The effect of different online education modes on the English language learning of media studies students

Abstract The implementation of e-learning methods and their efficiency in the process of teaching English for specific purposes, particularly media and journalism, is a rather new phenomenon in Central Europe. The experiment conducted within this study was focused on the efficiency of e-learning and blended-learning modes. We created three groups of students with diverse modes of online education used in each group. The first group (18 students) was educated through the purely e-learning way, the second group (20 students) was taught through the classical face-to-face method, and the third group (18 students) through the blended-learning approach. The online education mode included interactive webinars with a native speaker who was providing live feedback on students’ assignments. Within the study programme of media/journalism the blended-learning mode seems to be the most efficient. Comparing the results of pre-tests to post-tests enabled us to specify the language skills which were improved in the three test groups. Scores of the students in all the four investigated areas (i.e. reading, speaking, listening and vocabulary) increased significantly in the blended-learning group. The vocabulary was the most improved language skill observed in both groups, meaning the e-learning as well as classic group. On the other hand, the online method resulted to an observable improvement in students’ performance as far as for the listening and speaking skills as it might simulate their future workplace.


PUBLIC INTEREST STATEMENT
The study deals with a rather new phenomenon in the region of Central Europe, namely the methods of e-learning and their efficacy (efficiency) in the process of teaching English for specific purposes (media and journalism).
The experiment carried out by the authors focuses on verifying the effectiveness of various models and tools in the education of English language with three groups students: group with a classical education, with only the online learning and last group used blended learning, i.e. classroom-bound lessons and e-learning was applied as a supplementary method.
The experiment confirmed that students' scores in all investigated areas (reading, speaking, listening and vocabulary) increased significantly in the blended-learning group. The vocabulary was the most improved language skill in both the e-learning and the classic group. The online method has led to an observable improvement of students' performance in listening and speaking skills. each group. The first group (18 students) was educated through the purely e-learning way, the second group (20 students) was taught through the classical face-to-face method, and the third group (18 students) through the blendedlearning approach. The online education mode included interactive webinars with a native speaker who was providing live feedback on students' assignments. Within the study programme of media/journalism the blended-learning mode seems to be the most efficient. Comparing the results of pre-tests to post-tests enabled us to specify the language skills which were improved in the three test groups. Scores of the students in all the four investigated areas (i.e. reading, speaking, listening and vocabulary) increased significantly in the blended-learning group. The vocabulary was the most improved language skill observed in both groups, meaning the e-learning as well as classic group. On the other hand, the online method resulted to an observable improvement in students' performance as far as for the listening and speaking skills as it might simulate their future workplace.

Introduction
The active use of foreign languages is currently considered as a necessity for everyone looking for their application in any type of trade. Besides the classical classroom teaching modes, more and more space is dedicated to some alternative teaching models in the field of professional training. Such alternative models, including e-learning, are complementary, and in some cases can nicely replace traditional teaching approaches. E-learning is essentially a virtual form of education that may, in certain circumstances, be either fully or partially combined with the classical forms of education (Clark & Mayer, 2016). If we are to examine the so-called classical e-learning models of education, where the computer is understood in the strict sense (i.e. as hardware), then we can identify the following basic sub-modules of e-learning in this system: • Professional software (an interactive CD, DVD, etc., not requiring any access to the Internet); • Online education through the Internet and through programmes or websites intended for educational purposes; • Online education through the Internet, but without websites directly intended for educational purposes (e.g. e-mail consultations, sending documents, assignments via e-mailing, etc.).
Education and learning by means of the above-mentioned methods has been examined on a long-term basis, particularly in Western Europe, the United States of America and Canada, where to these methods of education and learning have been paid special attention for several decades (e.g. Major, 2015). Major (2015) lists and analyses video-conferences between secondary school students in New York City, USA and Moscow, Russian Federation (in the 1980s as the beginnings of online learning), while many scientific studies are based on empirical researches, in particular comparative ones. In this context, e.g. the experimental education performed under the supervision of Major (2015) may be mentioned. Also, Major (2015) analyses the outputs of education/ teaching the subject that was carried out with 4 students attending the classroom-bound teaching mode and 22 students present via the so-called distance mode of education (via the web camera). It is also necessary to highlight the growing body of professional literature reporting current experiments in online education through mobile equipment (hardware). The New Media Consortium (2017) has been pointing out to the increasing importance of mobile devices in the online education environment, continuously and perseveringly reporting this fact for several years. And in the end, this fact has already been accepted by the academia (Pachler, Bachmair, & Cook, 2010). It is also worth noting that the research in the online training starts to put an even greater emphasis on the student-student interaction as an integral component of the online training process (Anderson & Elloumi, 2008). This is caused by the massive presence of student generation aged up to 26 years in average in the online environment.
While focusing on the media studies students, there are certain specific features in the teaching of English. This results, in particular, from the fact that for students of this field, retaining context of the Slovak Republic, it is necessary to consider the primary objectives to include an English language course in their study programme. In principle, two basic purposes can be recognized: • Learning and mastering a professional language in order to be able to process media contents and similar materials given in the Slovak language (basically, such working positions as editors and presenters, etc.). This presumes the above-standard ability to search and adapt, or to locate the already existing information in the Slovak context. Adaptation and location of information are considered more relevant operations than translation into the Slovak language. When one is something adapting or locating, it is dealt not only with the translation of information but also with its adaptation, including the used terminology, to the conditions of the recipient in the Slovak Republic environment, thus written or articulated in the Slovak language. In practice very often, translations of media materials and articles from English into Slovak are factually incorrect, even incomprehensible from the percipient's perspective, since the secondary originator of the information did not adapt it at all (to the Slovak context), merely or faintly translated it, often actually literally.
• Learning and mastering a professional language for the purposes of a subsequent use in the active communication form (e.g. used in the jobs of editor, presenter and spokesperson) is usually a more demanding purpose of language use. The reasons vest in it that the student do not only use their skills to translate some information gathered and prepared in advance, but they are creating and shaping the information as such-either in the oral or other form (written, online, etc.). When teaching English as a foreign language, depending on the form, educators' attention should be focused on improving either speaking or writing skills in English (Galloway, 2013).
Notwithstanding the above, it is necessary to state that teaching English in media studies requires-besides teaching the terminology used in the media environment-to provide students some selected terminology from other fields of life as well that they may come across in their future job and professional activities (e.g. law, finances, statistics, culture, education, sciences, history, social sciences, mainly political science, public administration and international relations).
In their study, Guarda and Helm (2017) found that the professional development of lecturers should be focused not only on the language as such but also on its practical use and the inevitable multilingual environment of higher education sphere. The traditional teaching of correct pronunciation has been replaced by an alternative corpus-based approach. This allows corpora to be integrated into the academic environment as a perspective technology and method. The alternative corpus-based class showed better progress of students in English (Kuzminykh & Khoroshilova, 2017). Kitchakarn (2016) shares the opinion that Facebook and other social networks should have their place within the teaching and learning activities as a platform for students' selfstudy, allowing them to exchange their ideas, attitudes, comments, feelings and meet the assignments in order to improve their knowledge in the field of grammar and writing as well. Wang, Cheng, Chen, Mercer, and Kirschner (2017) implement the concept mapping as a tool for the group learning and interaction in classrooms. The sources of such concepts are numerous computer and web-based applications in the online environment.

Materials and methods
Within the context of the Slovak Republic, e-learning has not reached such an important position yet; it has had rather a supplementary or replacing nature. Therefore, our effort was to empirically verify some benefits of e-learning previously anticipated by some theories (Kurucova, 2010a). Our experiment was carried out during one semester in 2016 with the three groups of Media Studies students. All the three groups were taught by the same teacher within the course titled English in the Media, covering the same course syllabi. The course was focused on the professional English language used in mass media (Kurucova, 2010b). The groups differed in the teaching modes. In the first group (Classic), there were 18 students educated only in the classical face-to-face model of education. The second group (E-learning) consisted of 20 students who were educated only in the online learning mode. The third group (Combined) comprising of 18 students used the blendedlearning mode, i.e. students attended classroom-bound lessons and e-learning was applied as a supplementary method. The total number of lessons was identical for each group in order to eliminate any possible impact of the teaching extent in favour of the teaching mode as such.
Education in the Classic group was performed in the standard way, by personal interactions between the teacher and students. The E-learning group (20 students) was taught in the online mode, primarily utilizing the following instruments: (a) Webinars, meaning live lectures in an identical manner as in the classical mode. Online platforms allowed all students to ask questions, i.e. participate actively in lectures.
(b) Check-up tools: (b1) Online Internet link to a native speaker, who listened to the students' reports and then provided them with feedback in a live transmission; (b2) Online training exercises with immediate error corrections (especially grammar and vocabulary exercises); (b3) International online group-work where some closed groups of students worked on identical tasks. Subsequently, their assignments were evaluated.
The blended learning approach was used in the Combined group. The students attended the classical face-to-face teaching process and at the same time they took part in the selected online modules applying the instruments given above in (a) and (b) for the E-learning group. The extent of webinars and check-up tools in this group did not exceed 50% of the total number of lessons.
The students were divided into the three groups by random selection. They varied in their knowledge levels (linguistic knowledge B2 or C1 according to the Common European Framework of Reference (Council of Europe, 2001)), age (ranging from 20 to 55 years) and gender.
The language skills of the engaged students had been measured using a pre-test in all the three groups at the time before the start of our experiment. The pre-test comprised of the following four parts: writing, listening, speaking and vocabulary. Each part was divided into the three subparts as detailed below. Students could obtain maximum 100 points in each subpart.
The first, written part of the pre-test (Writing) focused on Grammar (G), Terminology (T) and Structure (S). We examined the ability of students to produce a coherent text that should use accurate grammar, utilizing also appropriate terminology.
The second part of the pre-test assessed students' listening skills (Listening), where we focused on the following subsystems of language and perception: Reproduction (R), Questioning (Q), and Terminology (T). The test was performed to find out whether the students understood the text as a whole (reproduction), whether they were able to perceive its specifics (thus to answer specific questions), and if they used correct terminology (T) in their replies.
In the third part of the test (Speaking), we assessed the fluency of speech (Fluency-F), the suitability of vocabulary (Terminology-T), and the accuracy of grammatical phrases (Grammar-G).
In the last (fourth) part of the test (Vocabulary), we tested the students' knowledge of general vocabulary (Universal-U), professional vocabulary (Professional-P) and the suitability of its application (Application-A).
In the present report, we analyse the outcomes and results achieved by the participants belonging to the various groups. Descriptive statistics, Kruskal-Wallis' test and the one-sample Wilcoxon signed-rank test were used to evaluate the pre-test results (Kralova, Skorvagova, Tirpakova, & Markechova, 2017). All calculations were made using the STATISTICA software. Table 1 shows the results of participants: average scores in the pre-test in individual groups of participants for each observed area.
We tested whether the differences in pre-test results were statistically significant. Whereas the presumption of normal distribution of observed variables was not justified, in order to verify the statistical significance of differences between the student groups in pre-test results we employed the non-parametric method referred to as the Kruskal-Wallis test (Wilcox, 2009). It allows testing the hypothesis H 0 that the k independent samples k ! 3 ð Þcome from the same distribution. The test criterion of the Kruskal-Wallis test is the statistics which under the validity of null hypothesis is asymptotically χ 2 -distributed with k À 1 degrees of freedom (Hettmansperger, Möttönen, & Oja, 1998). The tested hypothesis H 0 is rejected at the significance level α if H ! χ 2 α k À 1 ð Þ, where χ 2 α k À 1 ð Þis the critical value of χ 2 -distribution with k À 1 degrees of freedom. If the null hypothesis is rejected in favour of the alternative hypothesis H 1 , which means that k independent samples k ! 3 ð Þcome from the same distribution, a new question is raised which pairs of samples differ significantly (Anděl, 2003). Duncan test, Tukey method, Scheffe method or Neményi test can be used in order to answer this question (Voss, 2010). In our case, the multiple comparison of mean ranks for the Kruskal-Wallis test was performed in the STATISTICA software by means of the Kruskal-Wallis test of multiple comparison via Z-score Applying the Kruskal-Wallis test, we verified whether the three groups of students (E-learning, Classic and Combined) varied considerably in the observed variable that was the resulting score in the pre-test.
In general, performing the Kruskal-Wallis test in the used statistical software enables to evaluate the test using the so-called p value, that is the probability of error made when the null hypothesis is rejected (Ryan, 2006). If the value of probability p is small enough (p<0:05 or p<0:01), the tested hypothesis H 0 (at a significance level of 0.05 or 0.01) is rejected. In the opposite case, the null hypothesis cannot be rejected.
The report of the software comprised the value of the test criterion H ¼ 1:092 and p value, p ¼ 0:579. Since p is greater than 0.05, the hypothesis H 0 cannot be rejected, the observed differences are not significant. Pre-test results are illustrated in Figure 1. The Kruskal-Wallis test confirmed that our individual groups did not differ significantly as far as to the achieved pre-test score. Thus, we considered the three groups of students to be equivalent before the start of our experiment. In other words, we can anticipate that the observed students' progress is related to the teaching modality and not to random differences in the groups.
The students of the three groups went through the same post-test at the end of the experiment. The test comprised the same four parts and it was evaluated in an identical manner as the pretest. Significance of differences between the groups, i.e. whether the group of participants had an impact on the average value of the observed variable, was confirmed by statistical methods. We tested the null hypothesis H 0 that the scores achieved in the post-test are the same in all three groups against the alternative hypothesis H 1 by means of the Kruskal-Wallis test. In cases when the test confirmed that there was a statistically significant difference between the observed groups, we calculated which groups differ significantly by means of the Kruskal-Wallis' test of Further, we investigated which areas (Writing, Listening, Speaking and Vocabulary) improved significantly in each of the observed groups of students (E-learning, Classic, Combined). The onesample Wilcoxon signed-rank test was used for testing the statistical significance of differences between the scores achieved by the students of particular group (E-learning, Classic, Combined) in the pre-test and the post-test in each of the four areas (Writing, Listening, Speaking and Vocabulary). The one-sample Wilcoxon signed-rank test is a non-parametric alternative to the one-sample t-test when the data are not normally distributed.

Results
The average scores obtained by students in the post-test for each group of participants and for each assessed area are given in Table 2. Table 2 show that there were differences between the groups in the achieved posttest scores. The test criterion of the Kruskal-Wallis' test assumed the value H ¼ 12:760. The p ¼ 0:002 was smaller than 0.01; the null hypothesis was therefore rejected at the significance level 0.01 in favour of the alternative hypothesis. The test confirmed that the groups of participants (Elearning, Classic, Combined) differed significantly in the post-test scores.

Results in
Whereas the test confirmed that there was a statistically significant difference between the observed groups, we used the Kruskal-Wallis test of multiple comparison to calculate Z-score values and corresponding p values. Since the null hypothesis was rejected, we list the p values only (Table 3). Table 3 shows that with respect to the post-test results, the E-learning and Combined group (p ¼ 0:011) and also Classic and Combined (p ¼ 0:002) were significantly different. However, the difference between the E-learning and Classic groups (p ¼ 1:000) was not statistically significant. The situation is visualized in Figure 2.
Based on the results of the analysis, we conclude that an isolated application of classical teaching or an isolated application of e-learning had approximately similar or comparable effects

E-learning group
The one-sample Wilcoxon signed-rank test was used to test the null hypothesis H 0 that the distributions of pre-test and post-test scores are the same, or the differences between the participants' scores in the pre-test and the post-test in each area are not significant. The value of the test criterion Z in the one-sample Wilcoxon signed-rank test and the p values are summarized in Table 4. Also in this case, the test results were evaluated by the p value.
Whereas the calculated p values in three out of four investigated skills (Writing, Listening and Speaking) exceeded 0.05, the tested hypothesis H 0 was not rejected for the all three skills. In other words, the differences between the scores obtained by the participants in the pre-test and the post-test in Writing, Listening and Speaking were not significant.
Whereas for the Vocabulary area the p value was lower than 0.01, thus the null hypothesis was rejected at the level of significance α ¼ 0:01. The results of the Wilcoxon test indicate that a significant change in scores of the students of this group occurred only in the Vocabulary area.  Our results indicate that working with online tools did not have any greater impact on the fields of Writing and Vocabulary. This may relate to the fact that the most modules implemented in the research were the tools implying interactive communication. Students were provided with the possibility to have a grammar test checked on the spot by clicking on the icon "correct answers". However, except for providing an immediate feedback, this feature seemed not to have added any educational value.

Classic group
Similarly to the E-learning group, the null hypothesis H 0 was tested by means of the one-sample Wilcoxon signed-rank test in the Classic group of students, stating that the difference between the participants' scores in particular areas in the pre-test and the post-test was not significant.
The one-sample Wilcoxon signed-rank test was applied for each of the observed areas (Writing, Listening, Speaking and Vocabulary). The test criteria Z and the p value are summarized in Table 5.
Whereas the p values for Writing, Listening and Speaking (Table 5) were not lower than 0.05, the tested hypothesis H 0 was not rejected, indicating that the differences between the participants' scores in the pre-test and the post-test in the assessed areas were not significant. Only in the Vocabulary area the p value was lower than 0.01, thus the tested hypothesis H 0 was rejected at the significance level α ¼ 0:01. In the E-learning group as well as in the Classic group, a significant change in Vocabulary was found.
Although the differences between the scores in Writing, Listening and Speaking were not significant, the scores in all the three areas increased. We suppose that this is due to several reasons. The classroom-bound education mode should lead to a substantial improvement of individual language skills. However, none of these skills per se is said to profit from the personal interaction in any special manner. In addition, the insignificant improvement of students' performance may be caused by the dynamics of education in the classroom. The instruction scheduling (one 90-min lesson per week × 12 weeks) represents a limiting factor not only for practising and developing the existing skills, but in particular, for the acquisition of new knowledge, abilities and skills.
The observed significant increase in scores covering Vocabulary might be determined by the way in that the lessons were structured. Students acquired a new technical terminology while listening to the lecturers' explanation. Following, they produced and analysed texts that were reviewed by their peers in the class. There was only limited time for practising the speaking skills, and not all students were provided with equal time space for their oral presentation within the lesson. By contrast, the tutor in the online mode listened to the reports prepared by students and then provided them with an individual feedback in the course of live transmission. The advantage of online education in such cases, among other things, is the question of students' time optimization.

Combined group
The null hypothesis H 0 , stating that the difference between the scores obtained by the participants of the Combined group in the pre-test and the post-test in Writing (or Listening, Speaking or is not significant, was tested by means of the one-sample Wilcoxon signed-rank test. The test criterion Z and p value for each of the observed areas is reported in Table 6. Whereas the p values in all four areas Writing, Listening, Speaking and Vocabulary (Table 6) were lower than 0.01, the tested hypothesis H 0 was rejected at the significance level α ¼ 0:01. Combined-group students' scores were improved significantly in the all observed areas. In the experiment, the blended-learning mode was proven to be the most effective approach. It stems from the fact that this model of education may involve the benefits of the other two approaches, online and classical education modes. The scores of students studying in the blended-learning way increased significantly in all the observed variables. The improvement of scores of Combined-group students was the highest among the three investigated groups.
The improvement in students' post-test scores was in all the assessed areas and to the similar extent. Individual scores did not show any disproportions in values. Thus, it can be stated that within the blended-learning mode, all the investigated skills and knowledge were communicated effectively to the students. Moreover, it may be concluded that the most suitable methods of instruction were applied. Some students in the Combined group showed decreasing interest in the field of online activities. This decrease may be either due to some insufficient technical means (hardware and/or internet connection) for streaming the lectures, or due to the students' preference to read from a paper version and not preferring always only reading from a screen, as reported by several authors (Walsh, 2016).

Discussion
In this paper, we investigated the effects and outcomes of the three teaching modes (classroom-bound, e-learning and blended learning) on the students' scores in the four specific language skills (writing, listening, speaking and vocabulary). The performance of students who attended the blended-learning mode improved the most. Furthermore, students' scores achieved in any of the assessed areas did not increase significantly more than in the others. Abilities of students in the purely classroom-bound mode and in the purely e-learning mode were developed significantly only in the field of vocabulary.
Although the significant improvement in the all criteria was observed only in the Combined group, it is necessary to note that the examiner of the online group reported a progress of students in their listening and speaking abilities. This corresponded to the expected improvement in the language skills that are adequate to the extent of the course preceding the post-test. It may be related to the fact that in the media studies (focused on the training for future presenters and editors), the online teaching mode is not just only a way to transfer information (theoretical knowledge) but also some practical training in an environment simulating the real workplace conditions of presenters/editors, using camcorders, microphones and other relevant equipment. Therefore, the online education mode becomes not only the educational tool for the theory but also an environment imitating the future workplace of students. Nevertheless, as suggested by previous studies (Clark & Mayer, 2016), it is hardly possible to convert the entire classroom-bound education to the purely online instructions. One of the most discussed issues related to e-learning is its limited potential to develop students' practical skills (Ellaway & Masters, 2008;Holmes & Gardner, 2006;Kratochvil, 2014;Watkins, 2005). In contrast with the previously reported studies claiming that the e-learning efficiency is substantially lower in the education of practical skills, the teaching of English for media studies turns out to be the exact opposite to this assumption. In our study, it has been demonstrated that in case of students studying solely in the online mode, there has been a more observable improvement, precisely of those practical skills forming the core of the professions in question. In general, there is a fair imbalance between the number of lecturers teaching online via videos and audio-conferences and the amount of research discussing these particular forms involved in online teaching/learning. Moreover, there are only very few instructional guidance sources focusing on and covering just the video-and audio-conferencing area (Kozar, 2016).
Although the blended-learning mode proved to be effective, the students' interest in the online tools decreased gradually. This was also reported in earlier studies stating that . . .e-learners must be able to self-manage the entire learning process including self-regulation of behaviour, motivation, and cognition, proactively and deliberately. The core of self-regulated learning is self-motivation. Students' motivation is a major factor that affects the completion rates in the Web-based course and a lack of motivation is also linked to high dropout rates. (Saba, 2012) Our study has shown that the way of blending the classical and e-learning approaches is an extremely important factor for achieving the desired effect. We assume that it is the key indicator determining the intrinsic motivation of students to complete the course. This corresponds to the earlier findings reported by Drago and Wagner (2004) who stressed the role of the teacher as follows: The instructor in e-learning courses should facilitate, stimulate, guide, and challenge his/her students via empowering them with freedom and responsibility. Instructor feedback to students can improve learner affective responses, increase cognitive skills and knowledge, and activate meta-cognition, which refers to the awareness and control of cognition through planning, monitoring, and regulating cognitive activities.

Conclusions
Results of our research imply that the blended-learning mode is more effective than the purely classical or purely e-learning modes of education. Moreover, the scores of students improved similarly in the all areas. The careful arrangement of classroom-bound and online parts of the course appears to be the crucial factor influencing improvements in the communication and language skills of Media Studies students.
The reported experiment confirmed that the performance of students in all the three groups was improved. Thus, special attention should be paid to the relation between the particular language skill and the applied method of instruction. The online modules helped students to improve their listening and speaking skills, since the online exercises and tasks provided authentic opportunities to practising these skills via the actual and live communication with a native speaker. On the other hand, it is worth noticing that the communication served not only for foreign language learning but also as a simulation of students' future workplace conditions. This undoubtedly raised students' motivation to get involved in the online modules, since they might be perceived as an authentic activity. In spite of the fact that in training for other professions, the online education is a preferred tool, especially for the theoretical part of the curriculum, in case of journalists it may be regarded as a valuable tool for practical exercising of job-related skills. The e-learning approach could be effectively applied in all such fields where the process of online education imitates the real tasks and duties of a future job, and at the same time in such fields where the active oral communication and the ability to understand the spoken word is required.
Definitely, the use of information and communication technologies for simulating the workplace conditions of future journalists and its influence on foreign-language learning mode deserves more