1. Introduction
The number of educational institutions, companies and other users applying eLearning systems has grown significantly in the last decade, therefore they have become as important means and resources as other informational systems of the institutions [10]. However, there are several conditions and components of successfully using these systems in the educational institutions – for example what kind of system is chosen, how it is implemented and introduced. The operation of the system and the services available for the users (teachers, students) are also important [12].
Probably the most significant question is what teachers and students can profit from the system. What is the advantage of using it? How does it help the process of teaching and learning to become more effective and transparent? Does it support the management of institutional education? If it does so, what extent? Naturally, the organizational and economical aspects of the use of eLearning systems are also important. The application of the eLearning systems is gradually becoming more essential for those institutions, organizations and companies which have distant teaching and trainings and also for the improvement of human resources.
Our research objectives include the quality of eLearning establishment and the support of our institution. Accordingly, tasks were set for the examination of eLearning opportunities for quality improvement in education, improvements in this regard, proposals and recommendations for application development. The course structure recommended for the application of eLearning for institutional quality improvement together with other functions reachable as a module of the system, can ensure an integrated and comprehensive eLearning quality service. The strategy and implementation of quality improvement is only possible by providing qualified human resources. The basic objective of the introduced LMS is to improve the quality of education, which is one way where the students and instructors receive ongoing feedback about their experiences with the system. Corresponding objective is to compile a questionnaire, which was a result of useful information about the students and teachers eLearning system and application views [6].
The aim of the questionnaires were to compare the opinions of the students and the teachers and also to evaluate the Faculty of Economics and Business (FEB) of the University of Debrecen and the Corvinus University of Budapest (CUB) regarding the application of eLearning. The role of the questionnaire for quality development is to give guidance for the FEB in the application of eLearning. ELearning in the CUB is applied under certain organized institutional circumstances. The eLearning application of CUB works with an organization defined extended several faculties of the University, which can be a good example for our faculty.
The following hypotheses can be defined: The quality development of eLearning should be ensured under organized circumstances.
2. Applied Methods
2.1. Questionnaire Survey
Questionnaire survey was carried out at the FEB among of the economic and agricultural engineering and technical management of information technology students. The primary way of getting information (primary), as new, unknown and unpublished, original, primary data acquisition occurred during the survey.
There was an online way of response, which is a quantitative online CAWI (Computer Assisted Web Interviews) survey over the Internet. Rapid market surveys carried out by this method [1]. Our questionnaire was accessible through Limesurvey system (The evaluation of application of eLearning, 2015) (Figure 1), which is a free and open source online survey application written in PHP based on a MySQL database. As a web serverbased software it enables users using a web interface to develop and publish online surveys, collect responses, create statistics, and export the resulting data to other applications [7].
In this Thesis a research survey designed from which the predetermined group of users to get answers to important research questions. A questionnaire to evaluate the MannWhitney test was used. With this method we can determine significance of the averages of the two groups as follows: the two samples are ranked together, regardless of group that is ready to make the ranks. A questionnaire to evaluate the SPSS (Statistical Package for the Social Sciences statistical package for social sciences) program version 22 was used. The application of the questionnaire responses LimeSurvey SPSS syntax and data files were exported to the files that I imported into SPSS.
2.2. The Applied Statistic Methods
We used factor analysis and binominal logistic regression. We examined whether the background variables manipulating the variables are possible to be developed on the basis of the answers given. We used factor analysis to demonstrate this since it contracts the coherent factors into one common factor. Factor analysis is used to compress data and explore data structure (Szakaly et al., 2014, Balogh et al., 2015). This method contracts the basic variables into so called factor variables which cannot be directly observed. In most cases, factor analysis is used foremost in order to filter out multicollinearity [4].
Logistic regression quantifies the probability of occurrence of the category of a doubtful, category like dependent variable under the condition of the known outcomes of other explanatory variables. Logistic regression is a nonlinear classification method that does not suppose the continuity of explanatory variables neither the normality of multivariables. The decisionmaker can construct a decisionmaking rule relying on the hypothetical probability value in order to classify the given observation unit into a predetermined, result like category (Gal et al., 2013). If the number of the dependent variables' outcome is two, then the method is called a binomial logistic regression.
3. Results
The success and the efficiency of eLearning should be measured by a reliable method in order to use it effectively. Although, there are several studies about the success of eLearning systems, only a few of them is about the measurement of this success within the institutions. It is the study of DeLone and McLean (2003) which examines the success of the introduction of the eLearning system with the help of ELSS model (eLearning System Success). We made two questionnaires to evaluate the application of eLearning  one of them was for the students and the other one asked the lecturers. Therefore, the CUB's questionnaire responses are used as a basis of comparison.
The aim was to develop such questionnaires which are suitable both for the evaluation of the eLearning's quality and its economic benefits too. The basis of the eLearning's quality questions was Wang's article (2007), in which he measured the success of the eLearning systems. Therefore the questions of the students and the lecturers were the same.
There were 273 students and 50 lecturers from the CUB and 288 students and 46 lecturers from the UD FEB who properly filled out the questionnaires. We examined the answers about the eLearning's quality in this research on the basis of two criterions (studentlecturer, CUB  FEB). The basis of the answers' comparability was that both institutions applied the Moodle frame system. We tried to find out to what extent they exploit the facilities of the system.
The groups of questions are the following:
Quality of the system (17)
Quality of the information (812)
Quality of the service (1317)
Benefit of the eLearning system (1824)
Conclusions (2527)
The 27 questions could be answered in a scale of 10. Figure 2 represents a diagram that indicates the means of the answers of the two institutions' students and lecturers.
It is visible on Figure 2 that generally, according to the eLearning users of the CUB, the quality of the eLearning application is better. We have chosen the answers demonstrating significant differences based on the results of ttests, therefore we found significant differences in the answers.
3.1. tTest
The significant differences of the lecturers' answers is illustrated by the table 1. All results were higher at Benchmark excepting for the question 26.
Table 1
Regarding the result, it can be said that the eLearning application of the CUB is more successful than the FEB according to the students' and the lecturers' evaluation. It is also obvious that the quality of the system's operation of the CUB is higher than the FEB. This result supports our hypothesis according to which the FEB can evolve in the quality of eLearning application by ensuring the institutional frames for the system.
3.2. Factor Analysis
Hereafter, we examined whether background variables influencing the variables are possible to be formed. We used factor analysis to demonstrate this which contracts the coherent factors into one common factor. We examined the variables on the basis of the KaiserMeyerOlkin (KMO) criteria to determine whether they are suitable for factor analysis [2]. The value of the KMO is 0.886 (Table 2), which means that the variables are suitable for factor analysis.
The table also indicates the null hypothesis of the Bartlett test, which means that there is no correlation between the basic variables since the level of significance (Sig.) is smaller than 0.05. Consequently, the basic condition of the factor analysis, according to which the variables must correlate, is fulfilled.
Table 2


KaiserMeyerOlkin Measure of Sampling Adequacy.  0.886  
Bartlett's Test of Sphericity  Approx. ChiSquare  17 114.773 
df  351  
Sig.  0.000 
We used two methods to determine the number of the factors. One of them is the percentage of variance, which determines the number of the factors on the basis of the cumulated percentage of the variance, which means that it is necessary to establish such number of the factors which makes it possible to reach a cumulated minimal level of variance. The table 3 indicates the variance explained by the factors.
Table 3
The fourth row of the `Cumulative %' shows the cumulated variance of the four factors (64.219%) which were developed by the Kaisercriteria. It is above the necessary 60%.
The 5factor solution would have been reasonable regarding the methods but relying on the fulfilled factor analysis there would only be one variable in the factor 5. Therefore, we have chosen a 4factor solution which means that we replaced 27 variables which explains in 100% with 4 factors which explains in 64.22%.
After this, we rotated the factors during their selection to filter the correlated factors without relation and also in order to get a more simple and understandable solution. We used the Varimax rotational method during which the orthogonal rotation results in correlating factors.
Finally, we reached a 4factor solution as a result of the analysis, where the KMO = 0.886 and the explained variance is 64.22%. The names of the factors are the following:
FACTOR1: Quality of the service
FACTOR2: Efficiency of the system
FACTOR3: Quality of the online material
FACTOR4: Usability of the system
3.3. Binominal Logistic Regression
We used logistic regression for the results of the factor analysis. Our aim was to determine the importance of a given factor for the users of the CUB compared to those of the FEB.
Table 4 and 5 represent the first phase of the analysis. Table 4 shows the constant Waldstatistic in the preanalysis phase, which is the square of the beta (B) and the standard error. It demonstrated that there is not a significance.
Table 5 represents the individual effect of the independent variables yet not used in the analysis, according to which FACTOR1, FACTOR2 and FACTOR3 are also significant on their own, while the forth variable is not. The second part of the analysis demonstrates the final result. We used the,,Enter” method, which means that we used the four independent variables in the analysis at same time (Maslin, 2007).
Table 5


Score  df  Sig.  
Step 0  Variables  FACTOR1  49.687  1  .000 
FACTOR2  20.899  1  .000  
FACTOR3  22.418  1  .000  
FACTOR4  .401  1  .526  
Overall Statistics  97.189  4  .000 
Table 6 also applies the Waldstatistic. If the given variable is significant, then it supports the model. It is obvious that FACTOR1, FACTOR2 and FACTOR3 contributes to the model, while FACTOR4 does not. The Exp(B) indicates how each variables correct the estimation.
Table 6
Regarding this, FACTOR1 corrects the estimation the most (Exp(B)=2.135) with 113.5%, while FACTOR2 corrects it with 68.3%. FACTOR3 worsens the estimation with 44.9%, which means that according to the CUB users the first factor is twice, while the second factor is 1.683 times more important than according to the FEB.
4. Conclusion
We used statistic methods to examine the evaluation of the eLearning among the students and the lecturers of the FEB and the CUB. We found significant differences between the CUB's and the FEB's application as well as between the students' and the lecturers' evaluation by performing a ttest. We determined that relying on the result, that it is more effective and better in quality to operate the eLearning system under organized circumstances. This supports our hypothesis. We have created 4 factors from the 27 variables by factor analysis and we performed logistic regression on them. Our result was that according to the CUB users the quality of the service is more than twice as good, while the efficiency of the system is 1.683 times as more important than according to the FEB users.
The integrated eLearning Portal is being continuously developed and used since 2007. It is now an integral part of the educational process. The Moodle system is used more and more and functions of the education system and developed training room booking system are using an increasing number of computer rooms in online teaching employment exams, for exams. Several ways of student evaluation are also used by lecturers, over which the student evaluation system is available as we need to develop multifunction use. The proposed quality eLearning realized with the "quality eLearning" at FEB.