**Hypothesis 2 (H2).** *Attitude has a significant contribution to farmers' work performance towards fertilizer application in rice.*

Perceived ease of use (PEU) is one of the most important factors affecting technology use intention and a proxy to actual user behavior [46,47]. Farmers' decision to adopt technologies is dependent on how they perceive those technologies. Reimer et al. [48] informed that farmers' intention to perform certain practice in the United States Midwest region was found to be impacted by their subjective evaluation about the complexity of those practices. Bahramzadeh and Shokati [49] concluded that perceived ease of use is the most powerful factor in behavioral intention to perform a certain technology. Hence, it is clear that ease of use of technology is an important predictor of actual use of technology. Consistent with the previous discussion the following hypothesis has been formed:

**Hypothesis 3 (H3).** *Ease of use of technology has a significant contribution to farmers' work performance towards fertilizer application in rice.*

Motivation is a drive that stimulates people into action, direction to behavior and thus, their productivity [50]. An enormous impact of workers' motivation is visible on productivity and performance [51]. Previous studies provide evidence that motivation positively affects individual performance [52]. Hence, it is clear that workers' high motivation at work plays an important role in their satisfaction which ultimately reflects their higher work performance. In accordance with the preceding discussion, the following hypothesis has been developed:

**Hypothesis 4 (H4).** *Motivation has a significant contribution to farmers' work performance towards fertilizer application in rice.*

#### **3. Methodology**

The quantitative approaches seem to be the best when the focus of the study is to identify the factors that determine a certain behavior [53]. Hence, the researchers employed a quantitative approach to administering this study. A cross-sectional survey method was to collect data that helped the researchers to collects a larger set of data in relatively a shorter period of time [54,55]. In order to collect relevant data from a pre-determined sample, a structured questionnaire was carefully prepared, including open and closed form questions.

#### *3.1. Location, Population, and Sample*

A purposive and multistage random sampling technique was adopted to locate and select the respondents for the survey. This sampling technique was found to be successful in several cases [56,57]. This study chose a multistage sampling technique to minimize random errors and sample bias [57].

Gaibandha is considered one of the important districts in Bangladesh for rice production [58]. So, Gaibandha district was purposively selected as the study area. Then, considering the research cost, size of the area covered, time, human resources, accessibility, and availability of transportation, three Upazilas under the Gaibandha district, namely Gobindaganj, Palashbari, and Sadullapur, and seven villages from each Upazila were selected purposively. There were 3762 rice farmers identified in the 21 villages from three upazilas. Out of which, 355 farmers were randomly selected using Morgan recommendations [59]. A proportionate random sampling technique was used to determine the number of respondents from each selected village using the following formula:

$$\mathbf{n}\_1/\mathbf{N} \times \mathbf{n}\_2 = \mathbf{s}$$

where, N = selected total population of the study; n1 = proportion population in respected village; n2 = determined the total sample size of the study; and, s = sample size from respected village [60].

#### *3.2. Measurement of Variables*

In this study, all the quantitative data were coded into a numerical value. The suitable scoring procedures were applied to convert data to make them easier. Farmers' work performance towards fertilizer application in rice was the dependent variable while farmers' knowledge, attitude, ease of use of technology and motivation were the independent variables in the current study. All the independent variables were measured by using five-point Likert scale. Though the Likert scale is ordinal; however, Likert with five or more categories can often be used as continuous without any harm to the analysis that has been planned to apply [61,62]. In previous studies, scholars have treated their Likert scale questionnaire as interval scale [15,17,22]; hence researchers also follow the same in this study.

Fourteen statement items (I am able to apply recommended dose of all fertilizers to achieve targeted production, I am able to apply recommended dose of urea to increase plant growth and number of panicle in rice, I am able to apply recommended dose of Triple

Super Phosphate (TSP) fertilizer efficiently to improve yield of rice, I am able to apply recommended dose of Muriate of Potash (MP) efficiently to increase yield of rice, I am able to apply recommended dose of Gypsum (Sulphur fertilizer) efficiently to increase yield of rice, I am able to apply recommended dose of Zinc fertilizer (ZnSO4) efficiently to increase yield of rice, I am able to apply urea in three equal splits to rice, I am able to apply first split of urea to rice as basal method, I am able to apply second and third split of urea to rice as topdressing method, I am able to apply Triple Super Phosphate (TSP), Muriate of Potash (MP), Gypsum (Sulphur fertilizer) and Zinc fertilizer (ZnSO4) as basal method to rice, I am able to increase rice yield by improving the timing of fertilizer application, I am able to apply first split of urea final land preparation, I am able to apply second and third split of urea early tillering stage and just before panicle initiation stage of rice, I am able to apply Triple Super Phosphate (TSP), Muriate of Potash (MP) Gypsum (Sulphur fertilizer) and Zinc fertilizer (ZnSO4) during final land preparation) were adapted from Demba [17] to measure the level of farmers' work performance. Farmers were asked to report their views with corresponding statements based on five-point Likert scale and specified five possible responses range from 1 to 5 (1 = Strongly Disagree, up to 5 = Strongly Agree) [15,63].

For the measurement of knowledge, respondents were requested to specify their opinion against 10 statement items (I know that first split of urea should be applied as basal dose and rest of two split as topdressing, I know that the recommended doses of fertilizer are important to optimal rice yield, I understand that using appropriate method of fertilization application is important to extend rice yield, I understand that rice plants require urea at the early and mid-tillering stage to maximize rice yield, I know that urea fertilizer increase plant growth of rice, I know that Triple Super Phosphate (TSP) fertilizer is useful for flowering and panicle initiation in rice, I know that Muriate of Potash (MP) fertilizer is responsible for grain size and weight of rice, I understand that urea should be applied in three equal splits in rice field for higher yield and I know that Triple Super Phosphate (TSP) and Muriate of Potash (MP) should be applied during final land preparation) that were adapted from Ntawuruhunga [64]. A five-point Likert scale was also used to specify respondents' responses range from 1 to 5 (1 = Very Low, to 5 = Very high).

Attitude also measured by adapting 11 statements items (I know the recommended dose of fertilizer and follow it in rice farming, I will get lower yield due to fail applying recommended rate of urea, Triple Super Phosphate (TSP) and Muriate of Potash (MP) in rice, I know the timing of fertilizer applications can increase the yield of rice and follow it in rice farming, I think fertilizer application method is important yield and follow appropriate method in rice, I know urea should apply in three equal splits to rice and follow it, I think timing of urea application is difficult for me to apply in rice, I think I can reduce fertilizer cost by improving the timing of urea application in rice field, I know urea application at once in rice field just before transplanting is easier and follow it in rice, I know organic manure (cow dung) is less important for higher yield and do not use it in rice, It is good to apply fertilizers based on own experienced rather than external advice and I know excessive use of fertilizer is bad for rice production and follow recommended doses of fertilizer) from Ghosh and Hasan [48]. Respondents were asked to respond based on fivepoint Likert scale ranging from 1 to 5, where 1 indicates strongly disagree and 5 indicates strongly agree.

In case of ease of use of technology, 10 statements items (Use of recommended dose of urea for rice, Use of recommended dose of Triple Super Phosphate (TSP) for rice, Use of recommended dose of Muriate of Potash (MP) for rice, Use of recommended dose of Gypsum (Sulphur fertilizer) for rice, Use of recommended dose of Zinc fertilizer (ZnSO4) for Boro rice, Use of urea in three equal splits in rice field, Use of the first split of urea as basal after seedling establishment of rice, Use of the second split of urea at early tillering stage of rice, Use of the third split of urea at 5–7 days before panicle initiation of rice, and Use of all Triple Super Phosphate (TSP), Muriate of Potash (MoP), Gypsum (Sulphur fertilizer) and Zinc fertilizer as basal during final land preparation) were adapted from

Adrian [65]. Five-point Likert scale ranging from very difficult (1) to very easy (5) was used to measure farmers' ease of use of technology.

Finally, respondents were requested to specify their opinion against seven statements items (I apply fertilizer in my rice field as it gives higher yield, I use fertilizer in my rice field as it is easy to apply, Using fertilizer in my rice field gives me higher status in farming community, I apply fertilizer in my rice field because it's readily available, I use fertilizer in my rice field as I have received training on fertilizer application, I apply fertilizer in rice cultivation because my peers think I should use it, and I use fertilizer in my rice field as I have received sufficient extension support) in order to measure their motivation using a five-point scale ranging from "strongly disagree" (1) to "strongly agree" (5). Items of motivation were mostly adapted from Ryan et al. [66].

#### *3.3. Validity and Reliability Analysis*

In this study, the researcher adopted construct validity by measuring the content validation of the instruments. Content validity can be measured by seeking experts' opinions from the respected field of study to conform to the concept and measurements were clear and represented the concerned subject matter. In this procedure, experts' opinions were sought for all the items in the questionnaire and then validated by the supervisory committee. The questionnaire was finalized and sent to 35 non-sampled rice farmers who were randomly selected for pre-testing. Cronbach's Alpha test is used to measure the reliability of all the items under each construct in the questionnaire. Cronbach's Alpha of work performance, knowledge, attitude, ease of use of technology, and motivation was 0.862, 0.830, 0.770, 0.785, and 0.770, respectively. The value of Cronbach's alpha coefficient should be equal to or greater than 0.7 which means that the data is reliable and the internal consistency of the items in the scale is satisfactory [67]. Hence, the Cronbach's Alpha values of the items were found reliable.

### *3.4. Data Collection and Statistical Analysis*

Data were collected by the first author of the paper in a face-to-face situation, given respondents' level of literacy and other factors like their preparedness for this type of study. Data were collected from March to May 2018. The collected data were coded, entered, and analyzed using SPSS v23 according to the objectives and hypothesis of the study. Multiple linear regression with 0.05 and 0.01 levels of probabilities were used to explore the contribution of the selected factors on farmers' work performance and determine the highest contributing factors on farmers' work performance towards fertilizer application in rice. In the current study, multiple regression works with the following formula:

$$\mathbf{Y} = \mathbf{b}\_0 + \mathbf{b}\_1 \left(\mathbf{x}\_1\right) + \mathbf{b}\_2 \left(\mathbf{x}\_2\right) + \dots + \mathbf{c}\_i + \mathbf{b}\_k \left(\mathbf{x}\_k\right) + \varepsilon\_i \tag{1}$$

Here, Y is the dependent variable (farmers' work performance towards fertilizer application). X1, X2 ... Xk indicates the independent variables (knowledge, attitude, ease of use of technology, and motivation of farmers); b1, b2 ... bk are the regression coefficients of independent variables and b0 constant. Besides, ε<sup>i</sup> indicates the error term.

#### **4. Results and Discussion**

This section is organized into two sub-sections. The first sub-section deals with the findings of the study and the second sub-section present the test of hypotheses. While the third and last sub-sections discusses the findings related to the contribution of independent variables (i.e., knowledge, attitude, ease of use of technology, and motivation of farmers) on the dependent variable (farmers' work performance towards fertilizer application in rice).

Multiple linear regression analysis was executed to explore the contribution of selected factors that influence farmers' work performance towards fertilizer application in rice and finds out the factor that has the highest contribution to farmers' work performance towards fertilizer application in rice. There were four independent variables that influence farmers' work performance, were selected as predictors of the mentioned dependent variable. These

four independent variables—knowledge (X1), attitude (X2), ease of use of technology (X3), and motivation (X4) of farmers are expected to formulate a multiple linear regression model that could be explained the variation of work performance among farmers. Thus, the multiple linear regression equation of this study has been written as follows:

$$\mathbf{Y} = \mathbf{b}\_0 + \mathbf{b}\_1 \ (\boldsymbol{\aleph}\_1) + \mathbf{b}\_2 \ (\boldsymbol{\aleph}\_2) + \mathbf{b}\_3 \ (\boldsymbol{\aleph}\_3) + \mathbf{b}\_4 \ (\boldsymbol{\aleph}\_4) + \boldsymbol{\varepsilon}\_i \tag{2}$$

where, Y = Work performance of farmers; b0 = Constant; b1–4 = Regression coefficient; X1 = Knowledge; X2 = Attitude; X3 = Ease of use of technology; X4 = Motivation and ε<sup>i</sup> = Error term.

#### *4.1. Contribution of Selected Factors on Farmers' Work Performance*

Table 2 represents the model summary of multiple linear regression analysis. It showed the first statistics R known as the multiple correlation coefficients between all predictor variables and farmers' work performance and obtained 0.749. The next statistic is R2, the coefficient of determination that indicates the percentage of the total variance in a dependent variable explained by all the predictor variables. The value of R2 is 0.561 indicated that all the independent variables were simultaneously explained 56.1% of the total variance of the dependent variable. The next statistic is Adjusted R2, a modified version of R2 that calculates R<sup>2</sup> using only those independent variables, which was significant for predicting the dependent variable. Here, the adjusted R2 value (0.556) indicated that the significant predictor variables were simultaneously explained 55.6% of the total variance of farmers' work performance. In other words, the rest of 44.6% of the total variation of farmers' work performance has not been explained in the current study.

#### **Table 2.** Table of multiple linear regression model summary.


Significant: \* *p <* 0.05.

In addition, the value of the F-test was 111.783 which is significant at *p* < 0.05. It implies that the multiple linear regression model has a significant influence over the dependent variable of the study. In other words, it could be said that the combination of independent variables as a predictor has a significant contribution to farmers' work performance. Thus, the regression model was good or fit to predict the contributions of independent variables.

#### *4.2. Highest Contributing Factors on Farmers' Work Performance towards Fertilizer Application in Rice*

Table 3 recognized the independent variables that have a significant value of *p* < 0.05. It implies those respected variables have a statistically significant and distinctive contribution to predict the dependent variable of the study. However, the variables, which do not have a significant value of *p* < 0.05, are not considered a significant predictor of the mentioned dependent variable [68].

Table 3 shows the unstandardized regression coefficient (b) and standardized regression coefficients (*β*) taken to examine the contributions of selected independent variables on farmers' work performance. The strength of the contribution of the respected independent variables was compared to each other based on their standardized coefficient (*β*). Standardize coefficient (*β*) was estimated in units of standard deviation and not in a unit of the respected independent variables. The standardized coefficient (*β*) was calculated by multiplying the unstandardized coefficient (b) with the standard deviation of the independent and dependent variables. Thus, standardized coefficient (*β*) becomes normalized as a unit-less coefficient, also known as z-score. According to Table 3, the motivation of farmers had the largest standardized coefficient (*β*) value of 0.478. It implies that the motivation of farmers showed the highest contribution to predict the work performance of

farmers towards fertilizer application. The second highest *β* value was found for knowledge (0.265), followed by ease of use of technology (0.122), while attitude (0.073) had an insignificant contribution.

**Table 3.** Coefficients of multiple linear regression for farmers' work performance towards fertilizer application in rice.


Significant: \* *p <* 0.05.

The values of the unstandardized coefficients values for knowledge, attitude, ease of use of technology, and motivation of farmers were 0.281, 0.084, 0.146, and 0.544, respectively (Table 2). The unstandardized coefficients (b) value of the respected variables indicated the change amount in the dependent variable (Y) in accordance with the change of one unit of an independent variable (X). Thus, based on the estimated unstandardized coefficients (b), the multiple linear regression model has been obtained as follows:

$$\mathcal{Y} = -0.088 + 0.281 \left( \mathcal{X}\_1 \right) + 0.084 \left( \mathcal{X}\_2 \right) + 0.146 \left( \mathcal{X}\_3 \right) + 0.544 \left( \mathcal{X}\_4 \right) + \varepsilon\_i$$

Table 2 also revealed that knowledge (*t* = 6.315, *p* = 0.000), ease of use of technology (*t* = 2.905, *p* = 0.004) and motivation (*t* = 10.760, *p* = 0.000) significantly provide explanation of farmers' work performance towards fertilizer application in rice. In contrast, the contribution of attitude is insignificant to predict farmers' work performance as significant value (*p*) of attitude (*t* = 1.943, *p* = 0.053) is not <0.05.

#### *4.3. Test of Hypotheses*

**Hypothesis 1 (H1).** *Knowledge has a significant contribution to farmers' work performance towards fertilizer application in rice.*

According to the multiple linear regression analysis, the standardized coefficient (*β*) value for farmers' knowledge was 0.265 with a *t* value of 6.315 which was significant at *p* < 0.05. Therefore, the hypothesis (H1) of the study has been failed to reject and the null hypothesis (H0) has been rejected (Table 4).

**Table 4.** Summary of testifying the research hypotheses of the study.


**Hypothesis 2 (H2).** *Attitude has a significant contribution to farmers' work performance towards fertilizer application in rice.*

The value of standardized coefficient (*β*) for farmers' attitude towards fertilizer application was 0.073 with a *t* value of 1.943 which was significant at *p* < 0.05 (*p* = 0.053). Therefore, the hypothesis (H2) of the study has been rejected (Table 4) and the null hypothesis (H0) has been failed to reject.

**Hypothesis 3 (H3).** *Ease of use of technology has a significant contribution to farmers' work performance towards fertilizer application on rice.*

The value of standardized coefficient (*β*) for ease of use of technology was 0.122 with a *t* value of 2.905 was significant at *p* < 0.05 (*p* = 0.004). Therefore, the hypothesis (H3) of the study has been failed to reject and the null hypothesis (H0) has been rejected (Table 4).

**Hypothesis 4 (H4).** *Motivation has a significant contribution to farmers' work performance towards fertilizer application in rice.*

The value of the *β* for farmers' motivation towards fertilizer application in rice was 0.478 with a *t* value of 10.760 was significant at *p* < 0.05 (*p* = 0.000). Therefore, the hypothesis (H4) of the study has been failed to reject and the null hypothesis (H0) has been rejected (Table 4).

#### *4.4. Discussion*

According to the regression model, R<sup>2</sup> (coefficient of determination) and adjusted R<sup>2</sup> are 0.561 and 0.556 respectively. Moreover, the *F*-value 111.783 was significant at *p* < 0.05. According to these findings, the regression model is a good fit. That means, the regression model's estimated result is satisfactory as 56.1% of the total variance of farmers' work performance has been explained by motivation, knowledge, ease of use of technology, and attitude of farmers simultaneously. Hence, it can be assumed that these independent variables have adequate power for the explanation. The adjusted R2 (0.556) value also interpreted that only significant predictor variables have explained 55.6% of the total variance of farmers' work performance: motivation, knowledge, and ease of use of technology simultaneously. Therefore, it can be assumed that the regression model of the current study has explained a significant percentage of total variation that occurs in the work performance of the farmers towards fertilizer application in rice.

This finding is in line with Demba [17] executed a study on personality traits and work performance for paddy farmers and stated that coefficients for farmers' work performance model explained 59.5% of total variation on farmers' work performance in rice cultivation in the Gambia. Bagum et al. [15] also revealed that the regression model explained 49.2% of the total variance of farmers' performance regarding fertilizer application in Bangladesh. A similar trend is also found from the study conducted by Shah [63] and stated that the regression model explained 44.8% of the total variance of farmers' work performance in rice cultivation in Malaysia.

In this study, motivation was one of the significant predictors identified as the highest contributing factors to explain farmers' work performance towards fertilizer application. The value of *β*-coefficient for motivation suggests that with one standard deviation change in farmers' motivation, their work performance will be increased by 0.478 standard deviation. It indicates that the motivation of farmers mainly regulates their work performance towards fertilizer application. Motivation is the most important reason that influences farmers to practice a particular agricultural technology to achieve higher productivity [69]. Moreover, factors like education, experience, extension contract, and training help motivate farmers to improve work performance and increase output [70]. Since farmers are important for agricultural production, it is crucial to continually keep farmers' stimulation level up to perform the farming activity, especially fertilizer application.

Prior literature also mentioned that motivation is a significant predictor and the highest contributor to respondents' performance [71]. Ngima and Kyongo [72] also noticed a similar finding that motivation had a strong statistically significant influence on individuals' performance.

Knowledge the second-highest contributing factor was predicting farmers' work performance towards fertilizer application in rice. It indicates that the enhancement of knowledge can guide farmers to realize the appropriateness of using certain technology. Therefore, agriculture knowledge is vital for farmers to improve their performance to apply essential technologies and increase their productivity levels.

With references to knowledge, Bagum et al. [15] identified farmers' knowledge as an important predictor that significantly contributed to farmers' performance. Moreover, Campbell and Wiernik [28] argued that role-specific knowledge is one of the leading determinants of respondents' performance.

Farmers' attitude displayed an insignificant contribution to predicting farmers' work performance (*p* = 0.053) However, despite having found insignificant contribution, one should not ignore the importance of a favorable attitude in determining farmers' adoption decision of any farming practice [73]. Other studies provided evidence that attitude has a significant contribution to the performance of respondents [43,71]. However, sometimes, individuals' positive attitudes are not enough for performing a given behavior due to their different socio-economic circumstances. As per the researchers' observation, differences might be existed among the farmers according to knowledge, ability, attitudes, and these differences can influence their behavioral decision [74,75]. Such inconsistency might be prevailing among the respondent farmers in the study area. Thus, a farmer might possess a favorable attitude towards fertilizer application for higher yield yet not apply fertilizer at the recommended rate due to other factors like the high input cost or unavailability of fertilizers.

#### **5. Conclusions**

The overall findings of the multiple regression analysis explored the combination of significant predictors such as viz. knowledge, ease of use of technology, and motivation of farmers explained 55.6% of the total variance of farmers' work performance towards fertilizer application in rice. The rest of the variance of farmers' work performance may be explained by other factors that were not being considered in the current study. However, the estimated multiple linear regression model was good or fit to predict the contributions of independent variables. Therefore, the study concluded that the estimated regression model of farmers' work performance is suitable to predict the contributions of selected factors like knowledge, ease of use of technology and motivation of farmers in the current study.

Additionally, the motivation of farmers was recognized as the highest contributing factor followed by knowledge. Hence, it can be suggested that greater emphasis should be given to farmers' motivation and knowledge level to solve their problems and provide maximum effort for higher work performance.

Theoretically, this study will enhance the opportunity to execute new studies in the field of performance through providing critical literature support based on the significant contribution of knowledge, ease of use of technology, and motivation of farmers to their work performance. Besides, the current study's findings are significant to farmers as it focuses on the present level of farmers' work performance for establishing an effective working environment for them to ensure higher performance in applying agricultural practices and getting higher production of rice. Moreover, study findings will provide support as a basis of the national and local motivational campaign including training and technical support ought to be provided by the Department of Agricultural Extension (DAE) of Bangladesh, other GOs, and NGOs extension service providers to equip farmers with essential knowledge, high motivational level, and skill for strengthening their work performance.

Apart from this, the present study highlights only four variables: knowledge, attitude, ease of use of technology, and motivation of farmers that leads to better work performance of farmers towards fertilizer application in rice. Therefore, it is suggested that further research should be undertaken with other potential variables to explore the work performance of farmers. Moreover, other factors of rice cultivation such as irrigation, weed

management, pest and disease management, and intercultural operations can be taken under consideration for future research on farmers' work performance.

**Author Contributions:** Conceptualization, T.B., M.K.U., S.H., N.H.K., M.Z.R. and A.N.A.H.; Methodology, T.B., M.K.U., S.H. and N.H.K.; Formal Analysis, T.B. and S.H.; Writing—Original Draft Preparation, T.B. and M.K.U.; Writing—Review and Editing, M.K.U., T.B., S.H., N.H.K., M.Z.R. and A.N.A.H.; Supervision, M.K.U., S.H., N.H.K. and M.Z.R. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research paper was supported by Universiti Putra Malaysia and National Agricultural Technology Program (NATP): Phase-II Project (Vote No. 6282514), Bangladesh Agricultural Research Council.

**Institutional Review Board Statement:** Ethical review and approval were waived for this study due to collecting basic information from the respondents related to their farm practices. Disclose of these information were not harmful to any human. However, the authors were very careful to use these information just only for the study purpose.

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** The authors would like to express their sincere gratitude to Organization for Women in Science for the Developing World (OWSD) and Swedish International Cooperation Agency (SIDA) for their support through the fellowship in this research project. The authors would also acknowledge the National Agricultural Technology Program (NATP): Phase-II Project, Bangladesh Agricultural Research Council for the financial support, and the Universiti Putra Malaysia, Malaysia for the research facilities.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com

*Sustainability* Editorial Office E-mail: sustainability@mdpi.com www.mdpi.com/journal/sustainability

MDPI St. Alban-Anlage 66 4052 Basel Switzerland

Tel: +41 61 683 77 34 Fax: +41 61 302 89 18

www.mdpi.com

ISBN 978-3-0365-3448-0