ own texts, (c) rewriting of their original output (Stage 3) and (d) delayed post-test (Stage 4). The CG did not receive the models, but self-corrected their own texts, the TG was only exposed to the feedback at two times (one per Cycle), and the LTG benefitted from this technique during the two writing cycles and the period in-between.
Given the lexical gains that the models seem to offer to the detriment of grammar, we also opted to guide the children towards formal aspects through the technique of input enhancement. We chose to work with third person singular possessive pronouns (his/her) since it is a grammatical structure that Spanish children find it difficult to differentiate. In this way, the possessive pronouns present in all the models were underlined. With the aim to collect qualitative data on the children’s attitudes and motivation toward the task, they completed an individual questionnaire before Stage 1 and after Stage 3 in both cycles. In addition, another individual questionnaire was administered and focus group interviews were held with six students from each group randomly selected after the two cycles were completed.
The results obtained show that the use of model texts brought about an increase in the number of episodes generated and a greater attention to lexical and content aspects. Nevertheless, after prolonged exposure, the models helped the children diversify their linguistic concerns and so it was reflected on their drafts, as many of the aspects noticed were fully or partially incorporated. On the other hand, self-correction promoted attention to grammar, spelling and punctuation, which reflects the benefits of writing tasks to engage in languaging.
Regarding the effect of models on draft quality, in the short term the models led the children to reduce the number of pre-clauses and proto-clauses as well as to increase the grammatical complexity of their texts, which became visible in the high frequency of subordinate clauses. After sustained exposure to the models, the children were able to produce fewer proto-clauses and more clauses, have greater lexical diversity in their texts, and make fewer mistakes. With reference to input enhancement, no statistically significant differences were found between the texts written by the CG and the TG, but some trends were observed. However, we did observe statistically significant differences within the LTG as well as when the three groups were contrasted. Namely, the children in the LTG used all third-person possessive pronouns correctly in draft 6, reaching significance with respect to their first draft. Also in draft 6, they achieved better results than the CG and the TG and this result is likely to be attributable to the sustained exposure to the enhanced input.
Finally, concerning attitudes and motivation, the use of models elicited positive responses among the participants, especially in the case of the children who had been exposed to this form of feedback for a longer time. In general, some students showed a preference for a more explicit type of error correction, but their enjoyment, improvement in vocabulary, grammar, and overall written competence, along with their enthusiasm for working collaboratively, make models a very interesting tool to integrate into the EFL classroom. In light of these results, pedagogical implications will be discussed.