E-training based on Determination of Education and Training Models of Early Childhood Teachers Education Programs

Articles Info ________________ History Articles: Received 20 Juny 2018 Approved 15 July 2018 Published 30 August 2018 ________________


INTRODUCTION
The globalization era which is shown towards digital technology requires the availability of competent human resources in order to compensate.The development of digital technology may result the formation of human resources that are primitive and isolated if be avoided.The fulfillment of competent human resources to follow development of digital technology can be done through education, so it needs a good attention from a country.The sophistication of the technology utilized by the actors of education must be supported by quality of human resources who are capable in carrying out the work.
According to article 1 section 9 of the Constitution No 13 of 2013 about employment, training is the whole activities to provide, obtain, increase and develop work competence, productivity, discipline, attitude and work ethic on the level of skills and specific skills in accordance with the level and qualifications of the position and the job (Kemenaker, 2013).Meanwhile, according to Kamarainen (2002: 23) education and training are the enhancement activities of human resources in meeting the needs of future skills and requests of new qualifications.It is also said by Neo (2010: 5) that Education and training are ways to create intellectual ability that includes basic skills, advanced skills, and self-motivated creativity.Based on the understanding of education and training above, it is concluded that education and training are activities of providing the stimulation to do a better job in accordance with the company objectives to improve the skills of employees or workers and as a bridge to the development of knowledge and experience.
According to Widiasih and Suminar (2015) stated that the object of training is to achieve a change in the behavior of thoses trained.The purpose of Education and training is to achieve a change in the behavior of those who are trained.Education and training not only improve knowledge and skills, but also develop talents.Ratnasari (2013: 43) stated that the factors of training that consist of training materials the ability of trainers, training methods, and training participants partially have significant influence to the performance of the employees.So, the existence of education and training in an institution or agency will have an impact on the achievement that has been set and lead to the achievement of both the institution and personally.
Individual differences are significant determinant of interaction, learning and adaptation (Ojo, Raman, & Chong, 2016: 45-63) According to Tai (2006: 51-65) found that participants have more training motivation when they were assigned to attend training programs by the management than they make their own choices freely.
The Council of The European Union asserted that some changes show that training in the institutional and business context changes, producing new competency requirements for trainers both in terms of basic qualifications and ongoing professional development.
Meanwhile, according to Mamaqi, Miguel, & Olave (2011: 174-187) confirmed that teachers and trainers deal with the basic skills and competence such as curriculum development, tutorial support, skills assessment, management of learning environment, and teacher professional development that is sustainable.
According to Amara (2016: 83) Etraining is defined as the process of distance training through the use of the Internet or an Intranet, providing the necessary knowledge to some particular subjects or specialization which is selected, to increase the level of knowledge or to achieve rehabilitation by using computer, voice, video, multimedia, e-book, email, chat and discussion group.Meanwhile, according to Ramayah, Ahmad, & Hong (2012: 125-137) stated that e-training resembles e-learning in many ways especially in terms of delivery methods and technology used, except that it refers to the framework of a shorter learning time which is specifically designed to achieve the learning objectives or specific skills.
Based on the understanding above, it can be concluded that e-training is the utilization of electronic media for education and training activities in order to improve skills and performance of the participants of education and training.

METHODS
This study used a quantitative approach with survey methods.Quantitative approach is done by obtaining empirical data that allows researchers to see the general tendency behind the behavior of a person or group through analyzing the data in the form of numbers.Research with data in the form of numbers or is changed into numbers, so the approach used is a quantitative approach (Sugiyono, 2013: 7).
The population in this study were PAUD teachers who joined e-training-based education and training with a total of 2584 teacher.The sampling technique used in this study is nonprobability sampling with a purposive sampling technique.What is meant by nonprobability sampling is a sampling technique that does not provide equal opportunities/ chances for each element or member of the population to be selected as a sample (Sugiyono, 2013: 86).While purposive sampling is a technique of determining samples with certain considerations (Sugiyono, 2013: 85).Determination of the number of samples by using Isaac and Michel formulaa make it easy to determine the number of samples with an error rate of 1%, 5%, and 10% according to the desired.So, in this study, the numbers of samples based on the formula above were 183 respondents with an error rate of 5%.By the estimation method using Maximum Likehood (MC), 100 samples were needed in minimum and 200 samples were needed in maximum (Ghozali, 2007) In this study, there were variables that will be measured, namely: independent variables (exogenous), intervening variables and dependent variables (endogenous).There are seven independent (exogenous) variables, they are individual differences, job analysis, motivation, active participation, trainer competence, education and training methods, and learning principles.There was one intervening variable, it was the use of e-training and there was one dependent variable (endogenous), namely the effectiveness of education and training.
Data collection techniques used in this research were observation and questionnaires.The scale measurement instrument used in this study was Likert Scale with five alternative answers Likert Scale which was used to measure attitudes, opinions and perceptions of a person or group of people about social phenomena.
Data analysis techniques used in this study was Structural Equation Modeling (SEM).SEM testing in this study was using AMOS 24.0 statistical software.The first step was to examine the measurement model and structural model on exogenous and endogenous variables using confirmatory factor analysis (CFA) analysis to test whether the indicators used were valid and reliable indicators as a measure of latent variables by looking at the factor loading value ≥ 0.5.While to measure the consistency of the measuring instrument, it was done by calculating the construct reliability of ≥ 0.6.For structural models, SEM analysis was done by assessing the matching of the goodness of fit estimation in the research model and hypothesis.

Validity and Realibility Test
Based on the results of analysis, it is obtained the results of validity test from the the following table.Based on the table above, it is known that it is directly obtained through standardized loading factor for error measurement from each indicator is ≥ 0.5.So, it can be concluded that the entire variable can be explained from every its indicator or valid.
The table above also explains that the value of construct reliability is above ≥ 0.60, which means that instruments are reliable, however, the variance extracted value shows the construct= fulfilling the cut-off requirement of a minimum value of 0.50, which means that the indicators used as observed relative variables are able to explain the latent variables that are formed.The Variance Extracted value is ≥ 0.05, it indicates a good convergence (Ghozali, 2007: 135).

Confirmatory Factor Analysis
The test results of confirmatory factor analysis of exogenous variabe show goodness of fit which is presented in the table below.Based on the test results presented in the table above, it can be seen that the construct used in the confirmatory analysis research model has met the criteria of goodness of fit that has been set.According to the probability value table on goodness of fit, it shows a value of 0.054.This value is categorized as marginal, so it can be concluded that the model is accepted and feasible to use.
The requirements that must be fulfilled in SEM modeling are normal multivariate assumptions, the absence of multicollinearity or singularity and no outliers occur in the study.Ghozali (2011: 64) stated that the Maximum Likelihood (ML) estimation technique is more efficient and unbiased if the assumption of multivariate normality is fulfilled, that is by the amount of Critical Ratio (CR)> 2.58) multivariately normal data must be normal univariate.However, on the other hand, if the entire data that is normally univariate, it is not a guarantee that it will be multivariately normal outside the range of ± 2.58.The results of testing with the value of 2,125 shows that the research data used had meet the prerequisite of normality of the data or in other words that the data in this study has been normally distributed.In addition, the value of the determinant of the sample covariance matrix was 1.6441.This value was far from zero, so it can be concluded that there is no problem of multicollinearity and singularity in the data analyzed, the value of the mahalanobis distance is smaller than 72,20 based on the value of Chi Square at degrees of freedom of the number of indicators n=43 at a significance level of 0.001 (Ghozali, 2005: 130).It can be concluded that there were no multivariate outliers in research data.

Structural Equation Modelling Analysis (SEM)
The analysis of Structural Question Modeling (SEM) thoroughly or full model is the phase that is done after doing an analysis of confirmatory factor analysis on the variables exogenous and endogenous.Analysis of the results of data processing of the whole or full SEM model was conducted to test the suitability of the goodness of fit and statistical tests.The following is a presentation of the results of overall analysis or full SEM model in the form of images.The estimation parameter in testing the effect of active participation on the effectiveness of training shows a C.R value of 0.072 with a probability of > 0.05.So, it can be concluded that active participation variables have positive and insignificant effect on the effectiveness of education and training.The results are consistent with the research of Solansky and McIver (2017) which explained that we still have no a complete understanding of how the team actually increases or inhibits the learning outcomes, such as participation in training programs.Due to the tendency to use teams as part of education and development programs continues to increase, the need to fully understand the effectiveness of this strategy is very important.
The estimation parameter in testing the influence of trainers competence on the effectiveness of training shows a C.R value of 2.251 with a probability of <0.05.So, it can be concluded that the competence of trainers variable have a positive and significant effect on the effectiveness of training.The results are consistent with the research of Mamaqi, Miguel, and Olave (2011) which stated that the competence and new skills of trainers on identifying positive have an impact on training success.
The estimation parameter in testing the effect of the training method on the effectiveness of training shows that the C.R value is 1.933 with a probability of <0.05.So, it can be concluded that the training method variables have a positive and significant effect on the effectiveness of training.The results are consistent with training methods research of Mor, Glezer, and Zviran (2011) which explained that his study provides new knowledge on how an organization can effectively utilize training resources to achieve better operational training excellence.
The estimation parameter in testing the effect of learning principles on the effectiveness of training shows that a C.R value is -1.315 with a probability of > 0.05.So, it can be concluded that the principle study variables have a negative effect and insignificant on the effectiveness of education and training.The results are consistent with the research of Edward ( 2007) which stated that it is not surprising that a training to provide effective answers.Participants easily give up to study because they are no longer oriented to a task and obligation that must be fulfilled with too much methods that are not appropriate to the background of participants.

E-training-based Model of Education and Training of early childhood teacher's education programs
The estimation parameter in testing the effect of individual differences on the use of etraining shows a C.R value of 0.688 with a probability value of 0.504, then the probability value is > 0.05.So, it can be concluded that the variables of individual differences indirectly have a positive and insignificant effect on the effectiveness of education and training through the use of e-training.The result are consistent with the research of Robins and Pals (2002) individuals oriented to work goals and making mistakes in training with helpless attributes negatively affecting training.
The estimation parameter in testing the effect of job analysis on the use of e-training shows a C.R value of -0.598 with a probability value of 0.504, so that a probability value is > 0.05.So, it can be concluded that job analysis variables indirectly have a negative and insignificant effect on the effectiveness of education and training through the use of etraining.The results are consistent with the study of Klink (2002) which stated that job analysis does not fully fully affect training even though training is needed according to the field of work.
The estimation parameter in testing the effect of motivation on the use of e-training shows a C.R value of 2.075 with a probability value of 0.038, so that the probability value is <0.05.So, it can be concluded that the motivation variables indirectly have a positive and significant effect on the effectiveness of education and training through the use of etraining.The results are consistent with the research of Renta (2014) which stated that a significant influence between motivation to increase work, motivation to learn, coherence training on the effectiveness of training.
The estimation parameter in testing the effect of active participation on the use of etraining shows a C.R value of 3.091 with a probability value of 0.002, so that the probability value is <0.05.So, it can be concluded that the variable of active participation indirectly has a positive and significant effect on the effectiveness of education and training through the use of etraining.The results are consistent with Weaver's research (2017) the participation of trainees with higher education levels tends to have a positive effect on training.
The estimation parameter in testing the influence of the trainers competence on the use of e-training shows a C.R value of 2.338 with a probability value of 0.019,so that the probability value is <0.05.So, it can be concluded that the variable of trainer competence indirectly has a positive and significant effect on the effectiveness of training through the use of e-training.The results are consistent with the research of Mamaqi, Miguel, and Olave (2011) which stated that the competence and new skills of trainers have an impact on identifying positive impacts on training success.
The estimation parameter in testing the effect of the training method on the use of etraining shows a C.R value of 4.278 with a probability value of 0.001, so that the probability value is <0.05.So, that it can be concluded that the variable training method has a positive and significant effect on the effectiveness of education and training through the use of e-training.The results are consistent with the research of Kaptan-Mor, Glezer, and Zviran (2011) which explained that his study provides new insights on how an organization can effectively utilize training resources to achieve better operational training excellence.
The estimation parameter in testing the influence of the principle of learning on the use of e-training shows a C.R value of 0.194 with a probability value of 0.846, so that the probability value is >0.05.So, it can be concluded that the principle of learning indirectly has a positive and insignificant effect on the effectiveness of education and training through the use of e-training.The results are consistent with the research of Edwards (2007) which stated that it is not surprising that a training to provide effective answers.
Participants easily give up to learn because they are no longer oriented to tasks and obligations that must be fulfilled with too much methods that are not appropriate to the background of participants.
The estimation parameter in testing the effect of e-training on the effectiveness of training shows that a C.R value is 2.477 with a probability of <0.05.So, it can be concluded that e-training variables have a positive and significant effect on the effectiveness of training.The results are consistent with the research of Ramayah, Ahmad.,And Hong, (2012) while the organization is aware of the benefits that e-training brings, namely cost effectiveness, flexibility, convenience, consistency of content throughout the organization, many who do not recognize the critical factors that influence the successful implementation.

E-training-based Determination of Education and Training Model of early childhood education Teachers
Job analysis, trainer competence, and training methods have a direct and positive effect on the effectiveness of early childhood education teachers training with the magnitude of the effect in sequence is 28.9%, 27.5%, and 23.7%.Active motivation and participation directly and positively and insignificantly influence the effectiveness of early childhood education teachers training with the magnitude of the effect in a sequence are 12.7% and 0.8%.Individual differences and earning principles indirectly have no effect on the effectiveness of education and training of early childhood education teachers.
Training methods, active participation, trainer competence, and motivation indirectly have a positive and significant effect on the effectiveness of training through the use of etraining.The magnitude of the influence is 15.6%, 9.9%, 7%, and 6.3%.Individual differences and the principle of learning indirectly have a positive and insignificant effect on the effectiveness of education and training through the use of e-training with the magnitude is 1.9% and 0.6%.Job analysis indirectly does not affect the effectiveness of education and training through the use of etraining.The e-training variable directly has a positive and significant effect on the effectiveness of the training.with a magnitude of influence is 32.6% The magnitude of the total effect of the training method, trainer competence, job analysis, motivation, active participation, individual differences and learning principles in sequence are 39.3%, 34.5%, 27.4%, 19%, 10.7%, -4 5% and -12.2%.

CONCLUSION
The conclusion in this study is that job analysis, trainer competence, and training methods directly have a positive and significant effect on the effectiveness of early childhood education teachers training.

Picture 1 .
The Results of Structural Equation Modelling (SEM) The results of the SEM feasibility test or goodness of fit are shown in the following table.

Tabel 4 .
The Results of Analysis Model of Structural Equation Modelling (SEM) Education and Training Model for PAUD TeachersThe estimation parameter in testing the influence of individual differences on the effectiveness of the training shows a C.R value of -0.667 with a probability of > 0.05.So, it can be concluded that the variable of individual differences have no impact on the effectiveness of education and training.The results are consistent with the research ofRobins and Pals (2002) which stated that individuals oriented to work objectives and making mistakes in training with helpless attributes negatively affecting training.The estimation parameter in testing the effect of job analysis on training effectiveness shows a C.R value of 2.789 with a probability of <0.05.So, it can be concluded that job analysis variables have positive and significant effect on the effectiveness of education and training.The results are consistent with the research of Veen and Versloot (1993) which stated that training must be based on job profiles and even extends to the personal economic sector since it affects on the training.The estimation parameter in testing the effect of motivation on the effectiveness of training shows a C.R value of 1.235 with a probability of > 0.05.So, it can be concluded that motivation variables have positive and insignificant effect on the effectiveness of training.The results are consistent with the research of Tassone, Dik, & van Lingen (2017) which explained that motivation in individual training does not have an additional effect on knowledge and beyond the influence explained by education knowledge and pre-training, although the motivation of training and knowledge of post-training declarations correlates significantly.
Active motivation and participation have a directly positive influence and insignificant on the effectiveness of ECD educator training.Individual differences and learning principles directly have no effect on the effectiveness of education and training of early childhood educators.Training methods, active participation, trainer competence, and motivation indirectly have a positive and significant effect on the effectiveness of training through the use of etraining.Individual differences and the principle of learning indirectly have a positive and insignificant effect on the effectiveness training through the use of e-training.Job analysis indirectly does not affect the effectiveness of training through the use of etraining.The e-training variable directly has a positive and significant effect on the effectiveness of the training.
explained that as an alternative to the traditional training which has been running in advance, the web-based online training model that is cheap gives more open options for the customer to choose the type of training services that suit the needs and abilities of the customer.As an alternative to the traditional training which has been running before, the web-based online training model provides convenience for customers who have limited time and is difficult to reach the location of training to be able to get service training.

Table 1 .
Summary of Validity and Reliability Test using Loading Factor, Contruct Reliability,

Table 2 .
Solimun (2002) Feasibility Test of Exogenous Variable Confirmatory Analysis ModelBased on the test results presented in the table above, it can be seen that the construct used in the confirmatory analysis research model has met the criteria of goodness of fit that has been set.The probability value on goodness of fit shows a value of 0.634.This value has met the feasibility of testing and it is categorized as good although there are some marginal testing values.Referring to the opinion ofSolimun (2002)stated that based on the parsimony principle, if there is one or two fit criteria the model has met, the model is considered fit.Whereas, the results of confirmatory factor analysis of endogenous variables show the goodness of fit which is presented in the table below

.
Tabel 3. The Results of Feasibility Test of Endogenous Variable Confirmatory Analysis Model

Goodness of Fit Index Cut-off Value
The next step is hypothesis testing, this step was done after the SEM testing was done to meet the hypothesis that has been proposed in this study.The test of all hypotheses is done by looking at the Critical Ratio (C.R) value of cashuality relationship from the results of SEM processing as presented in the following table.

Table 5 .
Hypothesis Testing of Each Variable