1 Introduction

Information and communications technology provides a change in the educational context. Because applying technology in face-to-face instruction is an intriguing approach, a number of related studies have been conducted in novel ways. As stated by Peng and Fu, (2021), face-to-face instruction as a traditional method is less favored than online learning, because online instruction is able to provide varied teaching and learning materials, adjust students’ diversity, and fulfill varied students’ needs. Nevertheless, as added by Wu et al., (2010), face-to-face instruction should still be applied because social interaction and direct communication with peers in online instruction are limited. According to Ituma, (2011), for these reasons, in-person and online instructions are combined.

A fusion of traditional and online learning brings conceptualization, which offers a new method for teaching and learning. The idea to integrate teaching material with this online intervention is considered to improve the learning outcome from traditional to in-person learning fully. As stated by Broadbent, (2017); Yustina et al., (2020, 2022), this approach combines the benefits of in-person and online learning simultaneously if those are implemented properly.

Porter et al., (2014) state that blended learning refers to a technology-mediated face-to-face combination. Since the early 2000s, various blended methods of online with traditional face-to-face instruction, namely blended, hybrid, flipped, or inverted, have been adopted for educational institutions. Blended learning as a teaching innovation is closely related to technology. Currently, blended learning is the most effective and popular teaching model implemented for educational institutions as it is also considered effective in providing flexibility, appropriateness, and sustainability. In other words, as stated by Baris, (2015); Rasheed et al., (2020); Tong et al., (2022), several studies should be conducted to see how blended learning to be an effective learning for future trends. Graham et al., (2019) add that blended learning is able to increase the interaction between teachers and students, flexibility, and pedagogical improvements. It also contains values of learning interaction and involvement needed by students.

Blended learning, in fact, has not been well implemented in several countries, especially in developing countries. Some may not have stable access or affordable internet access, computer, or mobile devices required for blended learning (Abbasi et al., 2020; Seraji, 2022). The main problems in the implementation of blended learning are the contexts of local, cultural, and people needs (Ashraf et al., 2021; Seraji, 2022). Successful blended learning implementation requires an in-depth understanding of the factors. Several countries, especially in the developing countries, still maintain traditional values. Education is considered a hands-on activity process in the classroom wherein the teacher is directly present. People are not familiar with the concept and benefits of blended learning so they are not motivated to take part in the blended learning implementation (Ashraf et al., 2021; Bordoloi, 2021). That is, a new approach such as blended learning may not be readily accepted as it will change a new paradigm. There is a need to educate the public on the importance of blended learning in the future. The essential of the present study, therefore, provides an understanding for the public, especially for educational institutions/schools or teachers, to be able to design blended learning implementation, which is able to foster students’ positive attitudes and motivation, in order to achieve learning satisfaction.

Blended learning, which combines online and offline learning elements, has provided a number of attitude-related cases that would be the beginning of a change from the previous paradigm. A number of students had difficulty in using technology and accessing online learning platforms (Vallee et al., 2020). Students might have difficulty using online learning platforms or encounter technical issues such as unstable internet connection and difficult-to-operate hardware or software glitches. This can influence their attitudes on blended learning. Negative attitudes towards tools and online learning platforms can influence their contribution and participation. Blended learning requires a high motivation level. Students might have difficulty keeping motivated when they have to organize their time and study independently, especially in blended learning (Kenney & Newcombe, 2011). Students need to have good time management skills to balance online and offline learning. This incompetency can produce stress and motivation loss. Attitude and motivation can produce an equilibrium between students’ skills and challenge levels. This equilibrium can increase learning satisfaction. These three elements influence each other and create a positive learning environment. A positive attitude and motivation in learning, either online or offline learning, can increase learning satisfaction. As stated by Huang (2021), students with good attitudes tend to be more open to learning experiences and be more able to enjoy the learning process.

Several studies examined teachers' and students’ understanding of blended learning and the studies revealed the positive effect of applying blended learning concerning students’ learning achievement (Akgündüz & Akinoglu, 2017; Pinto-Llorente et al., 2017; Wichadee, 2017; Zheng et al., 2018). In addition, Lane, (2021) states that emotion is an indicator of students’ satisfaction that should be achieved. For learning improvement, in-depth investigation should be applied as a determining factor for students. The student satisfaction level is a major focus as an evaluation in blended learning whether it is effective or not. Several studies investigated students’ satisfaction in the blended learning implementation (Banerjee, 2011; Huang, 2021; Naaj, 2012; Prifti, 2020; Rahman et al., 2015; Taghizadeh & Hajhosseini, 2021; Wu et al., 2010). The academic discipline, however, still lacks in-depth studies about attitude and motivation as the factors that are involved in getting satisfaction in blended learning. The objective of the present study is to explore factors influencing learning satisfaction in blended learning. Attitude and motivation are the variables in this paper, which exhibits how the variables influence students’ learning satisfaction. This study, moreover, aims to explain the differences between attitude, motivation, and learning satisfaction of students based on gender and scientific field. As such, the research questions of this study are:

  • RQ1Is there an effect of attitude towards motivation in blended learning?

  • RQ2Is there an effect of attitude on learning satisfaction in blended learning?

  • RQ3Is there an effect of motivation towards learning satisfaction in blended learning?

  • RQ4Is there an indirect effect on learning satisfaction in blended learning wherein motivation is intervening?

  • RQ5How do students’ attitude, motivation, and learning satisfaction differ based on gender in blendedlearning?

  • RQ6How are there differences in students’ attitudes, motivation, and learning satisfaction based on the scientific field of blended learning?

In terms of answering the present study, this paper contributes to the theory and practice. The present research finding, practically, provides information to the teachers and educational institutions to be able to increase the students’ attitude and motivation in blended learning in order to achieve students’ learning satisfaction. Therefore, blended learning can be implemented sustainably by focusing on the students’ characteristics. In addition, this study provides useful information for the stakeholders who were involved in the blended learning implementation by giving students’ learning satisfaction structure models to improve the quality of education.

2 Theoretical background

2.1 Blended learning

Blended learning is an approach that combines face-to-face and online teaching instructions. Even though it provides many benefits, several challenges should be considered. A number of researchers have been investigating the challenges in blended learning. As stated by Ashraf et al.,(2021); Rasheed et al., (2020), and Kumar et al. (2021), technology is needed in blended learning as it can be a challenge for lecturers who are not convenient in using technology. This will certainly influence the blended learning implementation. Teachers and students use technology effectively, it can be a challenge if they do not get adequate training, or they do not have the required skills. Teachers as educators must have self-motivation to be able to improve their ICT skills (Copriady, 2014), hence blended learning can be implemented well and learning satisfaction can be achieved.

According to Ashraf et al (2021), assessing students’ learning activity in blended learning can be a challenge for teachers, especially if the students learn in different ways and lifting speed.

Adapt to blended learning: It requires students to have self-control, responsibility, and high digital literacy levels. As asserted by Ashraf et al (2021) and (Albeta, Suarman, et al., 2023b), planning, following, and adapting are important for students, as well as using the platform and online tools appropriately.

To overcome these challenges, it is crucial to make adequate training and support for teachers and students, ensure that every student has access to technology, plan and make an appropriate assessment in blended learning (Albeta et al., 2023a; Namyssova et al., 2019; Rasheed et al., 2020). By facilitating the existing learning, it ensures that knowledge can be obtained online and offline in blended learning. The improvement of attitude, motivation, and learning satisfaction in blended learning, therefore, can be achieved.

2.2 Students’ attitude in blended learning

Students’ attitude toward blended learning tends to be varied. In fact, it is not clearly stated how students’ general attitude toward blended learning influences the students’ comfort during the learning process. It provides an insight into students’ attitudes toward blended learning. A number of studies indicate that there is a more significant improvement in students’ knowledge and attitude in blended learning than in lecture-based traditional learning method (Bazelais & Doleck, 2018; Vallee et al., 2020). Blended learning enables the improvement of interaction between teachers and students that creates a better learning outcome.

As mentioned by Lozano-Lozano, (2020) Edward et al (2018) and Tong et al (2022), blended learning also provides positive impact on students’ motivation and achievement in every subjects. In addition, blended learning provides flexibility and self-sufficiency. The students have the opportunity to explore ideas and learn based on their lifting speed, anytime and anywhere. Hence, blended learning can increase students’ understanding and conceptual mastery, as well as students’ learning attitude (Youde, 2020; Zibin & Altakhaineh, 2019). In this study, we compared students’ attitudes towards blended learning implementation based on the genders and discipline. Thus, it can be knowledge for academics to take actions in the future.

2.3 Learning motivation in blended learning

Learning motivation is encouragement and power for someone to be involved in learning activity and achieve their learning goals. Motivation to learn refers to a combination of internal and external factors that influence how intense an individual’s effort is in understanding and mastering learning material. Two types of learning motivation are intrinsic and extrinsic motivation. Intrinsic motivation is a dominant type in blended learning. The students who are motivated intrinsically will complete tasks and produce better learning than the students who are motivated extrinsically (Law et al., 2019). A number of studies emphasize the essential of motivation as the impact on blended learning (Edward et al., 2018; Law et al., 2019; Rahayu & Iswari, 2021; Sabah, 2019).

Interest and fascination on a subject or topic can be a strong aspect for learning. High learning motivation tends to encourage active engagement in learning. This could involve participation in discussion, additional research, or various activities that increase students’ understanding and involvement. Motivation to learn is considered as independent variable or causative factors (Law et al., 2019; Peng & Fu, 2021; Sabah, 2019). A person’s motivation level, influenced by various factors such as goals, interests, and expectations, may cause involvement in learning. A person’s motivation level can influence the extent to which learning satisfaction can be achieved. The higher the motivation, the more likely individuals are to feel satisfied with their learning process and achievement. Success in achieving goals fueled by learning motivation can contribute positively to students’ self-confidence. In fact, the heterogeneity of students in online learning in blended learning cause bigger challenge for teachers to understand students’ motivation (Stavredes, 2011). More research is, therefore, required to provide information on how students’ motivation is formed in blended learning.

2.4 Learning satisfaction

Learning satisfaction is a positive feeling and satisfaction that a person feels towards the learning process and the achieved results. It involves individual’s subjective perception of the extent to which their needs, expectations, and learning goals are met. Learning satisfaction can be related closely to the gained understanding and the level of achieved achievement in learning. Learning satisfaction is used to evaluate and assess the learning effectiveness, including blended learning (Ma & Lee, 2021). One of main factors to measure satisfaction and learning success is through students’ attitude towards the learning (Selim, 2007). Learning satisfaction will continue to influence the students’ learning experiences. When the students met the criteria or exceeded expectations in learning, they would feel satisfied.

A number of experimental research indicates that learner performance and students’ learning satisfaction are influenced students’ attitude towards teacher, motivation, and their learning experience. The higher the students’ learning motivation, the learning satisfaction in the learning process will be greater (Huang, 2021). Based on those literature, the present study indicates that attitude and learning motivation in blended learning have a considerable effect towards learning satisfaction.

2.5 The present study

Blended learning began to be actively implemented in Indonesia during the Covid-19 pandemic and a number of schools still have been implementing it up to now. Thus, it is essential to see the students’ learning satisfaction as the effective assessment of blended learning. Regarding that goal, it is a similar line with this research (Ma & Lee, 2021), however we explain students' attitude and motivation factors to achieve learning satisfaction.. The present study also explains the difference in attitude, motivation, and satisfaction based on students’ gender and scientific field.

Through vosviewer application, we examine the research in the last three years that is related to student attitude in blended leaning. Based on Fig. 1, the research concerns students that are related to student experience, knowledge, blended learning course, and so on. Figure 1 presents that the study about student attitude and motivation achievement is still not linked yet directly to. Hence, a study is required to be able describe the interrelationships between attitude and motivation. In addition to Fig. 1 description, there are no studies about satisfaction, attitude, and motivation in blended learning conducted in the last three years. For that reason, we analyze satisfaction that is related to students’ attitude and motivation.

Fig. 1
figure 1

A Research Conducted in 2020–2023 Regarding Student Attitude in the Blended Learning Implementation

Full size image

3 Method

The present study utilized a quantitative variance-based structural equation modeling method to determine the procedure that should be followed by students so they could achieve learning satisfaction completely in the blended learning implementation through the variables of attitude and motivation. Smart PLS V.3.2.7 was used to analyze the structural models and measurements based on the data collected through a questionnaire. In addition, the present study applied IBM SPSS 22 to test the difference between the variables based on the student’s gender and scientific field.

3.1 Hypotheses and model

In particular, this paper discusses how attitudes, learning motivation, and students’ satisfaction in blended learning are interrelated with each other. Figure 2 is the Model, which are seen as follows:

Fig. 2
figure 2

Proposition for Satisfaction Model in Blended Learning

Full size image

Based on the research objectives and model Fig. 2, the research hypotheses are:

1. The effect of variables/constructs in the study.

This study investigates the factors influencing learning satisfaction in the blended learning implementation. We summarize four hypotheses that are tested by using structural equation modeling.

A person’s attitude plays an essential role in influencing their motivation. If a person has positive attitude towards learning, they tend to be motivated to accomplish academic tasks. Students’ attitude and motivation are the most significant influencing factors in learning (Pariwat, 2020). Students who are successful in learning have attitude and motivation. It results in a better attitude leading to an increase in commitment and an increase in class participation. (Ahmad Baaqeel, 2020). Therefore, the hypothesis that is proposed in this study can be described as follows:

H1: Attitude → Motivation. This hypothesis refers to the attitude that can influence motivation.

Attitudes possessed by students totally influence their level of learning satisfaction in the learning process. Students who have positive attitude towards blended learning technology are caused by several factors like provision of differentiated learning options through accessible and diverse materials. Thus, it can increase learning satisfaction and achievement (M J Kintu, 2017; Taghizadeh & Hajhosseini, 2021).

H2: Attitude → Satisfaction in blended learning. This hypothesis regarding attitudes directly influences learning satisfaction in blended learning.

Students’ learning experience will be influenced by their learning satisfaction. The positive feeling and satisfaction are obtained from a good learning process experienced by them in learning activities. The stronger the students’ learning motivation is, the greater their learning satisfaction is in learning process (Huang, 2021; Kuo et al., 2013; Lovecchio et al., 2015). As such, learning motivation in blended learning influences learning satisfaction. Therefore, the following hypothesis is:

H3: Motivation → Satisfaction in blended learning. This hypothesis comes from the influence of motivation on satisfaction in blended learning.

Of the explanation in favor of HI, H2, and H3, we extend the assertion that the attitude provides indirect effect on satisfaction in blended learning, wherein motivation is categorized as intervening. Therefore, the following hypothesis is:

H4: Attitude → Motivation → Satisfaction. This hypothesis indicates attitude that provides an indirect effect on satisfaction in blended learning where motivation is categorized as intervening.

2. The difference based on the students’ gender and scientific field on each variable.

We conclude that there are six hypotheses in the comparison of each variable based on gender and scientific field. The hypotheses can be described as follows:

  • H5a: There is a difference in students’ attitude on blended learning based on gender.

  • H5b: There is a difference in students’ motivation on blended learning based on gender.

  • H5c: There is a difference in learning satisfaction on blended learning based on gender.

  • H6a: There is a difference in students’ attitudes toward blended learning based on the student’s scientific field.

  • H6b: There is a difference in students’ motivation for blended learning based on the student’s scientific field.

  • H6c: There is a difference in learning satisfaction on blended learning based on the student’s scientific field.

3.2 Participants and data collection

We involved Senior High School students who have been implementing blended learning in this study. The sample selection was carried out using a simple random technique. Sample size is recommended to be carried out in order to gain consistent outcomes in terms of the required measurement model, which has high sample size qualification from the technique analysis. The participants involved Senior High School students in Indonesia, especially in Riau province, aged 16 up to 18 years old. The research subjects involved 488 students, which were 287 female students with a percentage of 58.8% and 201 male students with a percentage of 41.2%. Moreover, there were 264 science students (54,1%) and 264 non-science students (45,9%). We used IBM SPSS 22 to analyze the demographic data of the study participants. The demographic details of the students on Table 1 can be described as follows:

Table 1 Demographic details of the students
Full size table

For such investigations, the survey using a questionnaire are considered the most effective method for assessing the relationship between the dimensions in the research model (Al-Maroof & Al-Emran, 2018). Data collection was conducted from April to May 2023 through an online survey. The online questionnaire was shared among students willing to participate through social media platforms. The participants were provided with information to fill out the questionnaire. Firstly, the survey comprised the demographic information of the respondents. Second, the survey contains statements related to attitudes, motivation and learning satisfaction felt by students when implementing blended learning.

3.3 Instruments

In terms of developing questionnaire, we put items from the previous validated instruments. For attitude, we adopt an item from Ayanwale et al., (2023). Untuk motivasi kami mengadopsi item dari Raza et al., (2020) dan Wong et al., (2020). Redarding learning satisfaction, an item was adopted from Eom et al.,(2006) and Chiu et al., (2005). Based on the adopted items, we classify into several indicators to represent the measured variable or construct. Then, the constructs of the conceptual model are measured such as attitude, motivation, and satisfaction, which were measured by using a “5-point Likert scale”. Responses consist of “strongly disagree” (later coded as 1), “disagree” (2), “neutral” (3), “agree” (4), and “strongly agree” (5). It can be shown on Table 2 as the distribution of indicators based on each measured variable.

Table 2 Distribution of Indicators
Full size table

The measuring instruments in this study consisted of three constructs with 13 items that can be indicator from the three constructs. The instrument used in this study is tested to find out whether the instrument is reliable or valid. The resukt indicates that the steps have an acceptable benchmark with high score to realibity and validity (as seen on validity and realibility tests). The measurement includes attitude, learning motivation, and learning satisfaction on the blended learning implementation. It shows that the measurement rates the construction to be assessed accurately and question items put in this study are consistent. The steps used in this study, overall, are considered as an effective and appropriate tool for blended learning satisfaction based on the aspects of students’ attitudes and motivation. The questionnaire used can be seen on the appendix.

3.4 Data analysis

Smart PLS V.3.2.7 and IBM SPSS 22 were used to analyze the structural models and measurements based on the collected data through a questionnaire. Software SmartPLS V.3.2.7 is used to analyze the structural models. This study used PLS-SEM because this approach could manage the complex models with several indicators, construct, and correlational models with no causal-predictive focus and identification problem. Partial Least Squares Structural Equation Modeling (PLS-SEM) includes measurement model narrative and structural model details (Ringle et al., 2020). Measurement model description indicates that the indicator contains acceptable convergent validity, composite reliability, and discriminant validity, and subsequently is used on the structural model. The need and path coefficient evaluation is required for structural model evaluation. The selection of PLS-SEM is attributed to several reasons, 1) PLS-SEM is regarded as the ideal option for developing theories in the research, 2) PLS-SEM includes having moderated and mediating variables and the model complexity, 3) PLS-SEM can play complex models that include a lot of constructs, 4) the entire model as a single unit is tested through PLS-SEM rather than separating it into units, and 5) PLS-SEM indicates simultaneous analysis for both measurement and structural model, resulting in more accurate evaluation (Ringle et al., 2020). Chin, (1998) proposed sample size for PLS requirements should be aggregated by 10 times based on most research construct indicators. The most essential indicator is the construct of learning motivation. There were 5 question items. Hence, the minimum research sample should be 50. The research sample in this study was 488, which qualified the standard criteria of sample size. PLS-SEM is used to test the hypotheses of HI, H2, H3 dan H4.

IBM SPSS 22 is used to see the research result descriptively and the result of a comparative test from each variable of attitude, motivation, and learning satisfaction based on the student’s gender and scientific field. Independent sample t-test is used to analyze hypotheses H5a, H5b, H5c, H6a, H6b dan H6c. In conclusion, if the probability (p) Value is less than or equal to 0.05 (p < 0.05), the hypothesis is rejected. If p > 0.05, the hypothesis is accepted. Before testing the independent sample t-test, data should be qualified, namely normality and homogeneity tests. Normality testing in this study is used through the Kolmogorov–Smirnov test. While Anova test is used to test the homogeneity test.

4 Results

4.1 Exploration of effect between the students' attitude, motivation, and learning satisfaction variables in the blended learning implementation

4.1.1 Measurement model assessment

Hair et al., (2019) suggest assessing the measuring model’s construct reliability (Cronbach’s alpha and composite reliability) and validity (convergent and discriminant validity). In order to measure the construct reliability, the results in Table 3 reveal that Cronbach’s alpha values range between 0.720 and 0.908, which are above the standard value of 0.7. The results in Table 3 also indicate that the composite reliability (CR) values range between 0.830 and 0.932, which are above the recommended value of 0.7 (Hair et al., 2019). Hence, the construct’s reliability has been proven, and all the constructs are sufficiently error-free. In terms of measuring convergent validity, the factor loading, and average variance extracted (AVE) should be examined (Hair et al., 2019). Therefore, Table 3 indicates that the values of all factor loadings are higher than the suggested value of 0.7. In addition, Table 3 reveals that the AVE value ranges around 0.620 and 0.725, which are above the standard value of 0.5. For this reason, all constructs’ convergent validity requirements have been properly met. It can also be seen as follows:

Table 3 Convergent Validity and Reliability Measurement
Full size table

For the measurement of discriminant validity, three criteria are measured. It includes the Fornell-Larker criterion, cross-loadings, and the Heterotrait-Monotrait ratio (HTMT) (Hair et al., 2019). Table 4 shows that the Fornell-Larker criterion is validated due to the fact that the square roots of all AVEs are greater than its correlation with other constructs (Fornell & Larcker, 2016).

Table 4 Fornell-larcker scale result
Full size table

As illustrated in Table 5, the results indicate that the cross-loadings also met the criteria since the indicator loadings on each construct are higher than the loadings of its corresponding constructs.

Table 5 Cross-loading result
Full size table

Table 6 demonstrates the HTMT ratio results, which clearly exhibit that the value of each construct does not exceed the threshold value of 0.85 (Henseler et al., 2015). Consequently, the HTMT ratio is confirmed. In addition, the discriminant validity is ascertained. The analysis results furnish evidence that there are no issues regarding the assessment of the measurement model in terms of its reliability and validity, and hence, the collected data can be further used to evaluate the structural model.

Table 6 Heterotrait-Monotrait Ratio (HTMT)
Full size table

4.1.2 Structural equation modeling analysis

When evaluating structural equation modeling, Variance Inflation Factor (VIF) is higher than 5. It means that there is a collinearity factor from both dimensions. The structural equation modeling value of VIF in this study is less than 5, which is between 1 and 1.609. That is, there is no collinearity between the research dimensions. SRMR, NFI, and RMS theta are general indicators used to PLS-SEM to evaluate overall model fit. The value of SRMR is around 0 to 1. As suggested by (Hu & Bentler, 1998), if SRMR is less than 0.08, it is regarded as an appropriate model. The SRMR value from verified learning model evaluation, in this context, is 0.069. Hence, the model is quite appropriate in general. Table 7 is collinearity analysis and model fit, which can be described as follows:

Table 7 Collinearity analysis and model fit
Full size table

Furthermore, verified models are analyzed and explained with path analysis and R2. For path analysis, t-value is used to determine the validity of hypothesis. If t-value is higher than 1.96 (t > 1.96), it means that the significant level is 0.05 (shown with *). If t-value is higher than 2.58 (t > 2.58), it means that the significant level is 0.01 (indicated with **). If t-value is higher than 3.29, it means that the significant level reaches out to 0.001 (shown with ***). It all can be seen in Table 8, which means that H1, H2, and H3 have reached a significant level in which p-value is lower than 0.001. Therefore, the hypotheses of H1, H2, and H3 are supported in this study. Path analysis model of PLS-SEM is shown in the following Fig. 3.

Table 8 Hypotheses testing results
Full size table
Fig. 3
figure 3

Model of PLS-SEM Path Analysis Diagram

Full size image

Three variables are described in Table 2 above. The three variables have indicators. Attitude contains three indicators, motivation has five indicators, and satisfaction consists of five indicators. In addition, the relationship between the variables is analyzed by using PLS-SEM, which consists of the analysis of attitude on motivation (H1), the analysis of direct attitude relationship on satisfaction (H2), the analysis of motivation relationship on satisfaction (H3), and the analysis of indirect attitude relationship on satisfaction through motivation as intervening or mediation (H4).

Intervening effect testing can also be detected by the t value of the indirect effect. t value is higher than 3.29 (t > 3.29), which means that the significant level achieves the significance level of 0.001 and indicates that there is an intervening effect. Motivation is used in this study as intervening or mediation from the relationship between attitude and satisfaction. The indirect effect value is 0.344 and t-value is 10.348, which reaches a significant level. Hence, the hypothesis of H4 is supported. It means that there is motivation in the effect of attitude on blended learning satisfaction as intervening or mediation. Intervening effect testing is shown in Table 9.

Table 9 Intervening Effect Testing
Full size table

R2 value is used to evaluate the explanatory model. R2 value is between 0 and 1. The higher the value is, the higher the explanatory value is. When R2 value is close to 0.50, the model is categorized as a moderate explanatory value. When R2 is close to 0.75, the model is classified as high explanatory value. It can be seen in Table 10 that the explanatory value of attitude on motivation is 0.379. On the other hand, the explanatory value of motivation on satisfaction is 0.485. Hence, the model in this study presents latent variables properly and has equal moderate explanatory value.

Table 10 R2 value and f2 value
Full size table

Effect size is the exogen variable effect on the endogen variable using explanatory effect value f2 to detect it. If 0.02 < f2 ≤ 0.15, it has a small effect. If 0.15 < f2 ≤ 0.35, it has a moderate effect. In addition, if f2 > 0,35, it indicates a big effect. It can be seen in Table 10 that the explanatory effect value f2 of attitude on motivation is 0.609. It indicates the explanatory skills with a high effect. The explanatory effect value f2 of motivation on satisfaction is 0.376, which indicates high explanatory effect skills. It implies that the exogen variable is able to present endogen variables with a high explanatory effect value level.

4.2 The difference based on the students’ gender and scientific field on each variable

The data are analyzed by using IBM SPSS 22, namely an independent sample t-test. The pre-requisite tests are normality and homogeneity tests before doing an independent sample t-test. Normality testing data can be seen on Table 11. Every variable has a sig value > 0.05. Hence, the data on attitude, motivation, and learning satisfaction are categorized as normal.

Table 11 Normality Data
Full size table

It is similar to the homogeneity test on Table 12. The Variables of attitude, motivation, and learning satisfaction have p-value (probability) > 0.05 so the data is categorized homogeny.

Table 12 Homogeneity Data
Full size table

4.2.1 Analysis of gender differences

The differences in attitude, motivation, and student’s satisfaction with the blended learning implementation based on gender are analyzed in this study. The result of the Independent Sample T-test indicates that attitude, motivation, and student’s satisfaction with the blended learning implementation based on gender reaches a significant level. The variable attitude indicates p = 0.00 and the variable motivation shows p = 0.0 with P value < 0.05. Hence, a significant difference between male and female students in the blended learning implementation is found. Male students have a higher average value than female students’ average value. Otherwise, the variable of satisfaction indicates p-value of 0.45, which is higher than 0.05. Therefore, there is no significant difference between male and female students in the variable of students’ satisfaction on blended learning. Table 13 is a gender difference analysis, which can be shown as follows:

Table 13 Gender difference analysis
Full size table

4.2.2 Analysis of field differences

Students are divided into two fields, namely science and non-science. The differences between attitude, motivation, and students’ satisfaction with blended learning implementation based on the field are analyzed subsequently. P value indicates p < 0.05, which is analyzed by using an Independent Sample T-test. Thus, attitude, motivation and students’ satisfaction with blended learning implementation based on the departments imply significant differences. Science students have higher attitudes, motivation, and satisfaction than non-science students. The different analysis results can be seen in Table 14.

Table 14 Field-based Analysis Result
Full size table

5 Discussion

5.1 Exploration of effect between the students’ attitude, motivation, and learning satisfaction variables in the blended learning implementation

The present study, in this paper, analyzes the factors influencing learning satisfaction in blended learning. Attitude and motivation are the variables that were examined to see their effects on students’ learning satisfaction. In addition, this paper describes the differences between attitude, motivation, and learning satisfaction of students based on gender and scientific field. To see the effects of attitude, motivation, and learning satisfaction on blended learning, it uses SEM – PLS (Structural Equation Modeling – Partial Least Square). Based on the result of the measurement model assessment in this study, it can be concluded that every construct was categorized as reliable and valid so it could be applied analysis model. We discussed findings that are related to the study. The implementation of blended learning could be successful by combining various teaching methods, either online or offline, and also ICT-integrated learning activities. To ensure the optimal blended learning implementation, high integrity and dedication of students and teachers were required. The research findings in this study found that students’ attitudes influenced motivation in the blended learning implementation. Positive attitude on blended learning implementation influenced students’ learning motivation. It is also stated by Kaharuddin et al (2020), attitude is a reaction, response, or a person’s circumstance or feeling that tends to relatively be unchangeable into happy, unhappy, or neutral. The positive response to learning could increase motivation. Otherwise, a negative attitude could produce boredom that subsequently results in a lack of learning motivation. This finding, practically, could provide insight for teachers to be able to build students’ positive attitudes by communicating and teaching them online or offline in the blended learning implementation.

The present study reveals that motivation influenced directly on the students’ learning satisfaction during the blended learning implementation. The higher the students’ learning motivation was, the higher the students’ learning satisfaction was. It can be stated that high learning motivation could influence students’ interest and involvement in the learning process so they might feel more satisfied with the learning experience they obtained. Students encounter challenges during the implementation of blended learning (item M2). They did tests and tasks by finding additional references by using an online platform (item M4). Learning implementation that was not limited by time and space could make students enjoy to understand the learning material (item M1). As such, it could motivate students to learn. At the end, they would be happy with the learning experience in the blended learning class (item S1), which means that the learning satisfaction was achieved. The research by Martín-Blas and Serrano-Fernández, (2009) and Şahin Kızıl, (2014) suggests that lessons delivered in an easily accessible format can increase student learning satisfaction. Therefore, the teacher needs to notice the factor of students’ learning motivation in terms of increasing learning satisfaction in the blended learning implementation. Learning motivation in a blended learning context is able to identify factors influencing students to learn. By understanding these factors, educational institutions can develop strategies to increase students’ motivation, such as designing engaging learning material, providing constructive feedback, or providing social support to students.

The research findings in this study show that attitudes have a direct influence on students' learning satisfaction aligns with different studies (Alwerthan, 2024; Audet et al., 2021). Learning satisfaction is a positive value on the whole students’ learning experience that was measured after the learning activity. Students could express their ideas on how to solve the problem through blended learning implementation. By understanding students’ attitudes in blended learning, educational institutions and the government can identify the advantages and disadvantages of this approach. With this information, learning programs could be designed better, educational quality could be ameliorated overall, and more positive learning experiences could be provided for students.

The main research finding was under attitudes through motivation to achieve learning satisfaction have a greater coefficient value in Fig. 2 compared to attitudes that directly influence satisfaction. Motivation is intervening between attitude and satisfaction in implementing blended learning. Teachers must be able to make learning interesting, relevant, and interactive. The teacher must be able to competent and master technology as the basis to create a better teaching and learning process for students. Thus, students have a positive attitude, which is the key to fostering student motivation to learn and achieving learning satisfaction. The research findings in this paper also help to understand the factors that contribute to students’ learning satisfaction in blended learning. This information is used to ameliorate learning design, enhance interaction between students and teachers, and provide better support services to students, which in turn is able to increase student satisfaction levels. This research finding, thereby, could help accelerate change in education towards more innovative and responsive approaches to the needs of students and society at large.

5.2 Difference based on gender and scientific field

Furthermore, the study in this paper examines the differences between attitude, motivation, and satisfaction based on gender and students’ departments/fields in blended learning. There was no difference in satisfaction in learning on a blended learning application based on gender classification. Either male or female students have similar satisfaction with blended learning implementation. It is similar line with the findings conducted by Aditya et al (2019), Kobicheva et al (2022), and Naaj (2012).

Different analysis on the variable of attitude and motivation indicates that there were significant differences in attitudes and motivation in blended learning based on gender. Male students have similar attitudes and motivation and are higher than female students. It is similar line with the findings conducted by Huang, (2021). Male students were skilled in ICT mastery in blended learning and were challenged to follow the blended learning process.

Another finding in this study came from the different variables of attitude, motivation, and satisfaction based on the students’ fields. The finding of the study in this paper presents that science students have higher attitudes, motivation, and satisfaction than non-science students on blended learning implementation. Science students obtained satisfaction in learning through multimedia that was prepared by the teacher. Subsequently, the students obtained additional opportunities to find out problem solutions in learning through the implementation of blended learning. They could follow the learning process and activity anytime and anywhere. It is reinforced by Akti Aslan,(2022) in his study, he emphasized that blended learning influences students’ perspectives and skills in solving science learning problems. Kintu et al., (2017) also emphasized that student characteristics and background are predictors of achieving learning satisfaction in the implementation of blended learning. So that practically, the results of this research can provide information to teachers so they can adjust students' characteristics and backgrounds in order to achieve learning satisfaction in the blended learning system.

6 Conclusions

Blended learning plays a vital role as blended learning system provides flexibility, effectiveness, and efficiency in the learning process. As such, the present study provides an insight regarding the structural model related to the achievement of learning satisfaction through attitude and motivation. The structural model in this study is suitable and able to use. It is supported by the validity and reliability values that are excellent. In addition, there is no collinearity between the research aspects when evaluating structural models.

The present study obtains structural model and reveals that learning satisfaction is influenced by attitude and motivation. Yet, the main research finding is that to achieve learning satisfaction, students must have attitude on blended learning implementation wherein motivation as intervening. It can be concluded that a motivation as intervening between attitude and satisfaction is needed to increase students’ satisfaction in blended learning. The effect of motivation as intervening on the model provides higher influence on the satisfaction in learning than attitude directly on the satisfaction in learning. Therefore, teachers must be able to create a learning environment that encourages positive attitudes in students. This positive attitude will then increase learning motivation and ultimately achieve satisfaction and success in implementing blended learning.

The study in this paper provides empirical data evidence of attitude, motivation, and students’ satisfaction in blended learning. Based on the data analysis, we found that male students have higher attitude and motivation than female students. Male students are more challenged in blended learning implementation. They are also skilled in using ICT in blended learning implementation. Yet, the satisfaction levels between male and female students on the blended learning implementation are not different. It can be concluded that all students are able to achieve a satisfactory level in blended learning. Hence, it is recommended for teachers to be innovative constantly in the implementation of blended learning and certainly involve technology within.

In addition, the finding of this study indicates that science students have higher attitudes, motivation, and satisfaction in learning than non-science students. They also have additional opportunities to find problem solutions in learning through the implementation of blended learning. Moreover, multimedia in blended learning is one of the factors that increase science students’ learning satisfaction. For those reasons, this study is one of the overviews for science teachers to be able to apply blended learning by involving technology in teaching the materials, either in the form of multimedia for learning science concepts or laboratory activities into virtual lab activities. Also, it certainly becomes a challenge for teachers to be able to master technology in order to implement blended learning.

In conclusion, it should be underlined that the present study can be beneficial for teachers and stakeholders to obtain better insight into the blended learning implementation. The research finding indicates that student learning satisfaction will be maximized through attitudes and motivation as intervening. This means that teachers must be able to carry out student simulations to be able to participate in blended learning so that a positive attitude is developed, then motivation increases, and ultimately maximum learning satisfaction is achieved. Based on these findings, teachers must apply innovative and creative teaching techniques that are integrated with technology in order to achieve learning satisfaction through attitudes and motivation in implementing blended learning.

7 Limitations and further research

In the present study, there are a number of limitations that can be highlighted. This study, firstly, only examines two factors that influence learning satisfaction in the blended learning implementation, namely attitude and motivation. It is recommended for further research to add explanations of other factors that influence learning satisfaction in blended learning. It enables us to explore students’ subjective perceptions related to learning satisfaction in blended learning. Secondly, the sample size is limited in Riau province Indonesia, so it has the potential to restrict the generalizability of findings. Further research can establish this limitation by collaborating nationally or internationally to increase the generalizability level of the research findings. Moreover, this study is only conducted at a certain level, namely senior high school, will be beneficial if further research examines the elementary school level and or junior high school, so it can provide information for teachers and stakeholders related to the factors influencing the achievement of students’ learning satisfaction in the blended learning implementation. The present study, thirdly, only analyzes the difference of variables based on gender and the scientific field. It is more beneficial if further research compares with other students’ characteristics (for example, in or out of town, private or public school). As such, the finding can provide a description of the planned policy government on school preparedness to implement blended learning. This study, fourthly, only uses quantitative data so there is no open explanation from students for the answers to the questionnaire that had been given. It is recommended for further research to conduct qualitative research so that the findings can provide a more in-depth overview of the learning satisfaction model in the blended learning implementation. It can enable us to explore how students’ learning satisfaction can be achieved through students’ subjective perceptions.