DA has theoretically and experimentally been proven to be a valuable diagnostic tool primarily in psychological research and, later in educational research focused on teaching and learning of different school subjects (e.g., mathematics, physics). However, it has not been long before language educators have begun to examine its pedagogical applications of DA (e.g., Ableeva, 2008; Antón, 2009; Lantolf & Poehner, 2004; Poehner, 2005). This initial L2 research argues that assessment-with-mediation brings assessment and instruction together into an organic unity whereby learning is the consequence of mediation, which is then internalized and becomes accessible to be used later in similar contexts. DA is a connection between assessment and learning is grounded in SCT principles. What DA of SCT framework presupposes is that a learner’s independent performance can only capture a part of their abilities.

Engagement with learners by providing them a variety of prompts, leading questions, and demonstrations when learners encounter difficulties, and observing the extent to which learners respond to this support (scaffoldings) provides insights into the potential abilities that are still forming (Vygotsky, 1987). Vygotsky (1978, 2012) referred to this conceptual slot of learners’ potential abilities as the Zone of Proximal Development (ZPD) and noted that it was the optimal target for instruction to interfere in and guide the learners through with necessary scaffoldings throughout their development. DA comprises a range of procedures that researchers in various fields have formulated to identify the appropriate ZPD of the target learner (Lidz & Haywood, 2014; Sternberg & Grigorenko, 2002). From the outset of introducing DA to second language acquisition (Kozulin & Garb, 2002; Poehner, 2008), DA has been adopted often as a one-shot procedure. While an initial DA was used to determine each learner’s ZPD, subsequent instruction was provided, taloring to learners’ emerging abilities and needs. DA may be employed to diagnose learners’ needs and to inform instruction aimed at fostering learners’ development. DA studies fall into either an interventionist or interactionist approach (Lantolf & Poehner, 2004). While interventionist DA follows procedures in which mediation or scaffoldings in various forms are scripted in advance and provided during the procedure in a standardized manner. In interactionist DA, on the other hand, mediators follow a general principle of beginning with support/scaffolding that is implicit so that they can determine whether the support or scaffoldings given is adequate to prompt a conducive response from the learners. Then it becomes more explicit only depending upon the learners’ difficulties. To date, interactionist DA has been favored in classroom contexts and in one-to-one administrations where the goal is to optimally respond to learner difficulties and subsequently identify the ZPD.

In the case of DA in second language acquisition, studies have been conducted with learners in various levels ranging from primary to university learners (Antón, 2009; Davin & Donato, 2013), with target skills including reading (Poehner & van Compernolle, 2013), listening comprehension (Ableeva, 2010). In the second language writing, there has been relatively more interest in linking assessment with second language writing and locating assessment as an indispensable complement to effective writing instruction (e.g., Crusan, 2010; Cumming, 2009; Hamp-Lyons, 2003; Weigle, 2007). Attempts to make connections between assessment and the teaching and learning of second language writing have been taken on, using a few theoretical as well. Several studies have applied DA to contexts of L2 writing, most notably (Anton, 2009; Kushki, Nassaji, et al., 2022; Kushki, Rahimi, et al., 2022; Nassaji et al., 2020; Shrestha & Coffin, 2012). For example, taking a comparative, Nassaji et al. (2020) and Kushki, Nassaji, et al. (2022) administered both interactionist DA and interventionist DA with EFL learners in an Iranian university. Nassaji et al. (2020) found that interactionist DA afforded more diagnostic insights into the learners’ emerging writing abilities than interventionist DA. Kushki, Nassaji, et al. (2022) further reported that interactionist DA was more effective in fostering learner argumentative writing abilities over time.

As pedagogical frameworks, including learning-oriented language assessment (LOLA; Hamp-Lyons, 2017) and formative assessment (Barlow et al., 2007; Fernando, 2018). DA was also has been explored as a whole-class procedures (Poehner et al., 2018) or administrated via computers (Shrestha, 2020).

All in all, DA research on L2 writing has explored the role of the assessment in providing an in-depth and progressive view of learners’ evolving abilities and has provided empirical evidence that the procedures mediated learning supported learners’ development in terms of diminishing number and phases of support or scaffolding provided. Given that most of these studies have been administered short period of interventions, whether such an intervention is sufficient to elicit developmental changes in learners’ language abilities and whether the progress that learners demonstrated persist over time.

Second language acquisition has sparsely been examined in terms of its potential intersection with SCT. More specifically, the current lack of diagnostic assessment in language instruction and the development of listening proficiency has hardly been put in perspective. That is, the potential extension of the traditional understanding of listening assessment in foreign language contexts to the application of DA to the development of learners’ listening ability has been far and between (e.g., Aleeva, 2010; Roohani et al., 2018). DA is grounded in the Vygotskyan concept of ZPD and prescribes mediated instructor-learner dialogue during the assessment. The study by Abeeva (2010) investigates the effects of dynamic assessment on improving listening comprehension of intermediate university students learning French as a foreign language. The study compared the results in a traditional test of listening comprehension. The study demonstrated that DA, due to the mediated dialogue, shed lights on the sources of unsuccessful performance that usually go unnoticed during conventional assessments, which are non-dynamic in nature. DA, on the other hand, is able to inform the instructional process regarding specific areas where learners need improvement and during that process, it allows for appropriate intervention to help learners tackle and overcome these problems. The results of the study indicate that, through dialogic interactions in the ZPD, DA allows an instructor (moderator) to ascertain not only the actual proficiency level of learners’ listening ability but also to diagnose the potential level of their listening development, while at the same time promoting this development through problem-contingent scaffoldings.

One of the rare studies of the role and influence of DA in second language listening was conducted by Zhang (2023) based on the construction of a model of mediation grounded in DA for EFL listening comprehension. The model was implemented in a 1-semester comparative experimental study in 2 parallel classes of Grade 8 in a middle school in China. For mediation, the experimental class (EC) received four DA sessions adjusted to the learners’ emerging abilities at all listening stages. Statistical results showed that students in EC significantly outperformed those in the control class in both a posttest and a near transfer assessment test, a more complex task intended to assess development in a more accurate manner. Although they showed no significant edge in the delayed test, microgenetic analysis of the development of nine students demonstrated that the mode of mediation used during enrichment allowed the teacher to identify students’ listening problems, including their struggles with phonological, lexical, grammatical, strategic issues.

Two tertiary-level students from a general English (Movie Talk) class participated in the present study. The instructor and researcher met with each student via Zoom once a week over 6-week period. Movie Talk was geared towards providing students with learning opportunities to practice listening comprehension and critical thinking using English-medium films. The first female student, Hayoon (aged 21) was a sophomore majoring in sports management. She had an experience of a listening comprehension class in the previous semester. Hoon (aged 23), a third-year male student was majored in business management at the time of data collection. To determine their overall listening comprehension, they took an IELTS listening sample test offered by the British Council. Hayoon scored 32 points out of 40 and Hoon scored 22 in the listening comprehension test.

For the six Zoom-mediated sessions, a clip from TED Talk (www.ted.com) was selected by the instructor considering the participants’ level of English and topic of interests and issues relevance to their lives. The clip chosen was titled Why Domestic Violence Victims Don’t Leave presented by Steiner (2012). The clip revolved around domestic violence that the speaker underwent and the process how she was led into being victimized by the partner. The clip was chosen from a pool of TED Talk selections by the instructor/researcher and another experienced EFL listening instructor. The selection was based on the frequency of predicted difficult or unknown words and the comprehensibility of the audiovisual clip. The instructor and researcher went through 15 clips from TED talks on a variety of topics including psychology, environment, education, literature, international affairs, and medicine. Then these clips were ordered in terms of predicted comprehensibility and predicted frequency of unknown words. The two instructors produced separate lists of orders of the 15 TED talk clips and later discussed and resolved their differences. The clip finally opted for was placed in the 10^th in terms of the number of predicted unknown words (the higher, the higher the frequency of difficult words) and the third in terms of comprehensibility (the higher, the more difficult). The selected clip for the present study was then sectioned into several approximately 15 to 25 second-long portions deemed retrievable for their short-term memory capacity. To control for the potential effects of visual clues, the clip selected for the present study involved only one monolingual speaker who recounts her story about the domestic violence.

A total of 12 Zoom-mediated sessions were held over 6-week period (six sessions per each student) during the semester. Each session lasted between 45 minutes and an hour. Each portion of the audiovisual clip was played by the instructor/researcher twice and the participant was supposed to tell what they comprehended. Note-taking was allowed, and it was used as additional data to delve into what the participants focused on while listening.

Upon listening to the target portion, the instructor/researcher asked what the portion was about and listened to what the participant comprehended. Once the participant signaled an obstacle in comprehension, the instructor/researcher provided a general and comprehensive scaffolding and induced comprehension. However, the provided scaffolding did not help get the participant through an unknown word, more specific scaffolding tailored to the problem in question was given. The number of occurrences of interactionist scaffolding provided to the participant varied within and across individuals since lexical level and overall listening comprehension of them were different. Table 1 summarizes the procedure. The interactionist dynamic assessment sessions were all conducted in Korean, the participants mother tongue.

A total of 6-hour long Zoom recordings were first transcribed verbatim by the researcher. The transcribed dialogue between each participant and the instructor/researcher was read through several times in an exhaustive manner. The episodes were then identified and further analyzed in terms of the number of successful and unsuccessful interventional scaffolding were counted. Each episode of the participants was then analyzed in terms the lexical problem the participants were faced with, and the incremental scaffolding provided in accordance to whether or not the participants successfully made inferencing of the word(s) and in turn reached correct comprehension.

The number of correct inferencing and incremental scaffolding are summarized in Table 2. The number of interventional DA episodes that Hayoon made ranged from 2 to 4 (out of 4 to 6). As presented in Table 2, while Hayoon demonstrated increasing pattern in the number of successful inferencing when provided and induced by interventional scaffolding upon her lexical problems, Hoon remained at a similar level (1 or 2) throughout the 6-week period. The result appears reasonable given the fact that Hayoon showed better performance (30 out of 40 points) than Hoon (19 out of 40 points) in IELTS listening test administered in the beginning of the data collection. It may be a legitimate consequence since Hayoon already had a certain repertoire of listening comprehension strategies that enabled her to overcome and resolve lexical issues. On the other hand, Hoon, who had lower level of listening proficiency, did not strategic repository which includes inferencing skills.

Table 3 below represents the number of interventional DA episodes of dialogic interaction with Hayoon that ended up in successful inferencing. The number of interventional scaffoldings varied from 1 to 4, with 1 standing for the most general and comprehensive scaffolding and 4 referring to the scaffolding that began with general and developed into considerably direct and containing near-to-answer scaffolding. The interventional DA dialogue with Hayoon demonstrated certain pattern of changes. That is, as sessions progressed, the dialogue involved fewer number of scaffoldings, which means that Hayoon became capable of inferencing unknown or uncertain word(s) more effectively with less specific scaffolding.

On the other hand, as Table 4 briefs, notwithstanding considerably fewer number of successful inferencing episodes, the interventional scaffolding provided for Hoon remained relatively high (4, meaning considerably specific scaffolding) across sessions. It means that Hoon was able to successfully infer words only when near-to-answer specific scaffolding was provided across sessions. It seems that listening proficiency might have interfered with one of the essential skills of inferencing.

The results seem to be in line with those of previous research findings especially in L2 writing (e.g., Kushki, Nassaji, et al., 2022) that reported that interactionist DA was more effective in fostering learner argumentative writing abilities over time.

This section addresses the type of inferencing that the two L2 listeners during the interactionist DA.