Open science practices in higher education: Discussion of survey results from research and teaching staff in Germany

Aspects of open science and scholarly practices are often discussed with a focus on research and research dissemination processes. There is currently less discussion on open science and its influence on learning and teaching in higher education, and reversely. This paper discusses open science in relation to educational practices and resources and reports on a study to investigate current educational practices from the perspective of open science. We argue that offering students opportunities via open educational practices raises their awareness of


Introduction
Open science and open education are strongly connected through the concept of 'openness', but they approach this concept from different perspectives: Open sciencehere we mean as well open research, i.e. referring to natural sciences, social sciences and humanities likewise -mostly refers to research and researchers as well as aspects connected to the scientific enterprise such as scientific communities, publications, and research impact (Bartling & Friesike, 2014a;Herb, 2015). In contrast, open education is concerned with open strategies and approaches to learning and teaching in various settings like for example schools, higher education, vocational education, informal learning. Open educational resources is a key element of open education and well explored by the literature in this area (Hylén, van Damme, Mulder, & D'Antoni, 2012).
Most research discusses openness in either science or education and lacks the interrelation of both fields, specifically in higher education environments where a large number of employees are concerned with both, research as well as learning and teaching. Moreover, discussions within both movements, openness in science and education, mostly We conducted a study, which used a quantitative online survey to ask academic staff teaching at higher education institutions in Germany about their use of digital media, tools and OER, and their teaching practices. Results consider practices of teaching staff including resources, technologies and activities that relate to currently discussed aspects of open science. The leading research question is:  Which open science related practices are currently applied in German higher education?
In this paper, we briefly introduce aspects of open science and open education before we lay out opportunities of open practices. Afterwards, we report on our quantitative survey that provides a view on current practices of teachers. We summarize the survey findings and compare them to similar studies before we draw conclusions.

Aspects of Open Science
Open science or open research stands for a movement which suggests openness in all phases of the research lifecycle (European Union, 2016;Förstner, Hagedorn, Koltzenburg, Kubke, & Mietchen, 2011). It considers not only the use of new technologies in areas like content access, shared ideas and collaboration, but advances further discourses, some of which stand for a radical change in research behaviour, like open peer review (Ross-Hellauer, 2017), open grant writing and open evaluation. As such, in open science researchers move from publishing as early as possible to sharing as early as possible (European Union, 2016). Some researchers even talk of a second "scientific revolution" (Bartling & Friesike, 2014b;Friesike, Widenmayer, Gassmann, & Schildhauer, 2015;Nielsen, 2013). Researchers (Fecher, Friesike, Peters, & Wagner, 2017;Fender, 2015). There seems to be a tendency towards open access publishing (Bosman & Kramer, 2018) suggests that any open data should be findable, accessible, interoperable and reusable (Wilkinson et al., 2016). Its concrete application is relevant for data producers such as researchers and metadata editors and for infrastructure developers that give access to this data.
Larger bibliometric studies (Piwowar et al., 2018) analyse open access publishing and show a positive trend -however there are still great differences among diverse disciplines (Bambey, 2016).
Other recent studies aim at finding explicit explanations for researchers open practice behaviour. such as the study by Moksness and Olsen that shows attitudes and social norms as predictor for publishing open access (Moksness & Olsen, 2017). Other surveys showed that external factors like a researcher's institution or their personality influence the adoption of sharing one's research data openly (Kim & Nah, 2018;Kim & Stanton, 2016;Linek, Fecher, Friesike, & Hebing, 2017). Moreover, researchers define "openness" in different ways, which influence their practices (Levin, Leonelli, Weckowska, Castle, & Dupré, 2016), specifically in relation to their research impact in society and good research practices guaranteeing research quality (Grubb, Easterbrook, & Biondi-Zoccai, 2011). However, some other studies show that some researchers are still sceptical of sharing their data (Blahous et al., 2015).

Aspects of Open Education
Open education shall decrease learning inequalities and support lifelong learning (Blessinger & Bliss, 2016a;UNESCO, 2012). A core element of open education is open educational resources (OER). There is a common understanding of the nature of open educational resources -although there might be some disagreement on best practices and types of licensing to adopt. OER are educational resources and materials that users are able to retain, reuse, revise, remix and redistribute (Wiley, Bliss, & McEwen, 2014).
OER include all kinds of educational resources, including learning material, tools and software. "Access is fundamental to open education. [However] Open education goes beyond access" (Blessinger & Bliss, 2016a, pp. 13-14), practices need to include "the construction of new pedagogies and learning activities" (Kaatrakoski, Littlejohn, & Hood, 2016). Increasing the use of OER and at the same time adapting open pedagogies leads to an increase in open educational practices (Albion, Jones, Jones, & Campbell, 2017;Ehlers & Stracke, 2012) and fosters open education. Cronin (Cronin, 2017)  Studies on open education practices focus on applying OER (Bossu, Brown, & Bull, 2013;Boston Consulting Group, 2013a) or open textbooks (Seaman & Seaman, 2018), or discuss any influencing factors like policies and their potential to foster OER use and creation (Bossu & Stagg, 2018;Cox & Trotter, 2016). Researchers see potential in current initiatives, but see a need for improvements (Stagg & Bossu, 2016;Udas, Partridge, & Stagg, 2016). Kaatrakoski, Littlejohn and Hood (2016) still see tensions in practices between individual's needs and institutional policies, educators' amount of teaching responsibility and institutional accountability, and cost efficiency and learning objectives. In her qualitative study, Cronin describes four levels which educators can be distinguished with regard to their open practices: macro (will I share openly?), meso (who will I share with?), micro (who will I share as), and nano (will I share this) (Cronin, 2017). Cronin states that educators are influenced in adopting open practices by diverse factors such as the use and creation of OER that has a positive influence (compare (Wiley, 2015)). Reversely, open practices like networking foster the awareness or OER (Cronin, 2017). Stagg (2014) discusses not only open educational resources use, but practices like enabling an open environment for students (discussion options, options to share ideas and one's work), and formal credit, meaning that students' open behaviour find its way into the formal assessment process. With regard to open pedagogies, research discusses concepts of research-oriented learning, with forms of openness referring levels of student autonomy (Brew, 2013;Heck & Heudorfer, 2018). In this study, our understanding of practices refers to activities, behaviours and attitudes of teachers that contribute to more or less open learning and teaching environments, similar to prior discussions (Stagg, 2014;Väänänen & Peltonen, 2016), with the aim to get first insights into teachers' practices and their levels of openness.

Adopting Open Practices in Teaching and Learning
There seems to be several commonalities between open practices in science and in education. One of these similarities is presented by Väänänen and Peltonen (2016)

Survey on Open Practices
We conducted an online survey to investigate the status openness in higher education

Method
This is an explorative study that aimed at questioning current issues and ideas to implement open science practices in education. The target participants were any academic, professional and researcher with teaching responsibility at German higher education institutions, including universities and universities of applied sciences. As higher education systems and educational roles differ globally, we did not aim at designing this survey to be used internationally. However, we think that the design of the contextual questions (in contract to demographic questions) is adaptable and a comparative study in other countries would be beneficial.
Regarding our study, we aimed at doing a purposive sampling and involving people from current groups and communities that engage in discussions and activities about open science and open education. To reach them we sent the survey to diverse institutionalinternal and external mailing lists and via personal contacts. We also included mailing lists that were discipline-based, derived from higher education and higher education didactic communities as well as lists from open science, Science 2.0 and open educational resources communities. Additionally, personal e-mails were sent to presidents and contact persons from those communities, and Twitter was used to disseminate the survey.
We collected data anonymously and survey participation was voluntary. Thus, we did not seek approval by an ethics committee. Potential participants were informed about the study, data usage and its goals on the online survey landing page. They were informed that they give consent for their anonymous data being used for scientific purposes when starting the online survey. The survey was online from February 6 to March 3, 2017.
The survey structure and data is openly available (Heck et al., 2017b;Heck et al., 2017a).
It includes 20 topical questions which were separated into five major topics: demographics (4 questions), material used in courses (4 questions), open educational resources awareness, usage and development (6 questions), collaborative tools used in courses (1 question), assessment and participation options (5 questions). The question types differed, with mostly single choice questions, multiple choice where applicable (choice of applied tools), and 5-point-likert scale when participants had to rate the importance of resource characteristics (Fig 2). We offered a comment field when participants clicked the NO-answer and at the end of the survey. As well, participants had the option to add additional answers, e.g. tools they use that we did not list.
Questions on OER regard use and creation of OER and reasons for this behaviour. Data from earlier studies revealed that academics were confused about the proper definition of OER. Some seemed to understand OER as free resources, or only refer to open source software (Seaman & Seaman, 2018). Other studies (Seaman & Seaman, 2018) decided to give a broad explanation of OER, avoiding details to not tempt the participant to claim "awareness". However, there is a danger of having a bias when giving an explanation.
We decided not to give an explanation to participants about the definition of OER, but keep this question simple. We assume that either someone knows about OER or not. If they had not heard of the term before, they do not properly use OER (at least not consciously) or create them.
Demographic questions asked about the current professional position, the discipline, birth year and gender. The classification of research disciplines was adapted to general disciplines at German higher education institutions without any sub-classes. The job position classification refers to common positions in Germany: Professor (all with German professor title, includes associate, full and affiliate professors), special education teacher (staff with specific teaching responsibility like teaching literacy skills), academic (staff with research and teaching responsibility), lecturer (with teaching responsibility only), student assistant (supports teaching and research).
We used SPSS (v23) for statistical analysis, and provide descriptive analysis for all variables. We got 360 responses, whereof 210 were completes and 150 incompletes.
Results are based on the 210 complete cases. Significance tests (Chi-Square) considering the job position were done with 207 cases, where we left out two student assistants (not representative for group) and one case with an unclear job position. Two researchers analysed and checked open text questions. We show the most relevant results on specific questions in tables and figures below and discuss them in the subsequent section.

Limitations of the Survey
Using self-selective online sampling and a purposive dissemination of the questionnaire (Creswell, 2013) (Table 1), whereas the Arts and Humanities discipline is overrepresented. Despite this, we think our explorative study gives critical insights into the status of openness in higher education in Germany, with implications for further research in other countries. Table 1 summarizes the demographic data from 210 cases. The majority of participants was about 40 years old. Please note that this field had to invalid entries. Participants had a multiple choice option for their discipline and some felt to belong to two disciplines, i.e. n is larger than 210 cases. The Art and Humanities group is slightly overrepresented which might be due to the mailing lists where we promoted our survey.  Figure 2 shows the boxplots for the question on criteria considered for resource choice.

Results
The boxplots and the means (Table 2) show that all criteria are important for the participants, with means a rated value less than three (1 = very important). Currency of material and ease of use are the most important criteria for selecting resources for teaching, with also the lowest standard deviation. Open licenses are least important, with a high standard deviation. Table 3   Regarding collaborative tools used in courses, we asked the participants to distinguish if they use tools only for the provision of course resources, only for communication and collaboration between lecturers and students, or for both of the pre-mentioned tasks.
Participants had the option to state that they do not use any tool. Distinguishing between usage and non-usage, most participants used two collaborative tools (Fig 3). The tools used most often (Fig 4)    There is a tendency that professors and academics use "traditional academic tools" (such as reference management tools) more often than special education teachers and lecturers.
Special education teachers and lecturers tend to use non-academic tools like blogs (over 23 % compared to less than 15 % for both academics and professors) and editing tools like Google Docs (over 36 % compared to 26 % for academics). One reason might be that not all lecturers and special education teachers have access to academic tools (e.g., some fourth ranks. We found a tendency that lecturers and special education teachers use tools like blogs and Google Docs more often.

Fig 5. Tools used distinguishing between purposes.
The last part of the survey investigated questions around student participation, sharing and assessment, i.e. aspects mentioned with regard to open educational practices and pedagogy (Table 4). Although academics are the largest group supporting resource sharing, they do not explicitly require it from their students. Contrary, there are exactly twice as many professors who do require in-course sharing than those who only offer sharing options. Require in-course sharing from students was the most popular answer for all job positions except with academics. We asked if participants assess students sharing, that is if students' grading is dependent on sharing materials. Professors, who require sharing in their teaching more often, also assess students' sharing activities (48 %). Over one-third of special education teachers assess sharing, within the lecturers and academics group it is less than 25 %. In addition, 68 % of the participants stated that they offer opportunities for students to co-create and determine course content ( Table 5). The behaviour significantly correlates with the use of OER (χ² (1) = 7.07, p < .01, N = 210), although not with its creation.
Participants that opt for student co-creation said that most of the course content is predetermined with options to consider students' interests (50 %) or that the course basics are predetermined, but specific foci are determined together with students (43 %). Only 6 % of the participants opt for a more radical answer stating that course content derives out of discussions and determinations together with students during a running course.
Here, special education teachers and lecturers were more likely to choose the latter version, being 13 % and 12 % compared to less than 4 % for academics and professors.

Discussion
In the following, we discuss the highlight-findings, grouped in 1) open educational resources, 2) tools and activities, and draw upon challenges and opportunities for open practices.

Open Educational Resources: Awareness and Use
Overall, our findings match the results on open resources use that were reported earlier (Bossu et al., 2013;Boston Consulting Group, 2013b;Seaman & Seaman, 2018 Half of our respondents who know open resources also produce and share them ( Table   3). Those that do not produce them find it both too laborious and time-consuming, or they do not know how to do that. This confirms the reasons for not producing open resources as found by (2013), although this applies to less respondents from our sample.
Our answers suggest that the digitality of learning and teaching material, that provides options for easy access and distribution, is considered a more important aspect than openness. This may be an indication that the concept of open resources and the properties belonging to it are not fully understood or not valued equally. As proof for the latter we can look at the drivers behind resource selection. Although 77 % of the respondents know the license which determines a resource's options for reuse, the selection of teaching and learning resources is mainly driven by their currency and ease of access as well as by their relevance for the topic taught and their quality (Fig 2, Table 2). Open licenses, on the other hand, are neglected by the majority of respondents when choosing learning and teaching material. Respondents even stated: "Quality of content is key: whether I have to pay for it or not" and "negligible in as much as students have access".
There is a substantial number of respondents who do not use open resources (60 %) because they are either not aware of them or do not know what they mean, although we have not explicitly asked about the latter. These seem to be common problems faced by the OER movement, as demonstrated by other similar studies (Bossu et al., 2013). This finding is remarkable, though, despite major efforts from a range of stakeholders, national and international, to increase awareness and to provide access to open resources and promote activities via large initiatives. This still remains an considerable issue to be addressed, and perhaps one way to address this problem would be to increase capacity building and training of university teaching staff could.

Open Tools and Activities
It seems that there is a tendency that professors and academics prefer established academic tools such as literature reference tools. Special education teachers and lecturers tend to use non-traditional academic tools such as editing pads or open wikis.
One reason may be that the latter group do not have proper access to academic tools, for example because of license restrictions. Despite this tendency, email is still the tool most used (Fig. 4 and 5).
Results may be influenced by how the survey questions were formulated, though. We asked participants to state the current tool usage and did not ask them to state whether they have ever used those tools, whether they just do not know them, or whether they have particular reasons to not use them. It would be interesting to study whether their choice of tools is influenced by external factors -like institutional regulations, restricted options in designing a course -or if participants did test diverse tools in the past and came up with their personal favourites as a matter of best practices that also fit the current educational environment best. More research has to be done considering the teaching staff's opinion on and choice of good teaching practices and use of open technologies and pedagogy, specifically with regard to their specific educational contexts.
We were also interested in which ways teaching staff integrates and fosters open practices in education and what serves as incentives. We assumed that teachers do not feel too comfortable with using the technology (2008) and expected a conflict to occur between openness, collaboration, and assessment in class (2012). Our study reveals similar results regarding the use of tools defined as open Web 2.0 tools. Only 33% do use them.
However, over half of the respondents require students to share their works using any kind of digital technology (Table 4). We asked them, in which form students should share their work. One of ten respondents said they engage students to share work openly on the web. Brown's (2012) study revealed that some academics have difficulties in finding "an appropriate balance between assessment and student collaboration via Web 2.0" (Brown, 2012, p. 56 (Piwowar et al., 2018). If teaching and learning becomes more open and offers ways for students to access content, to participate and to co-create, this fosters a way towards openness in research, i.e. research that opens its community for students and is able to raise awareness of those critical issues beyond internal borders.
Our study shows a relation between the use of OER and supporting student co-creation. with it open practices among students, like co-creation of course content.
Regarding the latter aspects through an openness lens, we also must distinguish between levels of collaboration regarding diverse study and course forms. To teach courses with a high openness, like high levels of co-creation and communication options, might overstrain early semester students, whereas more experienced students in their Masters can benefit from those. Survey participants stated that the level of student experiences and skills influence their practices.  Although not focus of our study, we would like to make the point that where the interrelation and potential fruitful coaction between research and education become obvious. Pedagogical concepts of research-oriented learning focus on students as researchers and teaching research skills (Brew, 2013). "Learning through research" aims at letting students participate and engage in a research process. They need the opportunity to formulate research questions and co-design and reflect on research aspects (Reinmann, 2016). Aspects like student engagement and participation discussed within concepts of research-oriented learning are similar to those discussed within open science and education and would easily complement each other (Heck & Heudorfer, 2018). The open education concept emphasizes the importance of students being allowed to actively participate in the scientific community to understand what research is about and to apply this knowledge in their studies. "Indeed, one of the goals of open education is to move learners closer to the centre of a community of practice, specifically through providing opportunities and infrastructure for participation and collaboration" (Blessinger & Bliss, 2016b, p. 14). Brown (2012) emphasizes the high potential to build a bridge between teaching and current research, it allows students to become a member of a "knowledge creating collective" (Brown, 2012, p. 56), where they benefit from and contribute to the research community.

Conclusion
We discussed the interrelatedness between open practices in education and science and claimed that open science need to be fostered by educational practices that refer to goals in open science. We conducted an online survey to shed light on the status of those practices in German higher education institutions.
Our results point out that open practices have not yet been fully achieved in higher education. Open resources are not popular, and prevailing email as a digital teaching tool does not contribute to open practices that foster a community awareness and belonging.
Respondents undertake activities related to openness like encourage students to share their content and be co-creators of resources, but those activities are not common place.
Here, we still see challenges in bringing open practices and existing higher education practices together. However, independently from our aim to relate practices in science and education, we need to investigate what benefits and learning outcomes open practices in context of science and education will have. In addition, answers showed that teaching is very diverse and has different needs depending on the form and discipline of teaching.