Applying the UTAUT Model to Understand Factors Affecting Micro-Lecture Usage by Mathematics Teachers in China

Abstract

In the last decade, micro-lectures have been widely used to teach mathematics, but only a few studies have examined the factors affecting teachers’ intentions of using micro-lectures. As teachers are key to integrating modern technologies such as micro-lectures into students’ learning processes, knowledge about teachers’ intentions in this regard could be of particular importance. This study aimed to analyze the behavioral intention (BI) of mathematics teachers in using micro-lectures in mathematics in China, and identify the most influential factors involved, for the very first time. The Unified Theory of Acceptance and Use of Technology (UTAUT) model was used as a design model to investigate teachers’ BIs concerning the use of micro-lectures, and we used an online questionnaire to collect quantitative data. The participants in our research were 174 mathematics teachers from China, 166 of whom provided us with evaluable questionnaire data. Furthermore, partial least squares (PLS) regression was used, and hypothesis testing was performed with the Smart-PLS software. From the results, BI was positively affected by Performance Expectancy (PE), Effort Expectancy (EE), and Social Influence (SI). BI and facility conditions also had positive effects on user behavior; in contrast to other studies, SI had the most significant positive effect on BIs in our study. Our findings could provide insights into both future strategies for successful technology integration in mathematics classes and into mathematics teachers’ intentions towards integrating technologies into mathematics teaching.

1. Introduction

Mathematics is an abstract subject that can be difficult to understand for many students [1,2], but technology-based learning media could help teachers to design interesting mathematics lessons and thereby facilitate students in overcoming mathematical obstacles to learning [3,4,5]. This high potential of using technologies in teaching and learning mathematics has been identified by many researchers and practitioners, and today, many technology-based learning media are used by junior high school teachers to teach mathematics [6,7,8]. A micro-lecture (ML) is a video-based teaching and learning activity that is becoming more popular. The micro-lecture was first proposed by David Penrose in 2008 to help reflect and summarize core learning materials and direct students to explore the material more deeply [9]. According to Wang K et al. [10], micro-lectures are more attractive than MOOC and reading material for students. Compared to other computer-assisted mathematics learning programs, it has four beneficial features, namely: (1) shorter video duration so that students do not tire while watching micro-lectures; (2) a smaller theme that focuses on basic mathematical concepts which students must master; (3) design creation and explanation is easier, and hence mathematics teachers can modify micro-lectures according to class needs; and (4) the learning effect is very promising. Furthermore, it can be used for broader concepts, not only in school education, but also in adult education, training, and job training [11,12].

Micro-lectures can cover a wide range of new knowledge in the traditional classroom [11,13], and this is in line with teachers’ needs. They can help teachers focus on basic knowledge about a mathematics topic, and also reflect on their teaching as well as learning activities. This means the content in micro-lectures is more streamlined with a smaller resource capacity.

A micro-lecture is defined as a short 1-min video that focuses on concepts and specific knowledge points [14]. Furthermore, compared to the use of textbooks in traditional classrooms which have a duration of 45 min per lesson, and where students can typically only focus for 15 min, it was more effective [12]. However, according to Wang et al. [10] MLs in China usually last from 5 to 10 min. This extension of MLs could be problematic concerning the effectiveness of MLs, because the value of MLs decreases with increasing length [15]. In order to maintain the effectiveness of MLs, the designers of MLs, i.e., teachers, have a particularly important role to play.

As a variation of learning approaches, the use of micro-lectures can be adopted in online learning or in face-to-face learning in the classroom. Using MLs could provide a new learning experience for students, and at the same time utilizing MLs could improve students’ understanding of mathematical concepts. Micro-lectures are also one of the tools that have changed the traditional teaching of mathematics from teacher-centered to student-centered. Students can use MLs to learn new mathematics knowledge; hence, they become acquainted with the process of learning [16]. Once students begin to use micro-lectures, their mathematical, manipulating, and creative thinking skills improve at the same time. Moreover, they can be used to check whether they have mastered the knowledge points or not. Teachers can also take action for consolidation or improvement, and this increases student participation during micro-lectures and mathematics lessons [9]. This novel approach should enable students to review the content of mathematics lessons anywhere, anytime, and as often as needed, which should contribute to making mathematics learning more flexible, and easier for students to adapt the learning process to their needs [10,17,18]. The increasing availability of personal digital devices in recent years should give a further boost to this approach to learning, thereby helping students to better understand mathematical concepts [10,12,19]. One of several factors that could contribute to the success of using MLs in teaching and learning mathematics concerns teachers who develop MLs and integrate them into their teaching in a pedagogically sound way.

The number of studies focusing on developing micro-lectures for learning mathematics demonstrates the importance of MLs as a tool for teachers in the learning–teaching process [3,12,13,16,20]. Several studies have shown that micro-lectures can improve learning satisfaction [10,11], interest in learning [4,21,22], learning motivation [17,23,24], learning achievement [11,17,25], self-efficacy [26], and can provide new experiences to students [27,28,29]. As a consequence of the COVID-19 outbreak, students were not allowed to go to schools for several months, and the Chinese Ministry of Education invited 60 experts to develop free-access micro-lectures (NCPM micro-classes). These MLs could be used by students to study during the pandemic emergency [30], and could also persist in functioning as a high-quality enrichment of mathematics teaching in a post-COVID-19 era. The Chinese government recognizes the effectiveness of using various types of technology-based learning media, and focuses on developing learning approaches and models to improve high-order thinking skills. However, the efficacy of this technology-based learning media remains anecdotal. With the increasing quantity of technology-based learning media for teaching mathematics, it becomes necessary to examine the users’ attitude towards a new technology-based learning media and identify the determinants of their acceptance and rejection. Therefore, various studies have examined the acceptance and adoption of information and communication technology (ICT) by teachers [31,32]. They considered predictions and analysis for the factors influencing teacher acceptance and the adoption of ICT for teaching. These factors included Performance Expectancy, Effort Expectancy, Social Influence, and Facilitating Conditions. Wang et al. [10] investigated the behavioral intention of students to use micro-lectures. However, no study has focused on the factors that influence teachers to use micro-lectures for teaching, particularly in mathematics in junior high schools. Using the Unified Theory of Acceptance and Use of Technology (UTAUT) model developed by Venkatesh et al. [33], our study aimed to fill this gap by determining factors that influence teachers’ intentions and actual use of micro-lectures in China for teaching junior high school mathematics. Our study also aimed to identify the factors that most influence Junior High School teachers to use micro-lectures to teach mathematics in China. Specifically, our study pursued the following research question:

What factors could influence mathematics teachers’ intentions and actual uses of micro-lectures in China?

To be able to contribute to the research question above, on the one hand, the factors that indicate teachers’ intentions and actual uses of micro-lectures by mathematics teachers in China were identified, and, on the other hand, the extent to which each factor affected the intentions and actual use of micro-lectures was also identified. Through this knowledge concerning teachers’ intentions and teachers’ actual use of micro-lectures, our study provides both scientific knowledge regarding teachers’ behavioral intentions concerning micro-lectures, and practical knowledge regarding the integration of novel teaching media and approaches in teaching mathematics.

2. Literature Review

2.1. Micro-Lectures

The education sector requires more active learning environments in the era of the internet, and micro-lectures offer a learning medium with great potential [34]. They give students ownership of their learning, and provide many opportunities for them to manage their study time [9].

Micro-lectures were first used in formal education several years ago in China. Since then, video clips with a short duration that draw attention and focus on specific points of knowledge have been used on various learning platforms. Micro-lectures can be used in blended and mobile learning formats [35]. The concentration of study in China on micro-lectures has two focuses, including: (1) making and providing the creative design of micro-lectures; and (2) strategies for implementing micro-lectures in learning. Compared to other educational resources, micro-lectures can integrate text, audio–visual factors, interesting animation, and multimodality into classroom teaching. They also break down complex knowledge into fragments that can increase the interest in learning and understanding for students. Due to their small bit capacity, micro-lectures can easily be transferred, downloaded, played back on many variants of gadgets, and combined with an online learning environment. Finally, micro-lectures are one possible step to change learning approaches for the next generation.

Zhang [9] explained that micro-lectures based on their use can be divided into three categories.

• Knowledge point type: this focuses on discussing one important point in a single session. Teachers will use this type of ML to provide concrete examples or problem-solving steps. Moreover, this kind of micro-lecture is usually given by teachers to students before or after class;
• Creating learning context type: this type builds learning environments, gives students a problem, and promotes them to think. Background knowledge is created through micro-lectures, which can provide an active model for role play. Furthermore, micro-lectures necessitate student cooperation, questioning (problem-posing), and problem solving in the video. They aim to help students conduct deep learning and improve their inquiry skills.
• Presentation and the evaluation type: unlike the previous two types made by teachers, this micro-lecture is delivered by students in groups. They are tasked with selecting the topic of their own video, and designing and making micro-lectures according to their own style. Subsequently, micro-lectures that students have made in collaboration with teachers are evaluated.

2.2. Previous Studies about the Use of Micro-Lectures in Teaching and Learning Activities

Many studies that discuss the integration of technology in teaching and learning activities have been published [3,20,29]. However, those that focus on discussing micro-lectures are still limited. Micro-lectures are widely used in blended learning and online courses, and they have become suitable to be applied during the pandemic. They are also used in various fields of science, as well as English [36], psychological courses [10], computing [16], and engineering [11,35].

Using the acceptance model, Wang [10] conducted previous studies on micro-lectures, analyzing the factors that influence university students’ intentions to use micro-lectures through the TAM model. There were 27 questions collected from the psychological course using a 5-point Likert scale. According to the findings, perceived usefulness and user satisfaction were the biggest factors. Students access micro-lecture with ease and were satisfied with using it as a learning medium.

2.3. The UTAUT Model

The term “technology acceptance” originates from the business field which discusses the causal relationship between the use of technology and profit margin [37]. This is related to the willingness and constant use of technology to perform a daily task. In the context of education, technology cannot be effective unless it has users.

To examine a person’s behavior intention and Use Behavior, the Theory of Reasoned Action (TRA), Technology Acceptance Model (TAM), Theory of Planned Behavior (TPB), or Unified Theory of Acceptance and Use of Technology (UTAUT) models could be used [38]. Each model has its advantages and disadvantages: TAM is a tool that is used to explore technology adoption in a variety of contexts [39,40], but no more than 40% of the variance (R² = 40) in the dependent variable can be explained using this model, which leaves room for additional antecedents of acceptance [32]. Moreover, TAM has more limitations and is frequently redefined resulting in theoretical chaos and confusion [41].

It is at this point where the development of UTAUT proposed by Venkatesh, Morris, and Davis started [33]. Using this model, up to 70% of the variance (R² = 70) in Behavior Intention and 40% of the actual use can be explained [42], as compared to other models. UTAUT consists of exogenous variables of Performance Expectancy (PE), Effort Expectancy (EE), Social Influence (SI), and Facilitating Conditions (FC). The endogenous variable, Behavioral Intention (BI), is influenced by the model, which eventually leads to Use Behavior (UB) [43]. The UTAUT paradigm is used in this study because it is a valuable academic lens for addressing issues linked to information and communications technology (ICT) resource attitudes in the field of education [44]. To analyze the effect on the actual use of technology in line with BI, it can be demonstrated that the constructs within UTAUT can be appropriately related to specific contexts. Therefore, using the UTAUT model, this study could predict and explain Chinese mathematics teachers’ intentions and actual uses of micro-lectures in mathematics classrooms. Several mediating factors, namely sex, age, experience, and willingness to use technology-based learning media, have an indirect impact on Behavioral Intentions, together with four main variables [33]. In recent years, the UTAUT model has described user intentions and usage. This model is frequently used to assess the user acceptance of technology-based learning media such as video-based learning [45], blended learning [46], online learning [47], and interactive whiteboard [48]. However, no study on using micro-lectures for mathematics learning, particularly concerning Chinese teachers’ intentions and actual uses of MLs, has been conducted.

2.4. Proposed Study Model and Hypotheses Development

Mathematics teachers’ use of micro-lectures is the first step in integrating new technology into learning processes. In this study, the UTAUT model was used to analyze mathematics teachers’ acceptance of micro-lectures in Guangxi Province, China. The factors explored as determinants of the acceptance behavior included PE, SI, EE, and FC.

2.5. Performance Expectancy (PE)

The degree to which users believe that using technology can improve their job performance is referred to as Performance Expectancy (PE) [49]. PE has been shown in many studies to influence BIs to use technology-based learning media because of the benefits obtained. Furthermore, it was found to be a significant factor influencing teachers in Africa to use ICT in their classrooms [50]. Based on this background, our proposed hypothesis was:

Hypothesis 1.

PE influences BIs to use micro-lectures in mathematics.

2.6. Effort Expectancy (EE)

The level of ease of use for a technology medium is defined as Effort Expectancy (EE) [33]. According to previous studies, EE has a significant effect on the use of technology to teach [51,52,53]. Based on this background, the second proposed hypothesis was:

Hypothesis 2.

EE influences BIs to use micro-lectures in mathematics lessons.

2.7. Social Influence (SI)

The degree to which a person feels that important people around him/her believe that someone should use technologies or a technological system is referred to as Social Influence (SI) [33]. A previous study showed that SI is categorized by friends, family, coworkers, and students and found that SI affects BIs [32,50,54]. In addition, teachers are more likely to have a strong BI for using ICT [55]. Based on this background, our third proposed hypothesis was:

Hypothesis 3.

SI influences BIs to use micro-lectures in mathematics lessons.

2.8. Facilitating Conditions (FC)

The degree to which a person believes that an environment supports the use of technology is defined as Facilitating Conditions (FC) [33]. In this study, it is the accessibility of an appropriate learning environment and infrastructure within a school that can promote the use of micro-lectures by secondary school teachers in teaching mathematics. According to previous studies, FC can have a direct effect on UB without going through BI [55,56,57]. Due to its inconsistent influence on BI, FC is still being debated. Several studies have demonstrated that FC does not affect BI and UB [43,48,58]. However, others found a positive effect of FC on BI [31,59,60] and a significant effect of FC on UB [31,61,62]. Due to the background that, in an educational context, there is a direct influence of FC on UB, our fourth proposed hypothesis was:

Hypothesis 4.

FC influences UB to use micro-lectures in mathematics lessons.

2.9. Behavior Intention (BI) to Technology Use

Behavioral Intention (BI) to technology use is defined as a person’s intention to adopt and use technology in the future [33]. This is the most significant factor influencing the UB of technology. In this study, the objective model adopts the original UTAUT model, with all PE, EE, and SI factors affecting BI and only FC affecting UB. Therefore, this study can examine the variables that intervene in BI to positively influence UB. According to Venkatesh [33], there are fewer or no studies that show a relationship between user acceptance and individual usage outcomes. Based on this background, our fifth proposed hypothesis was:

Hypothesis 5.

BI influences UB to use micro-lecture in math lessons.

In line with previous studies [56,63], in our research, only the UTAUT model was used and moderators were frequently dropped. A moderator was not used because there was no difference in the moderator for the adoption and usage context. The following research model was authenticated from the above hypotheses, and was developed in line with the UTAUT model for micro-lecture acceptance among mathematics teachers (see Figure 1).

3. Research Methodology

To test the initial hypotheses listed above and provide suggestions based on findings about the factors that could influence mathematics teachers using micro-lectures, this study used a quantitative approach. The UTAUT model was used to determine junior high school (students from 12 to 15 years old) mathematics teachers’ intentions, and the actual use of micro-lectures in mathematics in Guangxi, China. Guangxi is a Chinese autonomous region located on the country’s south-eastern border. In terms of economy, technology, and education, this is also one of the regions that lag behind other provinces. To increase student-centered teaching and learning activities with the use of technology, the province launched the internet and education program (2018–2022) in 2019. As a result, studies on the factors that influence mathematics teachers’ intentions to use technology can provide useful information for the application of technology in disadvantaged areas.

UTAUT was chosen as the research model because UTAUT is a widely accepted model to investigate the acceptance and use of micro-lectures in middle schools. According to the UTAUT model, four factors may influence the BIs of junior high school mathematics teachers using micro-lectures, namely PE, EE, SI, and FC. Figure 1 shows the study model used.

3.1. Procedures

Data were collected from junior high school mathematics teachers in Guangxi Province, China. The questionnaire (see

Table 1. Questionnaire items to analyze the intentions and actual use of junior high school teachers using micro-lectures.

	Construct	References
	Performance Expectancy (PE)	[34,45]
PE1	I think that teaching using micro-lectures is more effective
PE2	Micro-lectures increase teaching productivity
PE3	Using micro-lectures can improve my mathematics teaching performance
PE 4	I think teaching using micro-lectures is efficient
PE 5	Using micro-lectures can increase my employment opportunities
	Effort Expectancy (EE)	[34,45]
EE1	Micro-lectures for mathematics teaching are easy to use
EE2	Interaction with micro-lectures is very clear and easy to understand
EE3	If I can make videos, I can easily make micro-lectures
EE4	It is effortless to use micro-lectures when teaching mathematics
item	Social Influence (SI)
SI1	Mathematics teachers around me use micro-lectures to teach mathematics	[33]
SI2	In general, my school supports me using micro-lectures to teach mathematics
SI3	My students think that I should use micro-lectures to teach mathematics
SI4	Using micro-lectures will increase my social status
	Facilitating Condition (FC)
FC1	Each classroom has the suitable equipment to implement micro-lectures in mathematics learning	[33]
FC2	People around me can help me to use micro-lectures
FC3	I have sufficient knowledge to be able to use micro-lectures
FC4	If I faced difficulties about micro-lectures, there are people or groups who can help me solve the problem
FC5	The school provides trainings on making and usage of micro-lectures for mathematics learning.
	Behavior Intension (BI)
BI1	I plan to use micro-lectures to teach mathematics	[33]
BI2	I plan to use micro-lectures more often
BI3	I feel that micro-lectures should be used to teach mathematics
BI4	I will recommend micro-lectures to other mathematics teachers.
	Use Behavior (UB)
UB1	I use micro-lectures to teach mathematics	[33]
UB2	Micro-lectures have become part of my mathematics teaching
UB3	I use a lot of micro-lectures on math lessons

) was modified from a previous study model of UTAUT questionnaire items [33] and others that used UTAUT in video-based learning [45].

There were two sections in the questionnaire. The first part focused on demographic information about the mathematics teacher respondents. Meanwhile, the second part was the features questionnaire that measured the constructs in the study model. The study model had 25 items in 6 constructs, consisting of PE, EE, SI, FC, BI to use micro-lectures, and UB. Moreover, the questionnaire used a 5-point Likert scale, anchored on “1 = strongly agree” and “5 = strongly disagree”.

Table 1 shows the list of questions that measured the constructs. The original questionnaire data were in English, and were processed by a strict and professional translation performed by three Chinese doctoral students fluent in English. This was then re-checked by two native speakers. Consequently, the initial questionnaire was reliable, while the final questionnaire was placed on a website and distributed to a random sample of junior high school mathematics teachers in Guangxi Province, China, trained in designing and using micro-lectures by Guangxi Normal University.

3.2. Participants

Junior high school mathematics teachers in Guangxi Province, China, were the target respondents. This study investigated their adoption of micro-lectures for teaching mathematics in the classroom. It was performed by contacting the learning media development division at the education department based in Guilin, China. The survey link was distributed to a variety of schools along with the study’s objectives, data collection techniques, and the consent of participation. Furthermore, this survey was performed from 4 to 14 January 2022 and included 174 junior high school mathematics teachers. The remaining 166 pieces of data were valid questionnaires for formal data analysis, except for 8 that were incomplete respondents. Among the 166 participants, 113 (68.1%) were female teachers and 53 (31.93%) were male. In terms of education level, 104 (62.65%) had Bachelor’s degrees, while 62 had Master’s (37.35%). Participants spent a time span ranging from 15 to 25 min filling out the questionnaire.

Table 2. Descriptive statistics of junior high school teacher respondents.

Items	Type	Frequency	Percentage
Gender	Male	53	31.93%
	Female	113	68.07%
Age	20–25	13	7.83%
	26–30	35	21.08%
	31–35	26	15.66%
	35–up	92	55.42%
Education	Bachelor’s	104	62.65%
	Master’s	62	30.72%
Teaching experience	0–5 year	22	13.25%
	6–10 year	40	24.10%
	11–20 year	73	43.98%
	More than 21 years	31	18.67%
Technology experience	Never	33	19.88%
	Rare	67	40.36%
	Often	45	27.11%
	Very often	21	12.65%
School location	Rural	112	67.47%
	Urban	54	32.53%

shows the complete data on the participants.

3.3. Data Analysis

This study used Smart-PLS with a Structural Equation Model (SEM) approach to test the SEM model (not hypotheses) hypotheses and assess the statistical significance of the path coefficients in the research model mentioned above. This approach is frequently used in social science studies because of its accuracy in psychometric model analysis [55]. According to previous studies, the Smart-PLS is used because of the following reasons: (1) hypothesis testing can be performed when the distribution is abnormal; (2) it can be used once the element is less than 3 factors; and (3) it can be used without thinking about the number of samples [64].

The PLS-SEM step consists of reflective measurement and structural model assessments. The reflective measurement model assessment reveals reflective indicator loading, internal consistency reliability consisting of Cronbach’s alpha and composite reliability, convergent validity through Average Variance Extracted, and discriminant validity using the Heterotrait–Monotrait ratio (HTMT). Meanwhile, statistical assessments such as VIF value, path coefficients, t statistics, and p-values are used to evaluate the structural model [55,65]. The t-test was used to compare the actual behavior of micro-lecture use among Chinese mathematics teachers in Guangxi Province, China.

4. Data Analysis and Results

In this study, the steps of assessing structural equation models were in accordance with the recommendations of Chin [66]. Firstly, the descriptive statistics of the measurement instruments are presented. Secondly, an analysis of the measurement models is presented. Thirdly, an evaluation of the structural model and a hypotheses examination is presented.

4.1. Descriptive Statistics

Descriptive statistics were used to respond to the characteristics of respondents’ answers. The mean, standard deviation, skewness, and kurtosis are all included in the descriptive statistics results.

Table 3. Descriptive statistics.

Items		Mean	Std. Dev.	Skewness	Kurtosis
		Statistic	Statistic	Statistic	Statistic
Performance Expectancy	PE1	4.084	0.812	−0.156	−1.467
	PE2	4.048	0.822	−0.090	−1.517
	PE3	4.072	0.743	−0.117	−1.172
	PE4	4.048	0.822	−0.487	−0.433
	PE5	3.771	0.951	−0.467	−0.640
Effort Expectancy	EE1	3.711	1.039	−0.314	−1.054
	EE2	3.205	1.065	0.313	−1.176
	EE3	3.361	1.004	−0.122	−1.200
	EE4	3.518	0.977	−0.249	−0.965
Social Influences	SI1	3.892	0.839	−0.167	−0.834
	SI2	3.807	0.770	−0.138	−0.444
	SI3	3.578	0.949	−0.357	−0.801
	SI4	3.771	0.799	−0.429	−0.096
Facilitating Condition	FC1	3.807	0.978	−0.312	−0.940
	FC2	3.831	1.042	−0.371	−1.080
	FC3	3.639	1.118	−0.300	−1.269
	FC4	3.699	1.030	−0.311	−1.028
	FC5	3.795	1.087	−0.500	−1.027
Behavior Intention	BI1	3.964	0.785	0.064	−1.372
	BI2	3.988	0.901	0.024	−1.778
	BI3	4.096	0.847	−0.186	−1.586
	BI4	4.096	0.847	−0.186	−1.586
Use Behavior	UB1	3.614	1.007	−0.102	−1.069
	UB2	3.831	0.970	−0.219	−1.073
	UB3	4.120	0.858	−0.584	−0.574

displays the results of the descriptive statistics.

The average of the respondents’ answers to each item was 3814, with UB3 (4.12) having the highest average, and EE2 (3.2) having the lowest. This means that most of the relative answers were positive responses. In the pattern for the average of each item in each construct, the average value was not significantly different. This pattern also occurred in the standard deviation. According to the previous study, the limit for skewness and kurtosis was│2.3│ [67]; therefore, based on Table 3, skewness and kurtosis were within the acceptable value range.

4.2. Evaluating the Measurement Model

The structural model produced was revealed by the outcomes of the study in the measurement model. Each relationship between the variables had a loading factor in this structural model. Figure 2 shows the final results.

Figure 2 shows the loading factor for each variable in the range of 0.680 to 0.957 which is a good value. Each variable shows a value that is almost evenly distributed and consistent. Although some loading factor values are still less than 0.700, they should be greater than 0.700, as noted by Hair et al. [68]. This means that the observed variables should be able to explain the latent variables. Furthermore,

Table 4. Results of loading factor, validity, and reliability.

Latent Variable	Indicator	Loading	t-Value	Composite Reliability	Cronbach’s Alpha	AVE
Performance Expectancy	PE1	0.893	57.89	0.932	0.907	0.735
	PE2	0.911	56.855
	PE3	0.937	114.399
	PE4	0.840	37.294
	PE5	0.680	13.589
Effort Expectancy	EE1	0.728	25.933	0.892	0.849	0.676
	EE2	0.848	16.899
	EE3	0.825	14.763
	EE4	0.879	20.875
Social Influence	SI1	0.828	28.518	0.931	0.902	0.773
	SI2	0.957	132.207
	SI3	0.858	28.298
	SI4	0.869	33.725
Facilitating Condition	FC1	0.682	14.534	0.910	0.878	0.672
	FC2	0.814	22.794
	FC3	0.846	33.06
	FC4	0.858	25.916
	FC5	0.883	32.379
Behavior Intention	BI1	0.798	19.337	0.954	0.934	0.839
	BI2	0.953	129.226
	BI3	0.952	130.964
	BI4	0.952	120.649
Use Behavior	UB1	0.895	57.503	0.913	0.854	0.778
	UB2	0.947	159.574
	UB3	0.797	30.215

contains information regarding the measurement model, such as loading, t-value, internal consistency, Cronbach’s alpha, and AVE (Average Variance Extracted).

The measuring model’s convergent validity was demonstrated in the second stage by examining: (1) item reliability; (2) composite reliability; and (3) Average Variance Extracted (AVE). For item reliability, Cronbach’s alpha was used. Table 4 shows that all the constructs of Cronbach’s alpha value with an average of 0.920 were greater than the threshold limit of 0.70. Subsequently, the factor loadings for each item were checked. Any loading factor for each item that was less than 0.5 must be discarded and discontinued in the next analysis [69]. In this study, all items had a factor loading of more than 0.5 with an average of 0.883. Each construct in Table 4 had composite reliability greater than 0.5, showing good internal consistency reliability among the latent variables [70]. Furthermore, to analyze variances, the AVE of all constructs had values greater than 0.5, indicating that they fit the criterion for convergent validity [70]. This means that the measurement process in the developed model was of high quality and can explain the model. The overall fit of the model was within the acceptable range.

To check discriminant validity, this study used the Fornell–Larcker criterion, which suggests using the square root of AVE for each of the latent variables. This is followed by the careful examination of the correlation coefficients between the other variables [70]. In

Table 5. Results of discriminant validity based on Fornell–Larcker criterion results.

	Behavior Intention	Effort Expectancy	Facilitating Condition	Performance Expectancy	Social Influence	Use Behavior
Behavior Intention	0.916
Effort Expectancy	0.566	0.822
Facilitating Condition	0.678	0.628	0.820
Performance Expectancy	0.658	0.317	0.432	0.857
Social Influence	0.811	0.453	0.71	0.557	0.879
Use Behavior	0.814	0.538	0.645	0.496	0.811	0.882

, Fornell–Larcker criterion for discriminant validity is presented by showing the inter-item correlation matrix (the diagonal elements representing the square root of AVE). The diagonal element was observed as greater than the other correlation values between other latent variables, thereby meeting the conditions of discriminant validity. However, several studies showed that using the Fornell–Larcker criterion is insufficient for discriminant validity analysis [70]. To determine discriminant validity, the HTMT ratio is required. According to Naveed et al. [71] and Theo et al. [72], the threshold value maximal value for HTMT is 0.9.

Table 6. Additional validity discriminant measurement results based on HTMT.

Construct	Behavior Intervention	Effort Expectancy	Facilities Conditions	Performance Expectancy	Social Intervention
Behavior Intervention
Effort Expectancy	0.59
Facilities Conditions	0.74	0.69
Performance Expectancy	0.71	0.29	0.49
Social Intervention	0.86	0.46	0.78	0.59
Usage Behavior	0.90	0.59	0.72	0.55	0.83

shows the HTMT statistics, which provides support for the existence of discriminant validity. Therefore, the developed instrument possessed good characteristics to explain the developed model. After analyzing the measurement model, the structural model was examined.

4.3. Evaluating the Structural Model and Hypothesis Testing

In this section, an evaluation of the structural model is performed, as shown in Figure 3. SEM was used to test the hypothetical relationship between variables [73], and between UTAUT factors, Bi, and the UB of micro-lectures in Guangxi Province, China. Figure 3 describes the final structural model based on the results of the refinement criteria of the measurement model sections, which had 25 items.

The overall model was then examined according to the structural model analysis. The most frequently used model sizes are summarized in

Table 7. Direct indirect and total effect implied in the path model.

Factor	Determinant	Effect
		Direct	Indirect	Total
Behavior Intention R2 = 0.751	Performance Expectancy	0.280	0	0.280
	Effort Expectancy	0.215	0	0.215
	Social Influence	0.558	0	0.558
Use behavior R2 = 0.678	Behavior Intention	0.689	0	0.689
	Performance Expectancy	0	0.192	0.192
	Effort Expectancy	0	0.148	0.148
	Social Influence	0	0.385	0.385
	Facilitating Condition	0.184	0	0.184

. Moreover, the results of using Smart-PLS 3 showed an acceptable fit to the data.

According to several experts [33,64], if R² value greater than 0.67 is considered high, the variance between 0.33 to 0.67 is considered moderate, while that between 0.19 and 0.33 is considered weak. Overall, the proposed model accounted for 75.1% variance in mathematics teacher intentions to adopt micro-lectures, and 67.8% of the variance in actual usage (Table 7). Standardized Root Means Square Residual (SRMR) was used to assess the fit in the PLS model. A good fit is defined by an SRMR value less than 0.10 [74,75] and in this study, the SRMR value was 0.123. Although this does not indicate a very good fit model, it has reliabilitsy, validity, and can explain the relationships in the hypothesized paths according to the R² measured. The analysis can then be continued at the next stage [76].

Table 7 shows the information on the direct, indirect, and total effects on each relationship between variables. The majority of each relationship had a moderate effect or more than 0.1. For example, the most influential determinant on BI was SI with a total effect of 0.558. PE was the second biggest factor with a total effect of 0.280, while EE was the last one, with a total effect of only 0.215. A direct effect was responsible for all of the total impacts. The two main determinants influencing UB were BI and FC, with direct effects of 0.689 and 0.184, respectively. Other determinants included the indirect effect of PE of 0.192, EE of 0.148, and SI of 0.385. According to Cohen [77], BI to use micro-lectures had a large total effect (more than 0.5), and all effect sizes were greater than 0.1, as shown in

Table 8. Hypothesis testing of factors affecting the use of micro-lectures.

Hypothesis	Relationship	Path Coefficient (B)	Sample Mean	Standard Deviation (SthEV)	t Statistic	p Values	Decision
H1	Performance Expectancy -> Behavior Intention	0.522 ***	0.252	0.500	5.566	0.000	Significant
H2	Effort Expectancy -> Behavior Intention	0.227 ***	0.215	0.063	3.587	0.000	Significant
H3	Social Influence -> Behavior Intention	0.553 ***	0.558	0.043	12.815	0.000	Significant
H4	Facilitating Condition -> Use Behavior	0.182 **	0.184	0.066	2.746	0.006	Significant
H5	Behavior Intention -> Use Behavior	0.689 ***	0.689	0.067	10.299	0.000	Significant

**, *** Indicate significance at the 0.05, and 0.01 levels (two-tailed), respectively.

.

Table 8 displays the path coefficient, sample mean standard deviation, t-statistics and the significance level (p value). Since the t-statistics are greater than 1.96 and the p-value is less than 0.05, all pathways exhibit significant results [56]. PE shows a significant positive effect on micro-lecture BI (β = 0.522, t-statistic = 5.556 p = 0.000), which supports Hypothesis 1. EE shows a significant positive effect on micro-lecture BI (β = 0.227, t-statistic = 3.587, p = 0.000), and this supports Hypothesis 2. For SI, a significant positive effect is observed on micro-lecture BI (β = 0.553, t-statistic = 12.815 p = 0.000) and this provides support for Hypothesis 3. FC shows a significant positive effect on micro-lecture UB (β = 0.182, t-statistic = 2.746 p = 0.006), and Hypothesis 4 is supported by this. Micro-lecture BI, as expected, has a significant positive effect on micro-lecture UB (β = 0.689, t-statistic = 10.299 p = 0.000), which provides support for Hypothesis 5.

5. Discussion and Implication

This study focused on identifying factors in the UTAUT model that could influence mathematics teachers’ adoption of micro-lectures. Our research also aimed to determine the factors that could have the greatest effect on BI and, in line with the results, all factors affected BI. This is consistent with other studies on technology acceptance [32,55,60,78] m-learning adoption [79], animation usage [80], and ICT-based instruction [55].

In this study, SI focused more on the teacher’s confidence in principals, curriculum leaders, friends, and students’ thoughts on the use of micro-lectures in teaching mathematics. A surprising finding was that SI was the most powerful factor motivating BIs to use micro-lectures, despite earlier studies indicating that SI did not affect BIs [52,56]. However, this is consistent with previous studies which showed that SI has a positive influence on teacher BIs to use ICT [32,50,60]. Therefore, this means that teachers tend to use micro-lectures once other important people around want them to use this tool to teach mathematics. SI is the biggest factor influencing BI, and this is supported by previous studies which showed that Chinese people have collectivist cultural orientations [32], in cases of strong “we” power. They are more concerned with work performance and prefer cohesiveness in the workforce [81]. In addition, the ministry of education in China strongly recommended using technology to teach in the classroom.

The second factor influencing BI to use micro-lectures was PE. This is consistent with previous results [50,60,82,83] that show the influence of PE on teacher acceptance of ICT. However, it is inconsistent with those studies who showed no relationship between PE and BI [53]. The teachers’ main objective was to increase their performance and quality in teaching mathematics, as expected, and they believed that micro-lectures were a tool that could assist them. According to these results, teachers, experts, administrators, and developers of micro-lectures should focus on efficiency and efficacy that could improve teacher performance and the quality of mathematics teaching.

The factor that had the least impact on BI was EE, and this is consistent with other studies [60,82,83]. However, this contradicts previous studies which show that EE has the largest statistically positive effect that influences teachers’ BIs [48]. A surprising finding was that this was inconsistent with previous results on micro-lectures with no effect of easy use on BIs [10]. Therefore, further analysis is required for clarification. Many findings revealed that EE does not affect technology use [45,53,61,78]. Following the results of this study, teachers found micro-lectures to be simple to create and use while teaching mathematics; thereby, using micro-lectures could have a favorable impact on BIs. In making this learning tool, academicians and experts should consider and design micro-lectures that are easy to use, easy to display on classroom computers and various devices, and do not need the internet. These circumstances could motivate teachers to adopt micro-lectures for learning mathematics in the classroom.

Previous studies revealed that FC did not affect teacher usage technology in classrooms [43,58,84]. However, this study showed that FC affects teachers’ use of micro-lectures for teaching mathematics. This finding is supported by Ma et al. [32], who suggests that teachers will use micro-lectures once the necessary resources and technical support are available. Mathematics teachers stated that using micro-lectures in mathematics once they were studying was not much different from using other technologies, namely dynamic mathematics software or game-based learning. Furthermore, this experience provides them with the knowledge and habits to integrate technologies into mathematics lessons. Once teachers run into difficulties creating or using micro-lectures in the classroom, there are always colleagues or micro-lecture experts willing to assist. This finding suggests to schools that teachers would be motivated to use micro-lectures if they were provided training on making them. Many micro-lectures can be used to teach mathematics, as can the facilities that support their use in classroom mathematics lessons. To assist teachers while experiencing difficulties both in the preparation and use of micro-lectures in the classroom, supportive and expert staff should be available at any time.

Finally, Behavior Intention (BI) was suggested as the greatest impact on teachers’ usage of micro-lectures. This shows that high levels of BI have an impact on usage.

Behavior Intention and the usage of micro-lectures by junior high school mathematics teachers are shown in this result to help decision-makers in schools to use appropriate technology to assist teaching and learning activities. Additionally, this likely improves teachers’ performance and teaching effectiveness to improve students’ academic performance. This could be important for students who find mathematics to be a difficult and boring subject [85,86]. Micro-lectures have been proven to increase students’ academic achievements and make mathematics more enjoyable [3,20,87].

Our quantitative study indicated that Social Influence had the greatest impact on the Behavior Intention of mathematics teachers in Guangxi Province, China. Social Influence implies, among other things, that teachers place a high value on the opinions of key players in their environment regarding using different digital technologies, in our case micro-lectures. In this context, it seems essential that pedagogical key players within a region communicate the preferred digital technologies of a modern classroom, and that such information is not based on hearsay. Since Social Influence is the most crucial factor for the teachers in our study, ambiguous or unknown information regarding the preferences of pedagogical key players on preferred digital technologies could lead to uncertainty among teachers, which could slow down or disrupt the integration process of digital technologies in classrooms.

Furthermore, in line with our study and other studies on using technologies (e.g., [50,60,83,84]), Performance Expectancy influenced the Behavior Intention of mathematics teachers. Performance Expectancy describes the extent to which teachers believe that the use of technologies can improve their work performance. For teachers’ Performance Expectancy to occur and for teachers to recognize their improved performance, it may be helpful for teachers to learn to set realistic performance goals and define indicators to monitor these performance goals. Setting performance goals and associated indicators can be challenging for teachers at the beginning of such an approach. In order for teachers not to be overwhelmed by this challenge, but still to recognize their increased work performance, it might be beneficial for teachers to be accompanied by experts at the beginning of this process. These experts should help teachers formulate ambitious but not unattainable goals, and define indicators to measure the achievement of these goals. In this way, the achievement of goals concerning the increase in one’s work performance could be objectified, which could reduce teachers’ uncertainty in connection with the use of digital technologies for mathematics learning processes.

The slightest but still statistically relevant effect on Behavior Intention in our study was the Effort Expectancy of mathematics teachers. In simple terms, Effort Expectancy is the ease of use of a technological medium. Since micro-lectures, very short learning videos of a mathematical concept, are basically easy to use, the digital learning environment should not cause any problems when using videos. Possible problems could be that new player software is needed to play the videos, or that the videos can only be played online, and therefore the students have to be on the internet if they want to access the videos.

6. Conclusions

This study’s main purpose was the evaluation of teachers’ determinants of behavior concerning the use of micro-lectures to teach mathematics in junior high schools. The finding showed that teachers adopting and using micro-lectures were influenced by Behavioral Intention (BI) which, in turn, was influenced by Performance Expectancy (PE), Effort Expectancy (EE), and Social Influence (SI). Furthermore, Behavioral Intention of mathematics teachers’ use of micro-lectures in the classroom was most significantly affected by Social Influence.

Since the effect of using technology on mathematics promises to improve students’ mathematical abilities, principals, leaders, the curriculum should support the use of micro-lectures to teach mathematics. This indicates that the schools should conduct training on the creation and use of micro-lectures in mathematics. Support facilities should also be provided as well as tools and facilities for teachers and students for teaching and learning activities.

7. Limitations and Future Research

In this study, the sample was limited to junior high school mathematics teachers in Guangxi Province, China. The factors influencing BI and the use of micro-lectures in mathematics lessons may differ in other countries. However, the headmasters and students were not included as samples. This study did not analyze the moderating effect of gender, age, and experience, and this can influence teachers’ use of micro-lectures in mathematics classes. Additionally, this offers the potential for further study to determine the effect of gender, age, and experience on the BIs of mathematics teachers, particularly on the use of micro-lectures in mathematics. Another limitation is that there were no negative statements in this questionnaire. Therefore, it may not have been able to accurately measure the consistency of respondents’ answers.

Another recommendation for further study is the exploration of all provinces in China, including South, North, and Central China. Secondly, further study can use other methods such as interviews and observations to ensure more specific and convincing results. Thirdly, comparisons between countries, such as that performed by Efiloğlu Kurt [88], can also be accomplished for a better understanding of the factors that influence teachers’ use of micro-lectures in global mathematics lessons.

In future research, we will also investigate a reverse influence of the factors in our study. Investigating a reverse influence of the factors means, for example, investigating whether teachers can convince themselves that micro-lectures are good after they have decided to use micro-lectures, because people tend to justify their decisions.