Technology-enhanced social learning (TSL) to foster critical thinking dispositions and thinking in writing

Abstract

The objective of this study was to develop a Technology-Enhanced Social Learning (TSL) model that effectively cultivates critical thinking dispositions (CTD) and subject-specific critical thinking in writing (CTW) to serve as a reliable standard and point of reference for instructors in English as a Foreign Language (EFL) classrooms. The objective was to assess the efficacy of the TSL approach in enhancing Chinese EFL students’ dispositions and subject-specific critical thinking competencies while utilizing a TSL conceptual framework. A 14-week TSL intervention was administered in a quasi-experimental design comprising a treatment group (TG) and a comparison group (CG) that included a total of 60 students. The adoption of the California Critical Thinking Disposition Inventory (CCTDI) and an existing writing rubric allowed the measurement of dispositions and subject-specific critical thinking in writing. SPSS ANCOVA was used to compare the improvement of the two groups while controlling the effect of pre-existing differences. The findings indicate significant differences in overall critical thinking dispositions and subject-specific thinking in written compositions between the treatment group (TG) and control group (CG) after undergoing the TSL intervention. This demonstrates that the implementation of TSL can effectively augment the critical thinking disposition of Chinese EFL learners while also facilitating thinking in writing.

Keywords:

• Technology-enhanced social learning (TSL)
• critical thinking disposition (CTD)
• thinking in writing
• EFL writing
• constructivism
• asynchronous online discussion (AOD)

REVIEWING EDITOR:

• John Chiang, Communication and Media, State University of New York, USA

SUBJECTS:

• Bilingualism / ESL
• International & Comparative Education
• Education - Social Sciences
• Language Teaching & Learning

Introduction

It is imperative for EFL learners to develop their critical thinking (CT) and writing skills. Individuals who possess strong critical thinking abilities are more likely to have better opportunities in various domains, including academia, career, and everyday life (Franco et al., 2017). This is because critical thinking is closely intertwined with writing abilities (Akbari et al., 2018; Sharadgah et al., 2019; Susilo et al., 2021). Critical thinking disposition (CTD), as one of the main dimensions of CT, is a vital factor that influences students’ academic performance (Ali & Awan, 2021; Karagöl & Bekmezci, 2015; Shida et al., 2018).

The integration of Internet usage in education has been steadily rising as technology continues to develop, a fact that has been noted in scholarly sources (Barlow et al., 1995; Hiltz & Wellman, 1997; Gabriel, 2004; Gherwash & Paiz, 2019; Sönmez, 2021). The incorporation of digital media into computer-mediated communication (CMC) has progressively augmented the conventional education format. The academic performance of students is influenced by the learning approach (Esmaeil Nejad et al., 2022) and the type of critical thinking pedagogy employed (Abrami et al., 2014). It is recommended that the development of critical thinking be prioritized with technology-enhanced instruction to improve students’ academic writing capabilities for its flexibility in allowing instructors and students to interact with materials without the limits of time and location. For instance, there are various forms of technology-supported CT, including game-based learning through a computer-supported environment (McDonald, 2017), mobile learning, computer inquiry-based learning (Hu et al., 2019), or web-blended learning (Astuti et al., 2020).

CT is usually regarded as relative skills and dispositions. CTD, or the so-called affective dispositions, means the willingness or dispositions to think critically (Facione, 1990; Karagöl & Bekmezci, 2015; Kwon et al., 2007; Yeh & Chen, 2003). A person who has good CT skills but does not have the willingness (e.g., open-minded, honest in facing personal biases, etc.) (Facione, 1998, p. 14) to use relative CT skills (e.g., analysis, inference, evaluation, etc.) is not a good critical thinker (Facione, 1990).

When it comes to the cultivation of CT abilities, the consciousness of higher-order thinking is achieved through social interactions (Vygotsky, 1978). Also, agreed by Cole and Wertsch (1996) and Wass et al. (2011), CT is a process of construction under a social environment. Although the evidence showed that social learning and online interaction have positive effects on CT ability enhancement (Chen & Hu, 2018; Hajhosseini et al., 2016; Zhang, 2018) while most previous studies focused on CT skills rather than CT dispositions. Critical thinking skills and dispositions are two vital but different competencies that a good critical thinker needs to possess (Facione, 2020; Paul & Elder, 2012). Therefore, it is necessary to create a technology-enhanced social learning (TSL) environment to investigate whether it can influence Chinese EFL learners’ dispositions of CT and subject-specific thinking reflected in writing.

In this research, a pre-post-test quasi-experimental design was conducted with one treatment group (TG) and one comparison group (CG) under the TSL approach and traditional process writing approach, respectively, followed by two main research questions and relative hypothesis:

Q1: Does TSL perform better than the traditional writing approach in terms of the affective disposition of CT?

Null hypothesis: TSL does not perform better than traditional writing approach in terms of the affective disposition of CT.

H1: TSL performs better than the traditional writing approach in terms of the affective disposition of CT.

Q2: Does TSL perform better than the traditional writing approach in terms of subject-specific thinking in EFL writing?

Null hypothesis: TSL does not perform better than the traditional writing approach in terms of subject-specific thinking in EFL writing.

H2: TSL performs better than the traditional writing approach in terms of subject-specific thinking in EFL writing.

Literature review

Critical thinking disposition (CTD)

To enhance comprehension of CT dispositions, the researcher conducted a comparative analysis of the constructs about CT dispositions. As presented in Table 1, there was a consensus in their interpretation of concepts, such as the pursuit of evidence or truth, possession of knowledge and curiosity, impartiality, evidence-based analysis, holistic thinking, confidence, and the demonstration of fairness with humility. The standardization of these traits is consistent with Facione’s constructs for critical thinking dispositions (Facione, 2020), and it is plausible that these seven fundamental dispositions may serve as a universally recognized framework for critical thinking dispositions. An added advantage of Facione’s model is its compatibility with the California Critical Thinking Disposition Inventory (CCTDI), which provides a reliable and objective means of measuring the impact of specific interventions on students’ disposition to CT.

Table 1. Taxonomy of CT dispositions.

Ennis (1996)	Paul and Elder (2012)	Facione (2020)
Ideal critical thinkers are disposed to Seek and offer clear statements. Seek and offer clear reasons. Try to be well-informed. Use credible sources and observations. Take into account the total situation. Keep in mind the basic concern in the context. Be alert for alternatives. Be open-minded: Seriously consider alternative points of view. Withhold judgment when the evidence and reasons are insufficient. Take a position and change a position when the evidence and reasons are sufficient. Seek as much precision as the situation requires. Try to ‘get it right’ to the extent possible or feasible. Employ their critical thinking abilities.	Intellectual unfairness Intellectual arrogance Intellectual hypocrisy Intellectual conformity Intellectual self-centeredness Intellectual cowardice Intellectual distrust of reason Intellectual disregard for justice Intellectual laziness	Truth-seeking Open-mindedness Inquisitiveness Analyticity Systematicity Confidence in reasoning Maturity of judgment.

According to Facione’s (1990) suggestion, incorporating instructions aimed at promoting learners’ dispositions could effectively enhance their CT not only in terms of acquiring cognitive skills but also in fostering their willingness and ability to utilize these skills. According to the report’s citation, individuals who develop these aptitudes are more inclined to proficiently apply these skills in their daily lives and effectively address specific problems or issues. This is in contrast to those who may master these skills, but lack the motivation to utilize them (Facione, 1990, p. 13).

Therefore, this research adopts Facione’s definition of ‘a probing inquisitiveness, a keenness of mind, a zealous dedication to reason, and a hunger or eagerness for reliable information’ (Facione, 2020, p. 10) with seven scales of dispositions (1) Truth-seeking, (2) open-mindedness, (3) inquisitiveness, (4) analyticity, (5) systematicity, (6) confidence in reasoning, and (7) maturity of judgement. Cited by Facione’s team, ‘The ideal critical thinker is habitually inquisitive, well-informed, trustful of reason, open-minded, flexible, fair-minded in evaluation, honest in facing personal biases, prudent in making judgments, willing to reconsider, clear about issues’ (Facione, 1998, p. 14). And as explained in the famous Delphi report, a person who cultivated these dispositions is ‘more likely’ to apply CT skills such as analysis or evaluation in daily life than those who only have these skills but do not have a willing to use them (Facione, 1990, p. 13).

Subject-specific critical thinking in writing (CTW)

According to Facione et al., the consensus CT concept stands for non-disciplinary ability without a subject-specific context. Besides general skills and dispositions, there is also an additional dimension when CT is applied to a specific subject, knowledge, or context (Facione, 1990; Facione et al., 1995; Wisdom & Leavitt, 2015).

In 2005, Paul and Elder assembled a team that encompassed diverse elements of thought (Table 2). They made it possible to regard thought through various elements such as ‘goal, topic at issue, information, interpretation and inference, concepts, assumptions, implications, and consequences, and ultimately point of view’ (Paul & Elder, 2005, p. 47). Paul and Elder closely resemble the fundamental elements of the thinking process, particularly in contexts such as text-based activities or those involving reading and writing. We contemplate Facione’s CT framework as comprising vital critical thinking proficiencies that individuals from various disciplines ought to possess. Paul and Elder introduced an adjunct dimension of critical thinking, encompassing intellectual standards (Table 2), which may be employed as a benchmark to assess the caliber of cognitive procedures in written compositions (Paul & Elder, 2005).

Table 2. Paul & Elder’s CT framework.

Elements of thought	Intellectual standards
Purpose Question at issue Information Interpretation and inference Concepts Assumptions Implications and consequences Point of view	Clarity Accuracy Precision Relevance Depth Breadth Logic Significance Fairness

Technology-enhanced social learning (TSL)

Drawing on Vygotsky’s (1978) concept of the Zone of Proximal Development (ZPD), critical thinking can be understood as a socially constructed phenomenon (Wass et al., 2011). It follows that individuals’ cognition is shaped by social interactions, reinforcing the notion of ‘active knowledge construction’ (Cole & Wertsch, 1996, p. 251) as a pivotal element of this process. Conducting oneself in a manner that is appropriate and consistent with the standards and expectations of one’s professional role. By blending personal experiences with community involvement, individuals can improve their cognitive abilities. In 1978, Vygotsky’s social constructivism theory posited that the cultivation of higher-order thinking is largely shaped by social surroundings and the mode of social interaction (Wan Husssin et al., 2019). Recent research has incorporated social interactions into CT-focused composition classrooms to explore their impacts on cognitive processes and perceptions. The cultivation of advanced cognitive abilities is heavily influenced by the social climate and associated interactions within it, as evidenced by prior studies conducted by Chason et al. (2017), Ismail et al. (2017), Miri and Azizi (2018) and Sharadgah et al. (2019).

Further, with the development of technology, recent trends have shifted from traditional classroom arrangements to digital platforms, primarily with the aid of technological tools, according to research by Suwastini et al. (2021). A limited number of contemporary investigations have endeavored to foster critical thinking in the realm of EFL writing, utilizing scaffolding tools that are enhanced with technology and implemented within a social learning milieu. Among the few studies (Arend, 2009; Hajhosseini et al., 2016; Rathakrishnan et al., 2017; Wan Husssin et al., 2019), most focus only on CT skills without exploring affective dispositions in the social learning environment. The main reason is the complexity of the CT concepts, which leads to an unclear understanding of CT constructs. According to a recent review paper related to CT application in EFL education, 43% of authors did not show clear and explicit CT concept descriptions in the recent ten-year study (Yin et al., 2023). Various researchers do not have a clear understanding of CT, so they only focus on a single skill dimension while ignoring the disposition angle. Therefore, there is a gap in exploring technology-enhanced social learning interventions for Chinese EFL learners’ dispositions of CT and subject-specific CT in writing.

This research fills this gap by creating a TSL model that aims to enhance EFL learners’ affective CT dispositions and subject-specific thinking in writing. Figure 1 illustrates the conceptual framework of the TSL model, which includes both face-to-face and online interactions under constructivism. In the TSL environment, students’ subject-specific CT can be cultivated during infused writing practices, where they can trigger CT habits and traits during the class, and dispositions of CT can also be further cultivated through further asynchronous online discussions (AODs). Through the discussion, outcomes, including learners’ sociability, sense of engagement, collaboration, and mutual influence, are generated through the interaction (Kent et al., 2016) where learners present, share, retrieve, and store information or knowledge (Azer, 2009; Din et al., 2015).

Figure 1. Conceptual framework of the TSL model.

Previous evidence showed the role of AODs in cultivating CT abilities. For instance, Wan Husssin et al. (2019, p. 6) concluded the interaction between learners were as:

1. Sharing/comparing of information

2. Discovery and exploration of dissonance or inconsistency among ideas

3. Negotiation of meaning/co-construction of knowledge

4. Testing and modification of proposed synthesis or co-construction

5. Agreement statement (s)/applications of newly constructed meaning

Abrami et al. (2014) also pointed out the necessity of combining authentic discussed situations and mentorship with CT cultivation with its convenience in triggering learners’ thoughts and interactions, which is a typical advantage to enhance CTD. Since the cultivation of CTD is more related to a habit or willingness to use these skills in an invisible way (Dong, 2017a, 2017c). And. compared with synchronous discussion, asynchronous online interactive discussion provides opportunities to be involved in learning better ‘at flexible time and place’ (Wan Husssin et al., 2019, p. 5). Therefore, students are more likely to trigger higher- order thinking via online social interaction mediums (Raspopovic et al., 2017) without the limits of time and space.

Methodology

Research design

An experimental study utilizing a pre-post-test design was conducted at a public university located in the province of Hebei, China. The experiment involved the participation of 60 first-year English major students who belonged to two distinct natural groups. The sample population consisted of 11 male and 49 female participants, with an average age of 18.4. One of the classes served as the treatment group (TG) utilizing the TSL intervention, while the other class served as the comparison group (CG) in a conventional four-step process writing approach. There is a comparable level of proficiency in the English language between the two classes, falling within the CEFR B1 to B2 intermediate range as stipulated by the Common European Framework of Reference for Languages and confirmed by Cambridge. Within the scope of the experiment, it was imperative for two distinct participant groups to participate in the California Critical Thinking Disposition Inventory (CCTDI) survey and produce 300-word IELTS argumentative writing samples both before and after the culmination of the semester.

Within the context of a traditional classroom environment, two distinct cohorts were instructed to engage in argumentative writing exercises facilitated by a single instructor. The process involved several phases, including brainstorming, drafting, peer review, and revision.

Following the session, the intervention cohort will be granted access to an array of online discourse prompts disseminated via ‘Xue Xi Tong’ or ‘XXT’, a highly sophisticated mobile learning platform designed for smartphones, tablet computers, and other portable electronic devices. XXT serves as a formalized academic application of the university, which supports educators and students alike, especially during the COVID-19 pandemic, when remote learning was mandated by the Chinese government. Consequently, with the university’s approval, the students were able to complete their registration using their assigned student identification numbers. evaluated in the system. The instructor has the option to form a group or grant students access to a particular Course Name in a designated class. Figure 2 shows a screenshot of when students post their ideas or thoughts through XXT platform which can help readers better understand this application.

Figure 2. Screenshot of a discussion page in Xue Xi Tong platform.

During the 14 weeks, Both the two groups had two-hour in-class teaching for each week. Except for the instructional segment within the classroom, the members of the TG are tasked with completing AOD activities. The educator will offer pertinent inquiries or material on the online platform and communicate to the students that they are encouraged to actively submit their reflections or suggestions regarding the designated queries. To better understand the research design, Figure 3 shows the brief experimental procedure for the TG and CG.

Figure 3. A brief experimental procedure.

Sample discussion questions

For instance, if the topic of this week’s classroom teaching section is ‘precision’, a situation-based standard of CT, the instructor will ask the class a question in XXT for them to discuss.

Can you think of a recent situation in which you needed more details to figure out something and, because you didn’t have the details, you experienced some negative consequences?

For example, have you ever been given directions to someone’s house, directions that seemed precise enough at the time, but when you tried to find the person’s house, you got lost because of lack of details in the directions?

First, identify a situation in which the details and specifics were important (for example, in buying a computer, or a mobile phone). Then identify the negative consequences that resulted because you didn’t get the details you needed to think well in the situation. Write out your answers.

(Paul & Elder, 2021, p. 95)

The underlying principle is that Paul and Elder’s critical thinking constructs include nine essential standards, with ‘precision’ being one of these standards, as depicted in Figure 4 of Paul and Elder (2005) (Figure 4).

Figure 4. Paul & Elder’s Nine CT standards.

Before the writing exercises, the instructor stressed the need for ‘precision’. The ‘precision’ concept is intended to give students a fundamental understanding of what ‘precision’ entails. The sample questions were as follows:

Consider whether the questions are precise:

• Is there any milk in the refrigerator?

• What should I do about my situation?

On the online platform, there will be a scaffolded discussion about how ‘precision’ can be further understood and how real-world experiences can be reflected.

Participants must actively post their responses in the discussion group, comment on the information shared by others, add new posts if they are motivated by others, and correct their posts if they have a new idea that differs from what was previously shared, according to the requirements that the instructor will specify.

Consequently, during the AOD process, learner-learner interactions in a constructivist social learning environment are generated. Learners gain a profound understanding of the concept of ‘precision’ by responding to the questions posed by the instructor. They apply the pedagogical definition to reality after giving the question some serious thought, which enables them to use the CT construct in everyday situations. Through learner-learner interaction, the learner will assimilate or construct new information through others’ sharing, experience, evaluation of others, or posts containing their content. In this way, in such social-communicative, collaborative, and constructivist learning, with the support of digital platforms, learners could interact and further benefit from their knowledge construction and in-depth learning, and critical thinking will be improved as well.

Data collection tools

Questionnaire: CCTDI

The CCTDI is a well-known instrument with 75 questions in a simple questionnaire format that assesses seven dispositional aspects of critical thinking using seven scales. Usually, it took 20 minutes to complete the questionnaire.

CCTDI has gained significant recognition as a widely used tool that involves 75 questions presented in an uncomplicated questionnaire format. The key objective of this instrument is to evaluate seven dispositional facets of critical thinking using seven distinct scales. Based on our prior experience, the estimated time required to complete the questionnaire was approximately 20 minutes.

The utilization of the questionnaire as the sole instrument can facilitate the evaluation of students’ perception of their critical thinking capacity. According to Nelson Laird (2005), it is viable to employ this tool as both a pre-post-evaluation mechanism to analyze the impact of curricular and extracurricular initiatives on students’ CT attitudes and conduct. It can be administered multiple times for assessment purposes of assessment (Facione, 1991).

To evaluate the dependability of the CCTDI, Facione et al. (1994) conducted a study that calculated preliminary reliability coefficients by applying Cronbach’s alpha. The outcomes of this analysis produced an aggregate score of .90 and personal scores of .80 for each of the seven scales. The tool showed considerable uniformity within a sample size of more than a thousand students in the higher education domain. The comprehensive assessment instrument, comprising 75 questions, demonstrated Cronbach’s alpha coefficients of .60 and .78 for its entire battery and each of the seven distinct scales, respectively. Hence, it can be employed with utmost dependability and authenticity to evaluate the impact of interventions on modifications in the dispositional performance of the study participants.

IELTS argumentative writing test evaluated with Dong’s rubric

Regarding the tool used for written assignments, considering that CT is a primary objective of instructional intervention and thinking in writing is the designated variable to be assessed through the writing test, it is imperative that the test be fashioned in a manner that permits students to exercise multifaceted thinking and leverage their cognitive faculties to substantiate claims, justify opinions, and solve problems. The IELTS (academic) test’s writing task 2 was selected as the primary source of writing tasks because its testing objectives of assessing candidates’ abilities to present problem-solving solutions, justify opinions, analyze evidence and opinions in a comparative and contrasting manner, and evaluate and critique ideas, evidence, or arguments (Esol, 2021) aligns with the aforementioned requisites.

Moreover, Dong’s criteria rubric (Dong, 2017b) applied to evaluate participants’ CTW was based on Paul and Elder’s ‘Intellectual Standards, which is a holistic criterion that can evaluate students’ CT in EFL writing products and test the effectiveness of CT in teaching writing. This rubric experienced revision and empirical verification in Dong (2017a) and helped to maintain high reliability and validity in collecting participants’ CT reflected in writing as well as further data analysis.

Data analysis

One-way ANOVA was used to confirm there was no significant difference between their pre-test scores which helped to control students in the two groups who were at the same level before the experiment. It also helped to test whether the two groups got changes through one semester. Treatment. To control the effect of confounding variables (Lu & Xie, 2019; Yang & Gamble, 2013), ANCOVA was used to compare two groups’ post-tests with pre-tests as covariance. After receiving the pre-post-critical thinking disposition and writing scores of the TG and CG, the initial phase involved conducting a normality test on the sample. To ascertain the normality of the data sample distribution, a normality test was performed using SPSS software.

Results

Scores of CTD

Descriptive statistics represent the mean in pre-CTD scores of the two groups. Their mean scores are similar to those for CG (285.97, and 283.77, respectively). There was no statistically significant difference, where F (df = 1, 58) =.094, p$>.$05, was not significant. Thus, there was no significant difference in the CTD between the two groups’ students before the study.

The presented descriptive statistics illustrate the average pre-CTD scores for each group. After conducting the analysis, it was noted that the average scores for both the CG and TG were comparable. Specifically, the mean score for the CG was 285.97, whereas that for the TG was 283.77. Statistical analysis revealed a lack of significant difference, as evidenced by F (df = 1, 58) =.094, p$>.$05. Consequently, no noteworthy variance in CTD was observed between the two groups of students before the commencement of the study (Figure 5).

Figure 5. Mean scores of TG and CG’s pre-post-CTD results.

The analysis of the CG's pre-post-CTD scores through descriptive statistics showed that the average value of pre-scores was 285.97, whereas the average value of post-scores was 286.43. No significant progress was observed in the CTD scores of the CG group. Based on the ANOVA analysis outcomes, as presented in Figure 6, the CG scores did not exhibit any statistically significant variance (p = .95) between the pre-post-experiment stages. Based on this evaluation, it appears that the CG did not achieve substantial advancements with regard to CTD.

Figure 6. ANOVA results through comparing pre-post-CTD scores of CG.

By comparing the TG’s pre-post CTD results, it is clear to see the mean of the two tests of TG in Figure 6. The mean pre score was 283.77, and the mean of post score was 302.60. There was an obvious 20-point improvement for this group under the TSL approach. The inferential analysis results (Figure 7) using ANOVA showed that this improvement was statistically significant with (p $=.$016, ${\eta }^2=.$096), where ${\eta }^2\gg .$06 represents a moderate effect size (Cohen, 1988; Hu & Dai, 2017). TG’s positive improvements to CTD level were noted, reflecting that the TSL treatment was successful in cultivating CTD by one-semester teaching.

Figure 7. ANOVA results through comparing pre-post-CTD scores of TG.

In addition, to investigate whether TG performed better than CG in terms of CTD, the researcher used SPSS ANCOVA to compare the two groups’ post-scores in CTD with the pre-test scores as covariates. Although the previous analysis showed that there is no significant difference between the two groups’ pre-CTD scores, ANCOVA analysis was conducted to prevent the effect of pre-test scores on post-tests. In this way, the results of the analysis will be more scientific and realistic. Through the analysis, the pre-scores of CTD did not significantly affect the post-scores (p > .05), which can be seen in Figure 8. According to the results, it showed a significant difference between the two groups (p$=.$037, ${\eta }^2=.$074), reflecting a moderate effect size. And since the mean of CG was higher than TG (${\mathit{\text{Mean}}}_{\mathit{\text{CG}}}\mathrm{ }$= 286.43; ${\mathit{\text{Mean}}}_{\mathit{\text{TG}}}$= 302.60) in the post-CTD test this demonstrated that TG performed much better than the CG in the post-test in CTD. Thus, the TG achieved positive progress than the CG in the CTD, demonstrating TSL approach was successful in improving CTD compared with the traditional process writing approach.

Figure 8. ANCOVA results through comparing post-CTD scores of CG and TG.

Besides, except for finding the difference in the scores of overall CTD, the difference in each aspect of the seven dispositions was also indicated in this research. The results in Table 3 shows the TSL approach successfully enabled participants to better perform dispositions of CT, especially in open-mindedness (p$=.$003, ${\eta }^2=.$146), inquisitiveness (p$=.$009, ${\eta }^2=.$112), analyticity (p$=.$018, ${\eta }^2=.$095), systematicity (p$=.$005, ${\eta }^2=.$129) and confidence in reasoning (p$=.$001, ${\eta }^2=.$188). Among them, the degree of TSL approach’s positive impact on students’ open-mindedness and confidence in reasoning was large (${\eta }^2>.14$) with a large effect size. And the results indicated a moderate effect on inquisitiveness (p$=.$009, ${\eta }^2=.$112), analyticity (p$=.$018, ${\eta }^2=.$095) and systematicity (p$=.$005, ${\eta }^2=.$129). Significant p values were bolded in the table. However, as for the truth-seeking and maturity of judgement, learners did not indicate significant progress after the intervention (p > .05).

Table 3. Differences in the scores of CTD of CG and TG after the intervention.

Areas	Group	Mean	SD	p
CTD Overall	CG	286.43	29.19	.037
	TG	302.60	31.06
Truth-seeking (TS)	CG	31.10	5.87	.406
	TG	32.40	6.53
Open-mindedness (OM)	CG	40.03	4.33	.003
	TG	44.03	5.24
Inquisitiveness (I)	CG	48.60	5.26	.009
	TG	52.13	4.74
Analyticity (A)	CG	40.53	4.58	.018
	TG	44.43	6.75
Systematicity (S)	CG	36.83	5.04	.005
	TG	41.33	6.83
Confidence in Reasoning (CR)	CG	43.03	5.66	.001
	TG	48.93	7.11
Maturity of judgement (MJ)	CG	43.77	6.95	.485
	TG	44.93	6.91

Scores of CTW

Their mean scores in CTW at the beginning of the semester were almost at the same level as that for CG is 22.03 and for TG is 22.13 (Figure 9). There was no statistically significant difference between the two groups (F (df = 1, 58) =.015, p$>.$05). Thus, there was no significant difference in the CTW between the CG and TG before the experiment.

Figure 9. Mean scores of TG and CG’s pre-post-writing test results.

As Figure 9 shows, the mean of the CG’s pre-scores was 22.03 and the mean of post-scores was 26.37, which had a four-point increase through the one-semester traditional writing approach. The inferential statistics through ANOVA (Figure 10) showed that there was a significant statistical difference (p$<.$05) with ${\eta }^2=.$311 between the pre-post-test scores of CG in CTW, where ${\eta }^2\gg .$14 represents a big effect size. Positive improvement of CG in terms of CTW was noted before and after the semester. This means the one-semester traditional writing instruction process was successful in enhancing the CTW level.

Figure 10. ANOVA results through comparing pre-post-writing scores of CG.

By comparing the TG’s pre-post-CTW scores, the descriptive statistics illustrated that the mean of pre-scores was 22.13, and the mean of post scores was 28.87, which had an obvious six-point progress in subject-specific thinking in writing. The inferential analysis results (Figure 11) using one-way ANOVA showed that this improvement was statistically significant (p$<.$05) with ${\eta }^2=.$562 which indicates a large effect size. TSL approach is also useful in improving the CTW level through one-semester treatment.

Figure 11. ANOVA results through comparing pre-post-writing scores of TG.

To see whether TG performed better than CG in subject-specific thinking in writing, the post scores of the two groups were analyzed using SPSS ANCOVA. Through the results shown in Figure 12, the significant difference between the two groups’ post-test was not significantly influenced by their pre-test performance (p > .05). After excluding the influence of two groups pre-tests level, the results showed that TG performed slightly better than CG in the post-test on subject-specific thinking in writing (${\mathit{\text{Mean}}}_{\mathit{\text{CG}}}\mathrm{ }$= 26.37; ${\mathit{\text{Mean}}}_{\mathit{\text{TG}}}$= 28.87). Although only two points were higher in the TG than in the CG, the inferential statistics showed that the difference was significant (p$<.$05, ${\eta }^2=.$143), representing a large effect size, as shown in Figure 12. This means that TG’s CTW scores were significantly higher than CG’s. Thus, the null hypothesis is rejected. The improvement in subject-specific thinking in writing in the TG was significantly higher than that in CG. Therefore, the TG made more progress than the CG in this aspect after the experiment.

Figure 12. ANCOVA results through comparing post-CTW scores of CG and TG.

In addition, difference in each sub-CTW is shown in Table 4. The results demonstrated TSL successfully took a positive impact on learners’ CTW especially in terms of clarity (p$=.$042, ${\eta }^2=.$070), relevance (p $=.$000*, ${\eta }^2=.$274), breadth (p$=.$004, ${\eta }^2=.$137), logic (p$=.$036, ${\eta }^2=.$037) and fairness (p$=.$024, ${\eta }^2=.$086). Among them, students’ relevance got the biggest improvement with a large effect size (${\eta }^2>.14$). Students also got some improvement in terms of clarity, breadth, and fairness, which indicates by the results with a moderate effect size (${\eta }^2>.06$). And the level of logic improved slightly with a small effect size (${\eta }^2>.01$). However, students’ accuracy, precision, depth, and significance did not achieve significance progress after the intervention (p > .05).

Table 4. Differences in the scores of CTW of CG and TG after the intervention.

Areas	Group	Mean	SD	p	${\eta }^2$
CTW Overall	CG	26.37	3.06	.003	.143
	TG	28.87	3.18
Clarity	CG	3.07	.58	.042	.070
	TG	3.37	.61
Accuracy	CG	3.17	.38	.260	.022
	TG	3.30	.47
Precision	CG	2.50	.57	.079	.053
	TG	2.80	.61
Relevance	CG	3.10	.55	.000*	.274
	TG	3.70	.47
Depth	CG	2.80	.55	.619	.004
	TG	2.90	.61
Breadth	CG	3.03	.49	.004	.137
	TG	3.43	.57
Logic	CG	2.83	.59	.036	.037
	TG	3.03	.49
Significance	CG	2.93	.58	.136	.039
	TG	3.13	.43
Fairness	CG	2.93	.64	.024	.086
	TG	3.20	.41

Discussions and recommendations

First, the TG taught by the TSL approach achieved CTD improvement through a one-semester intervention, while the CG taught by the traditional approach did not show any significant improvement in terms of CTD.

Certain dispositions were taught before the writing practice, which refers to the introduction of a person’s values, beliefs, and intentions toward CT. For example, there are seven items in Facione’s CT disposition model, which is regarded as a universal, widely shared framework: inquisitiveness, systematicity, analyticity, truth-seeking, open-mindedness, self-confidence, and maturity. Through CTD input teaching, the teacher taught these seven generic CTD constituents and allowed students to understand each disposition that a critical thinker should own before writing practice. Students were conscious of what kind of dispositions they needed when trying to use skills in whatever they did in daily life. To be more specific, for instance, when taught one of the dispositions ‘open-mindedness’, students in TG had such input that a person needs to have such willingness to be open-minded when they analyzed or evaluated things. When they were infused in online discussion, the precious ‘open-mindedness’ disposition input may be priming and remind the student to be open-minded when during specific writing process. In this way, such practices using the TSL helped students in TG increase their awareness of CTD and reminded them to have a relative willingness to be critical in a fair-minded way. The findings of CTD improvement in the TG proved the positive power of the TSL model in enhancing the dispositional aspects of CT.

Further, it highlighted the value of interactions between teachers and students as well as peer learners under TSL. Constructivism theory triggers the rethinking of the role of teachers and students in a CT-oriented social constructivist learning environment, which is favorable for the cultivation of CTD (Dong, 2017a). In the present study, interactions were observed in both groups. Although CG in the traditional process writing approach did not receive online scaffolded CT-oriented discussions, students in the CG still had opportunities to interact with each other. We cannot say that there were no interactions in the traditional process writing class for CG. For example, during non-CT-oriented process writing practices, students in CG also interacted in talking about ideas of the writing topics, and they evaluated each other about the advantages and shortcomings of peers’ compositions. During this process, students in CG interacted with each other and constructed their writing content in a social constructivist manner by receiving ideas and opinions from their peers.

However, no further online CT-oriented discussion for CG, where the interactions between the teacher and students, as well as between peers, were content-focused instead of thinking-focused. They received no CT-oriented guidance during the interactions, and this is one of the main reasons that the CG did not perform better after the study in terms of CTD. Their disposition level of CT was not significantly improved through the one-semester teaching in the traditional process-writing approach. Therefore, compared with the unchanged CTD level of CG, TG’s positive changes of CTD proved the role of AODs in promoting disposition of CT.

To be more specific, this result is consistent with previous studies such as Rathakrishnan et al. (2017) and Wan Husssin et al. (2019) which advocated the positive effects of online technology in scaffolding CT abilities. But it is worth noting that this finding result in this research further proves the scaffolding role of online technology in cultivating dispositions of CT. As the previous theoretical foundation showed that a social interactive environment has an advantage in terms of involving students in higher-order thinking processes especially the cultivation of CT disposition (Dong, 2017a). While different from Dong’s study, in this research the common social interactive writing practices under CG did not significantly improve CTD level. On the contrary, the TG under the technology-enhanced approach scaffolding with AODs achieved significant CTD improvement, which demonstrates the positive role of online technology in enhancing CTD with more flexible time and location (Wan Husssin et al., 2019). This provides empirical evidence that online technology especially the AODs on effectively in cultivating habits or willingness to use CT skills. During this process, online techniques encouraged learners to interact with each other, which promoted the process and quality of social learning. It plays more like a facilitator of social learning with the help of technical benefits in interacting more conveniently (Al-Rahmi & Zeki, 2017; Dwyer et al., 2014; Lewis, 2016; Wan Husssin et al., 2019). Future studies could make full use of the power of online technology integrated with common social learning activities in EFL writing classes.

Second, in terms of subject-specific thinking in writing, both CG and TG had improved CTW through one-semester teaching using the traditional process writing approach and the TSL approach, respectively, but TG under TSL had more improvement in CTW than CG under the traditional approach.

Among the previous results, the traditional approach did not affect the disposition of CT, but it could improve students’ CTW. This demonstrates that writing is a medium or tool for enhancing CT during the writing process. Previous researchers concluded that CT and writing share the common ‘procedural’ characteristic of the two systems (Mulnix, 2013) which requires a conscious focus on the process (Akbari et al., 2018; Crews-Anderson, 2007). Writing is also a process of thinking because of the close link between thinking and writing. Similar to Gnadimathi and Zarei (2018), students can exercise and shape independent patterns of thinking during ‘ongoing CT’ (Kaviani & Heidar, 2020, p. 310). Based on these, researchers recommended to trigger sustained CT during the process and highlighted the importance of considering ‘each stage’ involving the implementation of higher-order thinking when teaching writing (Singh et al., 2020, p. 196).

In this research, although there were no technology-enhanced constructivist elements for CG, the traditional process-writing approach did improve students’ specific CT in EFL writing shown by the finding results. The traditional process writing practices provided opportunities for students to trigger and practice thinking during each stage of writing, and the pattern of individual thinking for each participant in the group was shaped during the ‘ongoing CT’ described by Kaviani and Heidar (2020, p. 310) in the process of writing. Therefore, the improvement of CTW under the traditional approach for CG confirmed the writing’s role in presenting CT as a platform.

At the very least, the research findings proved the role of process writing with brainstorming, drafting, peer review, and revision stages in enhancing specific CT in EFL writing (Dong, 2017a). Participants under such traditional writing practices throughout the semester could perform better after the study. This proved the role of process writing in enhancing CTW with non-CT-oriented instructions, as the carefully designed process of writing could be beneficial for shaping and practicing sustained CT during the process, which is similar to previous studies such as Mall-Amiri and Sheikhy (2014), Huang (2016), Chason et al. (2017) and Sharadgah et al. (2019) in terms of cultivating CT through process writing. Although these studies have already advocated the positive role of writing in cultivating specific CT in writing, however, they did not show empirical evidence of technical scaffolding’s role in promoting higher-order thinking in writing class.

Therefore, writing, as a medium to practice CT, needs to combine the positive power of social learning and the technology scaffolding role. Without CT-oriented online discussion scaffolding, the benefit of improving CT is limited. According to the results, TG in the TSL framework with additional online CT-oriented discussions performed better than CG in subject-specific thinking in writing. This highlights the benefits of a technology-enhanced social constructivist learning environment in improving thinking in writing. In contrast to the non-TSL interactions that occurred in CG, TG built an online CT-oriented social constructivist learning environment in which students in TG could interact with each other in a thinking-centered rather than content-centered way, which provided opportunities for the learner to have a chance to be involved in high-quality social interactions without limits of time and space (Carr, 2012; Din et al., 2015; Kent et al., 2016; Wan Husssin et al., 2019). They interacted with each other based on a CT-related context and constructed and upgraded their thinking through CT-oriented interactions. During this process, their willingness to use CT skills was triggered and cultivated (Dong, 2017c), which had positive effects on them to perform better during the writing process.

Implications and future directions

As evidenced by the study’s positive findings on the efficacy of the TSL framework in the development of students’ EFL writing and CT competencies, educational administrators and teaching faculty within EFL departments should place greater emphasis on the critical thinking component of EFL writing instruction and evaluation criteria in domestic assessment modalities. Educators must have a comprehensive approach to EFL writing that encompasses not only language forms but also nurtures and cultivates students’ critical thinking and analytical skills.

Writing is an effective tool for CT practice. Teachers need to consider ‘each stage’ that involves the implementation of HODs for teaching writing since CT requires a conscious focus on the process (Akbari et al., 2018; Crews-Anderson, 2007).

‘Interaction’ plays a continuous and key role in cultivating the habits of CTD. Technology-scaffolded social learning techniques are effective in improving CT disposition and subject-specific thinking in writing. Building an interactive environment during the thinking and writing processes is necessary. Interaction is useful for the enhancement of higher-order thinking such as CT.

This study highlights the crucial role of skills and dispositions in the development of general critical thinking abilities. However, it is worth noting that the cultivation of CT dispositions may require an extended period. Subsequent research endeavors may entail a more protracted timeframe spanning at least one year, with the aim of delving into the progression of students’ critical thinking dispositions beyond the duration of the current investigation.

Finally, given the growing dependence on virtual learning platforms in the realm of EFL education, this research only adopted online discussions as a scaffolding tool for developing students’ in-depth thinking after face-to-face teaching practice. Further studies could be designed to modify the proposed approach to an online interactive environment instead of a blended approach to enhance its suitability in a more dynamic educational medium. Various research methodologies can be employed to delve deeper into students’ overall CT learning experiences, specifically in EFL writing, within a virtual learning environment.

Conclusion

In conclusion, the findings indicate that the TSL model has positive effects on Chinese EFL learners’ disposition toward CT and subject-specific CT in writing. Although writing is an effective medium to practice and improve CT in writing, the improvement of CT dispositions needs the participation of technological applications for its efficiency and convenience. It will provide more benefits if scaffolded with online interactive social learning techniques to cultivate the habits of CT and specific thinking in writing.

This study makes three key contributions, encompassing methodological, theoretical, and empirical aspects. The research shows the positive influence of a technology-based social learning intervention conducted over a 14-week semester in fostering students’ critical thinking disposition and writing skills. Theoretically, this study fills the knowledge gap that effective TSL models can successfully increase Chinese EFL learners’ dispositions of CT and emphasizes the importance of an online discussion-scaffolded social learning interactive environment for CT cultivation in EFL writing teaching. of both capabilities. Empirically, this study provides practical experience and recommendations for educators and teachers to improve students’ development of critical thinking dispositions and writing qualities.

One of the limitations is that the sample scope was limited because the participants were only from one public university in Hebei Province which is in the Northern area of China. Further studies can expand the scope of the research sample to change the university background and select EFL learners from other universities. Another limitation was the educational level of the participants. In this study, participants were only CEFR B1 or B2 intermediate-level students. Findings may differ among other language-level students, such as at a low or high level. Researchers in this area can also duplicate similar studies in other countries since this research only has implications for Chinese EFL learners. Finally, though the TSL instruction has already proved to be effective in enhancing Chinese EFL learners’ disposition of CT, it is necessary to deeply explore the connection between technology and social learning and critical thinking. The technical benefits of traditional methods in increasing social learning were not explicitly investigated with evidence. Therefore, future studies could be more concerned on this point.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Xue Yin

Xue Yin received the master’s degree from University of Liverpool in in TESOL. Now, she is a PhD candidate in Department of Language and Literacy, Faculty of Education, University of Malaya. Her main research area is second language acquisition, EFL writing, critical thinking, critical thinking pedagogy, CT-oriented instruction etc. E-mail: [email protected]

Mohd Rashid Bin Mohd Saad

Mohd Rashid Bin Mohd Saad (Senior Lecturer), PhD (2006–2008) Sociology in Math and Science, University of Malaya, Malaysia. MEdu (1997–1998) TESOL, University of Leeds, UK. Award and Stewardship:

• • Best TESL Lecturer for MEd Program 2020, Faculty of Education, 2020 (Ptj)

• • Overall Best Presenter, IPCORE USM-UPI, 2018 (International)

• • Best Paper Award, IPCORE USM-UPI, 2018 (International)

• • Most Innovative Research Award, IPCORE USM-UPI, 2018 (International)

E-mail: [email protected]

Huzaina Binti Abdul Halim

Huzaina Binti Abdul Halim (Senior Lecturer), PhD (Education), Imperial College London, UK. M.Sc (TESL), Universiti Putra Malaysia, Malaysia. Areas of expertise: teacher education (ESL curriculum), literature education, creative and critical thinking. E-mail: [email protected]