1 Introduction

Given that vocabulary knowledge serves as the foundation for developing second language (L2) knowledge in general [1, 2], a great deal of research in L2 education has targeted finding effective methods and pedagogical techniques for facilitating the development of vocabulary knowledge [3,4,5,6,7,8]. Despite significant progress in this area, vocabulary learning has remained one of the significant challenges that EFL (English as a Foreign Language) learners confront during their L2 literacy development [9]. In recent years and with the prevalence of portable devices for language learning [10], Mobile Assisted Language Learning (MALL) has gained increased attention in L2 acquisition research [11,12,13]. Relatedly, the application of MALL has been particularly an effective instructional approach in the context of L2 vocabulary learning [14, 15].

MALL provides learners abundant opportunities to interact, use language to meet everyday communication needs, and engage in different cultural experiences [16,17,18]. The expanding body of knowledge on technology-assisted vocabulary learning indicates that the implementation of mobile devices and associated technologies generally leads to improved learning outcomes compared to using traditional materials [14, 19,20,21,22]. Additionally, the application of various digital technologies for vocabulary learning has been associated with increased learning enjoyment and motivation [23], collaboration, interaction, and improved performance among language learners [24]. Moreover, it has been argued that MALL provides learners with significant affordances that extend beyond the physical limitations of the classrooms [23]. Despite these developments, the implementation of mobile devices for vocabulary learning beyond the classroom and self-regulated learning (SRL) conditions has remained relatively less explored. The current study aimed to address this gap and explored the impacts of mobile-assisted SRL on learning academic vocabulary among university students. The following sub-sections turn to the importance of academic vocabulary for university students and situate the current study within the literature of self-regulated vocabulary learning.

2 Literature review

2.1 Academic vocabulary

In a well-established model for describing and categorizing vocabulary for pedagogical purposes in language teaching programs, the lexicon of English has been divided into four main groups, including high-frequency (general service), academic, technical, and low-frequency words [25, 26]. Defined as vocabulary items used with relatively higher frequency in academic discourse (rather than non-academic texts) [27], academic vocabulary is employed by researchers for describing abstract ideas and processes in science [28], and also for the rhetorical organization of academic writing [29, 30]. Research on linguistic aspects of academic discourse has indicated that these medium-frequency words constitute a considerable proportion of academic texts (such as textbooks and research articles) that range from 10 to 14% of all words [27, 31]. Some widely used academic words in English include analyze, approach, area, assess, assume, authority, available, benefit, concept, and consist [31].

Research on identifying and teaching academic words is essential for several reasons. First, it has been strongly argued that knowing such words is necessary for university students’ academic literacy development and professional identity construction [30, 32]. Second, research has shown that learning and using academic vocabulary for EFL university students poses on them a significant learning burden [33, 34]. Third, a growing number of studies in recent years have indicated that there is a considerable variation in using academic words across disciplines [35, 36], and such findings underscore the need for addressing academic vocabulary learning needs in light of their disciplinary practices [37]. Accordingly, although several core academic word lists are available for a systematic focus on teaching academic vocabulary to university students [27, 31], the presence of a large number of words in those lists necessitates the need for more fine-tuned instructional interventions and materials for teaching academic words [20].

2.2 Self-regulated vocabulary learning

The concepts of self-directed learning (SDL) and self-regulated learning (SRL) are sometimes used interchangeably in the literature; however, as explained below, these modes of learning are different concerning the responsibility assumed by the learners. According to Brandt [38] SDL “represents a process of learning that is individual, purposeful, and developmental” (p. 3). More specifically, in its broadest sense as a basic human competence, SDL is a learning process in which taking the initiative (individually or with the help of others), identifying personal needs, setting achievement goals, collecting required resources, employing effective strategies, and evaluating outcomes are undertaken by students [39]. Additionally, a key component in SDL is that learners determine the environment for learning [38]. Nevertheless, an important distinction between SDL and SRL is that in SRL, the content to be learned is selected by teachers [40]. In this regard, the SRL theory is more concerned with factors involved in dealing with pre-specified learning content, such as processes of metacognition, motivation, and strategic action [41, 42]. Training students to become self-regulated has been regarded as a fundamental goal of modern education [43], and there is also a growing recognition that SRL is a significant component of L2 education that results in better learning outcomes [44, 45].

Over the past years, there has been a surge of interest among researchers in exploring the relationship between SRL and learners’ vocabulary acquisition [46,47,48,49]. This line of inquiry has highlighted that the ability of learners to engage in SRL significantly affects their performance in learning English vocabulary autonomously. For example, [46] explored how an English vocabulary learning app with a self-regulated learning mechanism assisted learners in improving their SRL abilities, intending to enhance their learning performance and motivation in a MALL context. The study findings indicated that learners with the SRL mechanism exhibited significantly better learning performances and motivation. Moreover, [49] explored how motivational factors and using learning strategies work in tandem to influence L2 vocabulary knowledge among 230 Korean high-school students. The study findings revealed that motivation significantly affects EFL learners’ vocabulary knowledge through the mediation of vocabulary learning strategies. Additionally, [47] developed an instrument for assessing the self-regulatory capacity of language learners in vocabulary learning in information and communication-based environments. The study findings indicated that self-regulated use of technology could serve as a valid and reliable measure of learners’ SRL capacity. In the same line, [48] explored the effects of self-regulated vocabulary learning processes on learners’ self-efficacy. Drawing on questionnaires and a vocabulary test, the study findings indicated that the SRL process could boost learners’ self-efficacy and increase their vocabulary knowledge. In other words, the process of self-regulated vocabulary learning not only increases self-efficacy, but it also leads to the development of vocabulary knowledge.

The above body of knowledge highlights that the process of self-regulated vocabulary learning influences the degree of learners’ self-efficacy and motivation, which in turn leads to the growth of learners’ vocabulary knowledge. In addition, the existing literature foregrounds that SRL ability significantly affects learning performance in an autonomous learning scenario, particularly the ability to learn English vocabulary [7, 8, 20,21,22]. Nonetheless, using the affordances of mobile devices for scaffolding academic vocabulary development is a relatively under-researched area [21]. Although previous studies have provided empirical evidence for the effectiveness of SRL using traditional materials, the contribution of emerging technologies, such as digital flashcards as an effective strategy for SRL, has remained less explored [20]. In this regard, developing university students' SRL abilities is critical in helping them plan and learn academic vocabulary in a technology-assisted learning environment. As SRL encourages students to be actively responsible for their learning, it involves goal setting, strategy implementation, self-monitoring, and self-adjustment [50], which can significantly affect their achievement. Therefore, the present study aimed to explore university students’ academic vocabulary learning in a mobile-assisted SRL intervention and compared the learning outcomes with traditional learning strategies. In so doing, the following research questions were addressed:

  1. 1

    Does mobile-assisted self-regulated learning result in improved academic vocabulary knowledge among university students?

  2. 2

    Which intervention (mobile-assisted learning with digital flashcards or learning from word lists) results in learning more academic words?

3 Methods

3.1 Participants

A convenience sampling procedure was used to recruit 49 Iranian university students majoring in Teaching English as a foreign language (TEFL) as the study participants. The mean age of the participants was 23, and they were selected based on their availability in the research context. Their proficiency level was examined using the Cambridge quick placement test [51], which indicated that the majority of them were at the B2 level (upper independent English level) based on the Common European Framework of Reference for Languages (CEFR) standards [52]. Following [53], the assignment of the participants to the experimental and control groups was based on their preferences for using digital or paper-based materials for self-regulated and outside-class vocabulary learning. Accordingly, 28 students were assigned to the experimental group (i.e., e-flashcards for mobile devices), and 21 preferred to use traditional materials (i.e., word lists). Informed consent was obtained from the participants, and they were briefed on the nature and goals of the study, data collection procedures, and confidentiality of their personal information.

3.2 Materials and testing instruments

3.2.1 Corpus-based word lists

Target words for the present study were selected from the list of frequently-used academic vocabulary in applied linguistics research articles [54]. The list was extracted systematically from a 15-million words corpus by applying frequency, range, and specialized use criteria. Accordingly, to be considered frequently used items, academic words had to occur more than 30 times per million words in the corpus and in at least 500 research articles. Additionally, the selected items were beyond general vocabulary in English. As represented below, 70 words in seven sets were used in creating digital flashcards and word lists for self-regulated learning of academic vocabulary among the participants. Given that these words were unknown for most students (see the following sub-sections), learning them contributes significantly to applied linguistics students’ comprehension of academic discourse.

3.2.2 Digital flashcard application

The present study used the Lexilize application (https://lexilize.com/) as the primary learning environment for the participants in the experimental group. The application provides learners with some affordances for vocabulary learning, such as importing words from Excel files, Spaced Repetition System, automatic pronunciation of words, supporting different languages, and working without an Internet connection. Moreover, the application incorporates gaming features for reviewing words, recalling their meaning, selecting their translation, and typing the word.

3.2.3 Vocabulary knowledge tests

In order to measure the knowledge of the target academic words before and after the treatment, the study used vocabulary knowledge scale (VKS), which is a widely employed self-assessment tool for measuring the depth and breadth of vocabulary knowledge [55, 56]. In addition to identifying whether they know a specific term, the VKS asks learners to rate their knowledge of the target vocabulary on a 5-point scale. Following this, participants must offer further proof of the correctness of their self-assessment by translating or using the target term in a phrase and in a proper context. Figure 1 shows a representative item from the VKS test as utilized in this study.

Fig. 1
figure 1

A model test question created using Wesche and Paribakht’s VKS [55]

Full size image

3.3 Procedure and data analysis

The study was conducted over an academic semester that lasted for 15 weeks. Data collection was started by administering three VKS tests designed based on the 360 frequently-used academic words in applied linguistics. Accordingly, the participants completed three diagnostic tests, each containing 120 items over three weeks. After analyzing the results, 70 words (Table 1) were identified as mostly unknown academic words for the majority of the participants (i.e., more than 80% of them). During the following seven weeks, the participants studied target academic words using different materials. More specifically, those in the experimental group used the Lexilize application, and those in the control group were given word lists. The content presented to both groups was similar; the only difference was in the learning medium. The post-test was administered during the eleventh week of the semester. In order to investigate the delayed impacts, another test was administered after four weeks.

Table 1 Target academic words
Full size table

The data in the form of participant scores on word knowledge tests was analyzed using IBM SPSS software (version 25). To that end, descriptive and inferential statistical methods were used. In descriptive statistics, mean values, standard deviations, standard errors of the mean, and 95% confidence intervals for means were obtained for VKS scores on both pre-, post-, and delayed post-tests. Regarding inferential statistics, mixed between-within-subjects analysis of variance (ANOVA) [57] and follow-up posthoc tests were used. This statistical test allows combining between- and within-subjects effects and accordingly provides a more reliable analysis as compared to t-test, which is normally used for comparing two groups. Additionally, as the study measured the students’ vocabulary knowledge in three different times, ANOVA is a more robust statistical test for studies with repeated measurements [57]. The independent variable was the group, which had two levels (experimental and control), while the dependent variable was VKS, which had three levels (time 1, 2, and 3). This analysis indicates whether there are discernible main effects for each independent variable and if there is a significant interaction between them [57].

4 Results

The results of the vocabulary tests administered to the participants are summarized in Table 2. As represented below, the mean value for scores obtained on the pre-test was 43.64 (SD = 4.39) for the experimental group. The control group had a mean score of 41.57 (SD = 4.28) on the same test. To check for any pre-existing differences before the treatment, the scores obtained on the pre-test were compared using an independent samples t-test, which pointed to no significant differences among the two groups, t (47) = 1.65, p ≤ 0.105 (two-tailed). Although the mean scores were not significantly different in the pre-test, those participants in the experimental group scored higher in the post-test (M = 119, SD = 12.94) compared to the control group (M = 110, SD = 10.45). Similar results were obtained on the delayed test, and the experimental group had a higher mean in VKS (M = 118.72, SD = 11.52) than the control group (M = 107.29, SD = 11.60).

Table 2 Descriptive statistics for the scores obtained on VKS
Full size table

The scores were further analyzed for within-subject effects (i.e., time). Prior to examining the main effects, preliminary analyses, including Levene's test of equality of error variances (assumption of homogeneity of variances) and Box's test of equality of covariance matrices (assumption of the equality of covariance matrices) were performed, and no violations of the assumptions required for ANOVA were observed. Moreover, the results of the multivariate analyses (Table 3) indicated a significant interaction effect between time and group (Wilks’ Lambda = 0.85, F (2, 46) = 3.95, p = 0.026, ηp2 = 0.147). This interaction effect indicates that the experimental and control groups had differing changes in their scores over time (see Fig. 2 below), and the effect size for the differences was large. In addition, the results revealed a statistically significant main effect for time (within-subjects variable), Wilks’ Lambda = 0.23, F (2, 46) = 983.20, p ≤ 0.001, ηp2 = 0.97. As a consequence, the changes in scores throughout the three testing periods were statistically significant, and the data indicated a relatively large effect size for the observed differences [58].

Table 3 Multivariate testsa (Wilks’ Lambda)
Full size table
Fig. 2
figure 2

Visual representation of the mean scores over time

Full size image

The next step in data analysis was investigating the between-subjects effects (i.e., grouping variable) (Table 4). Considering the significant interaction effect among the time and group variables, the scores obtained by the participants were compared across three testing times. The results showed that although both groups improved their vocabulary knowledge over time (i.e., within-subjects effect), the experimental group outperformed the control group in post- and delayed post-tests.

Table 4 Tests of between-subjects effects
Full size table

A visual illustration of the participants’ scores over time is shown in Fig. 2. This figure is useful in understanding the primary effects of the two variables. As shown in the chart, both groups improved their vocabulary knowledge considerably from the pre-test to the post-test, and the experimental group obtained higher scores as reflected in their overall mean score. From the post-test to the delayed test, the mean values for the scores dropped only slightly, with more noticeable changes in the scores obtained by the control group.

5 Discussion

The first research question explored the impacts of mobile-assisted SRL on academic vocabulary development among EFL university students. The findings, as summarized in Tables 2 and 3, provided empirical evidence for the effectiveness of mobile-assisted SRL for developing academic vocabulary knowledge. More specifically, the participants in both groups learned a considerable proportion of the target academic words from the pre-test to the post-test, and the results obtained on the delayed post-test were significantly better than the pre-test. The findings align with earlier studies that reported positive learning outcomes for self-regulated vocabulary learning in L2 acquisition [46,47,48,49]. Additionally, the findings of the current study are congruent with the studies that investigated SRL for academic vocabulary among university students [20, 21]. There might be a number of factors contributing to the effectiveness of SRL for learning academic vocabulary as observed in this study. First, given the nature of the target vocabulary items, participants in both groups considered academic words relevant to their needs and invested time and energy in learning them. Accordingly, they had metacognitive awareness with respect to the role of academic words in their field of study, which contributed significantly to their overall achievements [41]. Second, the participants employed intentional learning mechanisms for learning academic words, and there is ample evidence for the effectiveness of intentional vocabulary learning over incidental learning processes that need long periods and substantial amounts of input [1, 9]. This intentional learning might be considered strategic action within the SRL theory that resulted in long-term gains among the participants [41, 42].

The second research question compared the two strategies employed for SRL of academic words among the participants. The findings indicated that mobile-assisted SRL is more effective for learning academic words compared to traditional strategies such as learning from word lists. The participants in the experimental group outperformed the control group in the post-test and delayed test. This finding is congruent with studies that compared learning gains from mobile-assisted and traditional materials [20, 21]. In this regard, the findings extend our understanding with respect to the functionality of mobile-assisted learning of academic words in SRL environments. These results might be interpreted in light of the following considerations. First, as the mobile-assisted learning condition involved using digital flashcards with spaced repetition technology, the participants in the experimental group learned target words systematically and had more chances to review and recycle them. This impacted their overall learning and resulted in higher scores on vocabulary tests. Second, as highlighted by earlier studies, introducing new technologies for learning is generally associated with increased motivation [59], and increased motivation (or engagement with learning materials) is a strong predictor of success in language learning [60]. Third, considering the affordances of mobile devices for extending vocabulary learning to anytime and place [14], mobile-assisted learning materials were easily accessible to the participants in the experimental group, and this might have resulted in more interaction with content and eventual learning gains.

The study findings also have some implications for language teachers in academic contexts, university students, and English for Academic Purposes (EAP) programs. First, mobile-assisted SRL with digital flashcards has a significant potential for addressing university students’ academic vocabulary learning needs and might be considered an effective strategy with considerable learning outcomes. Accordingly, language teachers are encouraged to incorporate such strategies into their instructional repertoire by providing learners with appropriate materials and supporting them in undertaking SRL to develop their knowledge of academic words. Second, considering both the short- and long-term effectiveness of mobile-assisted SRL, university students can address their vocabulary learning needs using the available resources. Currently, there are a number of systematically developed lists for general and discipline-specific academic words [30, 61], and with freely available digital flashcard applications (similar to the one used in this study), university students might be able to create their flashcards. Finally, as most EFL students are struggling with building a vocabulary foundation for EAP courses, mobile-assisted SRL seems to have significant potential for bridging the vocabulary gap for those students [62, 63]. In this regard, EAP programs might benefit from systematically integrating mobile-assisted vocabulary learning beyond-the-classroom instructional programs for scaffolding vocabulary learning in short-time periods. Earlier studies have indicated that such interventions are helpful for learning a substantial number of words by EFL students [7, 8, 64, 65], and the growing evidence for the effectiveness of such platforms for learning academic words provides EAP programs with new opportunities for supporting university students [21].

6 Conclusion

The current study explored the effectiveness of mobile-assisted SRL in developing EFL university students’ academic vocabulary knowledge. The findings indicated that mobile-assisted SRL with digital flashcards contributed significantly to learning academic words and resulted in both short- and long-term gains. The findings also provided further empirical evidence for the effectiveness of mobile-assisted vocabulary learning compared to traditional materials (such as learning words from word lists). The study had some limitations that should be acknowledged. First, as it is a common practice in educational research, convenience sampling procedures were employed, and the sample size was relatively small. This point should be taken into account in generalizing the findings to other EFL contexts. Second, the study explored learning 70 academic words. Considering the large number of academic words in both core and discipline-specific lists, the findings provide only a partial picture of the possible learning outcomes for addressing academic vocabulary learning needs. Accordingly, there is a need for additional studies focusing on learning a more significant number of academic words in SRL interventions. Finally, this study collected quantitative data only, and the participants’ perceptions and attitudes toward mobile-assisted SRL were not explored. As such factors are important in learning outcomes from MALL [66, 67], there is a need for more investigations using mixed methods approaches. Despite these limitations, the study sheds light on learning outcomes and the relative effectiveness of mobile-assisted SRL for addressing EFL learners’ academic vocabulary learning needs. Future studies might focus on the abovementioned gaps and also consider investigating the impacts of mobile-assisted vocabulary learning in the more extended term periods by administrating more delayed post-tests following the interventions.