MOOCs and the Claim of Education for All: A Disillusion by Empirical Data

MOOCs have shaped the discussion on learning with digital media for the last few years. One claim of MOOCs in the tradition of Open Educational Resources is to expand access to education, mainly in the field of higher education. But do MOOCs meet this claim? The empirical data in this article confirm the suspicion that, despite all the heterogeneity of the participants, MOOCs are mostly used by people with a higher level of education. Data of participants from two MOOCs from Germany, as well as, empirical data from large providers and universities are used. But due to the different forms of MOOCs there is no comprehensive proof possible. With respect to the Knowledge Gap Theory and the Digital Divide, a theoretical framework is provided to explain possible causes of a different usage. The aim of the article is to point out the risks of an increase of inequalities as a consequence of hyping MOOCs and to stimulate a discussion about possible answers to make MOOCs an instrument of education for all. Introduction Unequal access to information and related unequal educational chances are considered to be the basis for the existing imbalance of power and differential possibilities for individual development. The permanent availability of information on the internet establishes the possibility that the traditionally institutional and status related barriers to education should not exist anymore. Therefore, increasing access to the internet in the last decade is a hopeful sign1 even though 1 In 2013 76.5% of the German population was using the Internet, which makes a total of about 53.7 million citizens of Germany aged 14 and above. This means that the amount of internet users in Germany has more than doubled since 2001 (TNS Infratest, 2013). In the USA in 2013 84.2 percent were using the internet. In 2001 the individuals using the internet nearly reached 50 percent already (ITU, 2014). MOOCs and the Claim of Education for All: A Disillusion by Empirical Data Rohs and Ganz 2 access, especially in developing countries, is still significantly lower than in industrialized countries. In order to promote the chances of a better education for all via the internet, various initiatives for open educational resources (OER) have been launched. The aims of this movement are written in the “Paris OER declaration” which includes the recommendation: “Promote and use OER to widen access to education at all levels, both formal and non-formal, in a perspective of lifelong learning, thus contributing to social inclusion, gender equity and special needs education” (UNESCO, 2012). This would imply that not only will access to educational resources be generally improved, i.e. more people will be using educational resources, but also that inequalities will be reduced, i.e. people who previously had no access to education will now get access. The core principle of MOOCs is to provide access to education for a wide audience and thus also increase the access to education. Therefore, we clearly see MOOCs in the tradition of Open Access, Open Education, and the OER movement (Klobas, Mackintosh & Murphy, 2015) (see figure 1). Figure 1. MOOC Timeline (Yuan & Powell, 2013, p. 6) This intention is often emphasized in relation to MOOCs, e.g. by EdX CEO Anant Agarwal in an interview done with Forbes online. In the interview, he named the goals of MOOCs “to increase access to quality education, to improve teaching and learning on campus, and to conduct research into how students learn” (Kanani, 2014). EdX also wants to “provide truly world-class online courses to everyone, everywhere, regardless of social status or income, while also improving oncampus education” (Kanani 2014). MOOCs and the Claim of Education for All: A Disillusion by Empirical Data Rohs and Ganz 3 Despite the hope for more equal access to education through MOOCs, the empirical data show (section 4) that MOOCs potentially reinforce inequality. In this article we will give a theoretical background to explain why MOOCs are mostly used by more highly educated people (section 2) and stimulate a discussion on if and how MOOCs can contribute to equal access to education promoted by Open Educational Resources (OER) (UNESCO, 2012). Theoretical Background As part of the theoretical considerations, two related approaches will be discussed that provide an explanation for the unequal use of information resources: The Knowledge Gap Theory developed in an era of traditional mass media and the follow-up expansion of that approach concerning the Digital Media. Knowledge Gap and Digital Divide An explanation of inequality within the prerequisites of access and the forms of use of open access information is provided by the Knowledge Gap Theory (Tichenor, Donohue & Olien, 1970), which argued that the increase of information in society leads to differing reception dependent on socioeconomic status: “As the infusion of mass media information into a social system increases, segments of the population with higher socioeconomic status tend to acquire this information at a faster rate than the lower status segments, so that the gap in knowledge between these segments tends to increase rather than decrease” (Tichenor, Donohue & Olien, 1970, p. 159). With regard to digital media, which is increasingly taking over the former role of mass media, the term digital divide has evolved to describe the aforementioned theory since the 1990s (Brown, Barram & Irving, 1995). Since that time, a number of researches have been carried out which led to a diverse state of knowledge (for a summary, see Visvanath & Finnegan, 1996). The central constructs of the Knowledge Gap Theory are: growth in knowledge, socioeconomic status, and time. Wirth (2006) outlines this relationship with the following arguments: 1) Communication skills: People with a higher socioeconomic status possess better competence in reading and understanding for the acquisition of political and scientific topics. 2) Existing knowledge: People with higher socioeconomic status possess a better topical prior knowledge, due to their formal education and their usage of media, which leads to easier acquisition of knowledge. 3) Social contacts: People with higher socioeconomic status have more social contacts, which leads to a higher possibility for exchange about relevant topics. MOOCs and the Claim of Education for All: A Disillusion by Empirical Data Rohs and Ganz 4 4) Selective use of information: People with higher socioeconomic status have better information literacy skills. 5) The special position of print media: Political and scientific topics are mainly covered by the print media, which are more often received by the formally higher educated (Wirth 2006, p. 169f). Due to these different causes which inter-relate, reinforce, or raise their effects, there are three theses on the digital divide to be discussed below: Access Gap. It is not surprising that there are still differences in access to the internet between developed industrial nations in North America, Europe, and the 3rd world developing countries, in particular, as shown in Figure 2. Figure 2. Individuals using the internet (developed and developing regions) Above all the reasons for this are attributed to the individual economic situation, but also to the technical conditions in developing countries especially for the poorer sections of the population. But the access gap isn’t only a topic between developing and industrialized countries; there is also an access gap within industrialized countries. Although in Germany about 75% of the population have access to the internet, it can be seen that “despite numerous initiatives the digital divide has MOOCs and the Claim of Education for All: A Disillusion by Empirical Data Rohs and Ganz 5 not yet been overcome in Germany” (TNS Infratest, 2013, p. 19; translation by authors2). These inequalities are dependent on Gender, Age, Employment, Educational Background, or Household Net Income (TNS Infratest, 2013):  Gender: 81.3% of males and 71.8% of females are online.  Age: 97.5% of the group aged 14-19, but only 30.2% of people aged 70 and above use the internet.  Employment: While 89.1% of the employed population is online, only 61.2% of the unemployed are.  Educational Background: There are 91.7% of people having a university degree can be considered as being online, while it is only 60.7% of the graduates from a secondary school. .  Household net income: Only 55% of people with a household net income of 1000 Euro and less per month are online, whereas 93% of the people with a household net income of 3000 Euro and above. In summary, this demonstrates clearly that not only are there differences between developing and developed countries, but also that socioeconomic factors within industrial countries influence access to the internet (Zhang, 2013). Usage gap. A further dimension is how people are making use of freely accessible content in the internet. It can be observed that different segments of the population also tend to differ in their usage strategies. Therefore, technology can act as a resource for equalizing but usage and interpretation are dependent on dispositions, which means without the proper basic skills that facilitate the acquisition and usage of available media, these resources cannot be used at all or only to a very limited degree (Zillien, 2009). Thus the access to digital media is not equivalent to being able to make use of them in a productive way. This false conclusion is also frequently connected to the so called “net generation” that grew up with the internet being available all the time and therefore supposedly show better usage competences than older generations. However, a closer look reveals a more generation-specific differentiation of use practices (DIVSI, 2012). The differences in cultural capital3, being habitus and skills, which are developed by socialization and education, are often held to be responsible for different use


Introduction
Unequal access to information and related unequal educational chances are considered to be the basis for the existing imbalance of power and differential possibilities for individual development.
The permanent availability of information on the internet establishes the possibility that the traditionally institutional and status related barriers to education should not exist anymore. Therefore, increasing access to the internet in the last decade is a hopeful sign 1 -even though 1 In 2013 76.5% of the German population was using the Internet, which makes a total of about 53.7 million citizens of Germany aged 14 and above. This means that the amount of internet users in Germany has more than doubled since 2001 (TNS Infratest, 2013). In the USA in 2013 84.2 percent were using the internet. In 2001 the individuals using the internet nearly reached 50 percent already (ITU, 2014). access, especially in developing countries, is still significantly lower than in industrialized countries.
In order to promote the chances of a better education for all via the internet, various initiatives for open educational resources (OER) have been launched. The aims of this movement are written in the "Paris OER declaration" which includes the recommendation: "Promote and use OER to widen access to education at all levels, both formal and non-formal, in a perspective of lifelong learning, thus contributing to social inclusion, gender equity and special needs education" (UNESCO, 2012). This would imply that not only will access to educational resources be generally improved, i.e. more people will be using educational resources, but also that inequalities will be reduced, i.e. people who previously had no access to education will now get access.
The core principle of MOOCs is to provide access to education for a wide audience and thus also increase the access to education. Therefore, we clearly see MOOCs in the tradition of Open Access, Open Education, and the OER movement (Klobas, Mackintosh & Murphy, 2015) (see figure 1). (Yuan & Powell, 2013, p. 6) This intention is often emphasized in relation to MOOCs, e.g. by EdX CEO Anant Agarwal in an interview done with Forbes online. In the interview, he named the goals of MOOCs "to increase access to quality education, to improve teaching and learning on campus, and to conduct research into how students learn" (Kanani, 2014). EdX also wants to "provide truly world-class online courses to everyone, everywhere, regardless of social status or income, while also improving oncampus education" (Kanani 2014).

Figure 1. MOOC Timeline
Despite the hope for more equal access to education through MOOCs, the empirical data show (section 4) that MOOCs potentially reinforce inequality. In this article we will give a theoretical background to explain why MOOCs are mostly used by more highly educated people (section 2) and stimulate a discussion on if and how MOOCs can contribute to equal access to education promoted by Open Educational Resources (OER) (UNESCO, 2012).

Theoretical Background
As part of the theoretical considerations, two related approaches will be discussed that provide an explanation for the unequal use of information resources: The Knowledge Gap Theory developed in an era of traditional mass media and the follow-up expansion of that approach concerning the Digital Media.

Knowledge Gap and Digital Divide
An explanation of inequality within the prerequisites of access and the forms of use of open access information is provided by the Knowledge Gap Theory (Tichenor, Donohue & Olien, 1970), which argued that the increase of information in society leads to differing reception dependent on socioeconomic status: "As the infusion of mass media information into a social system increases, segments of the population with higher socioeconomic status tend to acquire this information at a faster rate than the lower status segments, so that the gap in knowledge between these segments tends to increase rather than decrease" (Tichenor, Donohue & Olien, 1970, p. 159).
With regard to digital media, which is increasingly taking over the former role of mass media, the term digital divide has evolved to describe the aforementioned theory since the 1990s (Brown, Barram & Irving, 1995). Since that time, a number of researches have been carried out which led to a diverse state of knowledge (for a summary, see Visvanath & Finnegan, 1996).
The central constructs of the Knowledge Gap Theory are: growth in knowledge, socioeconomic status, and time. Wirth (2006) outlines this relationship with the following arguments: 1) Communication skills: People with a higher socioeconomic status possess better competence in reading and understanding for the acquisition of political and scientific topics.
2) Existing knowledge: People with higher socioeconomic status possess a better topical prior knowledge, due to their formal education and their usage of media, which leads to easier acquisition of knowledge.
3) Social contacts: People with higher socioeconomic status have more social contacts, which leads to a higher possibility for exchange about relevant topics. 4 4) Selective use of information: People with higher socioeconomic status have better information literacy skills.
5) The special position of print media: Political and scientific topics are mainly covered by the print media, which are more often received by the formally higher educated (Wirth 2006, p. 169f).
Due to these different causes which inter-relate, reinforce, or raise their effects, there are three theses on the digital divide to be discussed below: Access Gap. It is not surprising that there are still differences in access to the internet between developed industrial nations in North America, Europe, and the 3rd world developing countries, in particular, as shown in Figure 2. But the access gap isn't only a topic between developing and industrialized countries; there is also an access gap within industrialized countries. Although in Germany about 75% of the population have access to the internet, it can be seen that "despite numerous initiatives the digital divide has not yet been overcome in Germany" (TNS Infratest, 2013, p. 19 In summary, this demonstrates clearly that not only are there differences between developing and developed countries, but also that socioeconomic factors within industrial countries influence access to the internet (Zhang, 2013).

Usage gap.
A further dimension is how people are making use of freely accessible content in the internet. It can be observed that different segments of the population also tend to differ in their usage strategies. Therefore, technology can act as a resource for equalizing but usage and interpretation are dependent on dispositions, which means without the proper basic skills that facilitate the acquisition and usage of available media, these resources cannot be used at all or only to a very limited degree (Zillien, 2009).
Thus the access to digital media is not equivalent to being able to make use of them in a productive way. This false conclusion is also frequently connected to the so called "net generation" that grew up with the internet being available all the time and therefore supposedly show better usage competences than older generations. However, a closer look reveals a more generation-specific differentiation of use practices (DIVSI, 2012).
The differences in cultural capital 3 , being habitus and skills, which are developed by socialization and education, are often held to be responsible for different use practices of digital media. In fact empirical studies show no clear border between an "information elite and entertainment proletariat" 4 (Eichmann, 2000; translation of book title by authors), but a stronger orientation towards information and educational usage can be identified within higher status groups, which is supported by recent analysis of further education (Leven, Bilger, Strauß & Hartmann, 2013).

Reception gap. The reception gap finally focuses on the individual information literacy:
If users have the same access conditions, the individual capacity for dealing with information, e.g.
content search and rating strategies as well as cognitive dispositions, can lead towards differences in the reception of information. These findings can be condensed to the assumption that people with higher socioeconomic status are able to derive a higher benefit from a wide variety of educational possibilities provided by digital media, specifically the internet (Zillien, 2009, p. 241).
In the following section, how these arguments meet the use of MOOCs will be investigated.

Digital Divide and MOOCs
Massive Open Online Courses (MOOCs) have dominated the discussion on learning with digital media during the last years. Though, after excessive research, it appears to the authors that up to now the connection between MOOCs and the digital divide has only been established implicitly but never explicitly. Therefore, this theoretical framework has been chosen for this article. At first glance it seems that MOOCs are a useful as a strategy to close the gap between the information rich and the information poor by opening up the traditional universities and their content in new ways. But this view leaves aside the influence of socio-cultural and educational backgrounds, which may lead to a different reception of freely accessible resources like MOOCs which may involve OER (see Chapter 2.1). This means, with respect to the knowledge gap hypothesis and digital divide that the availability of those educational resources (MOOCs; OER; etc.) is especially useful for people with higher socioeconomic status and / or educational background and is associated with a different kind of motivation and reception of learning offers. Therefore, the chance of widening participation in education faces the risk of reproducing existing inequalities and may even be increasing them. This means that people with higher socioeconomic or educational status will benefit more from the opportunities MOOCs offer than other people. In consequence, MOOCs can enlarge the knowledge gap rather than close it. Therefore the question arises, if the expectation associated with MOOCs is that they will improve access to education by online teaching, which will lead to a higher participation by traditionally educational disadvantaged groups (Lane, 2013) or whether MOOCs will contribute to increase unequal chances for different population groups. In order to explore this assumption, the dimensions described in Section 2.1 will be discussed. familiarity with culture which can only be produced by family upbringing when it transmits the dominant culture." (Bourdieu, 1977, p. 494) 4 germ.: "Informationselite und Unterhaltungsproletariat" These resources can easily be used in order to create MOOCs, which can be freely accessible and may therefore be one important pillar for education in poorer regions of the world or for targetgroups whose financial background does not allow them to take part in educational offers that are not freely accessible. Therefore, we assume, that MOOCs and OER are very closely connected in their ideals and history but do not determine one another.
It proves to be especially problematic that in developing countries the internet is available to those who hold the necessary financial resources. It also proves to be not solely a problem of developing countries, as people with a low income in developed countries show a tendency not to invest in an internet connection (Zhang, 2013). But basic exclusion already happens when it comes to availability and reliability of ICT (Simon et al., 2011). This leads to the fact also that  (Ho et al., 2014). Breslow (2013) showed also that there is greater participation in urban areas than in rural areas and that there are nearly no participants from central Africa. Therefore, before MOOCs can help people in developing countries to become more educated, the infrastructural issues have to be solved (IIE 2014). Concerning the bank account-blindness addressed by Agarwal, it is interesting that the German MOOC platform iversity.org now differentiates between an audit track and a certificate track.
Both tracks offer almost the same, as the content does not differ. By choosing the latter track the learner can receive a certificate from iversity.org. This includes a graded online test and an actual certificate indicating the exact course contents and records about the personal performance in the course. While the audit track is free-of-charge, the certificate track is charged with a fee of 99€ which, depending on the socio-economic status of a learner, might be an unaffordable amount of money creating a high barrier. Regardless of the socio-economic status, the course contents can be accessed for free by everyone who has a computer and a stable internet connection. It is bankaccount blind in this way. But assuming, those who are most in need of further education have a rather low socio-economic status, this practice might also contribute to fostering the access gap.
This practice could equally belong to the access and the usage gap while looking at non-traditional target groups because the course can be accessed without further fees (beside computer and internet) but it cannot be used to achieve a qualification without being charged.
Finally, in the first world, it might make the impression that access is of no importance any more. This is mostly true, but new barriers might occur, like the capability of the available ICT to display graphically advanced and therefore highly demanding content. According to van Dijk ( Even common languages like English, in which most MOOCs are offered, are an insurmountable barrier for the underprivileged which again brings an advantage to those who are able to speak English because of their education. But compared with English, few OER (and thus also MOOCs) are available in languages such as Arab or Swahili which are spoken in countries with a high developing demand (Willems & Bossu, 2012).
Furthermore, the different forms of use and levels of motivation are of relevance. Thus, the German Conference of Rectors (HRK) states in its comment on MOOCs that they are mainly "exploited by those who are already highly committed to their studies, while low achieving students tend to not make use of them" 5 (HRK, 2014, translation by authors).
Reception gap and MOOCs. OER and cMOOCs are online-resources which significantly challenge the users' self-directing capacities in the learning process. Knox (2013) argues that OER will mostly be used in the context of higher education as resources for attending courses or in the form of class records acting as useful, and rather easily made, side products for marketing purposes.
While it is obligatory for the usage of OER within higher education institutions to have support and certification for the resources, these same services are not normally provided outside of higher education institutions 6 . This is especially true for MOOCs which are often criticized for the low-level of support. But MOOCs require a high capacity of media competence and self-regulation skills from the user and will therefore disadvantage weaker learners if the course design did not address this concern.
In summary, a hypothesis can be developed against the theoretical background outlined above, stating that there is a certain possibility that MOOCs are and can be especially exploited by those who already have a higher socioeconomic status and are often better educated. At the same time they also benefit more from MOOCs due to their better preconditions than people with low formal education status. The consequence of this would be that MOOCs contribute to a widening of the educational gap in society rather than helping to close it. Unfortunately, there is currently no sufficient empirical data to prove this thesis. Therefore, as a first step, it will be proved that people with a higher educational level are overrepresented in MOOCs in comparison to the population.

Methodology
The empirical basis for the research of possible usage inequalities of MOOCs is still limited (Spence, 2013). Often only a small amount of user data is collected (Breslow et al., 2013;Ho et al., 2014); therefore, it is difficult to make differentiated statements about the composition of participants. On

Research Results of Adult Education MOOC
The results of the survey show, that, including multiple choice, 27% of the participants declared   The results of these two MOOCs confirm that the proportion of persons with academic degrees in MOOCs is disproportionately higher compared to the population of adult educators. It should be emphasized that both MOOCs were not offered by universities nor did they try to solely attract students with an academic degree. Thus, both MOOCs did not address fields or professions in which it is mandatory to have an academic degree in Germany.

Additional Statistics
The data of the presented MOOCs only give a very limited insight into the diversity of MOOCs. It is shown by Ho et al. (2013), that in many courses most participants who finished the course and received a certificate had a bachelor's degree or above (compared with all subscribed participants). These differences vary, sometimes only by a few percent but they can also show a The high ratio of people having an academic degree is not surprising when looking at the MOOCs organized by universities because the content they offer is aiming at people who are genuinely interested in education. At the same time it can be assumed that universities aim to broaden their target group in a quantitative way rather than to make a qualitatively broader access possible.
Although, whilst it means that universities are not acting in the tradition of OER, MOOCs have the potential and possibility to attract non-traditional learners and might therefore be capable of facing the challenges of the demographic shift in the industrialized countries. Furthermore, an ethical obligation derives from the public responsibility of the universities to make education as accessible for the population as possible. The question of access therefore is not only on the side of the participants. It has several dimensions as shown above, including also organizers and society.

Critical Discussion
The claim of this paper is to gather empirical evidence for the unequal use of MOOCs. Whilst there has been popular discussion on the unequal use of MOOCs, the scientific debate is just beginning.

3.
There is no clear differentiation between the developed world and the third world. This is because the claim "education for all" is meant to address everyone. There are differences between countries, cultures, and people which certainly determine the possibilities in the respective countries but the thesis behind the inequalities remains similar, if not the same, only with qualitative differences. There is still an access gap, no matter if a person does not have a computer, or if the available computer is outdated and will not allow the user to participate in an online learning offer.

4.
The increasing heterogeneity of MOOCs complicates generalizations about their effects. But generally, as a result of their lower learning support, MOOCs make higher demands on self-directed learning, media literacy, and learning motivation. And most disadvantaged people do not have these skills.

Conclusion
The empirical data indicates that MOOCs are mostly used by people who already have a higher educational status. It is not only true for university MOOCs; it seems to have become a common phenomenon which also appears in other MOOCs that are not designed for academics in the first place. A particular reason might be in the courses themselves, which offer content at a level mainly designed for students in higher education. Furthermore, the theoretical considerations on the digital divide provide evidence that barriers exist regarding access, usage, and reception among the potential participants, which keep them from using these offerings properly. This again results in disadvantages for these groups compared to groups that have a higher educational and / or socioeconomic status.
Therefore, based on the knowledge gap theory, it is possible to identify a danger that MOOCs rather contribute to reinforce or expand existing inequalities in education than help to reduce the differences. MOOCs should not solely be understood as marketing instruments for the universities, they should also address the educationally disadvantaged, especially given the consideration of public investment. It is therefore necessary to become active on three different Currently there are insufficient research data to allow a better analysis of the socio-cultural conditions of MOOC participants. In particular, the large MOOC providers do not provide detailed data of participants for research. But projects like MOOC Knowledge 7 will systematically collect sufficient data of MOOC participants that can be useful to validate the first results.