Studying in a foreign language: Study performance and experiences of German students at a Dutch university

1.1 Academic performance of German students in a Dutch university

In this introductory section, we report on the academic performance of German students at our university in relation to proficiency in Dutch as a Second Language to provide quantitative evidence on the differences between Dutch and German students.

The study load for each year of study is expressed in European Credit Transfer and Accumulation System (ECTS) credits. Students who obtain the full complement of 60 ECTS credits are deemed to have completed the first year of the Bachelor (BA) programme. University reports show that study results of German students are lower than those of Dutch students. Results of recent research amongst the same group in the 2016 cohort by De Vos (2019) show the same trend. Figure 1 shows the percentages and numbers who passed their first year (with 60 credits) between 2011 and 2017. There was no obvious pattern of increasing or decreasing pass rates over the studied period.

Figure 1:

Percentages of Dutch and German students in 2011–2017 who received a positive BSA after one year (Source: Rapport Bindend Studie Adviezen, Radboud University)

BSA Binding study advice: positive when students have enough (at least 40) ECTS credits, negative when the number of ECTS credits is too low, in which event students are forced to stop.

In our previous study (Zijlmans et al. 2016), 36% of our sample from the 2014 cohort did not proceed through the first year. To enrol at the university, the German students of the intensive preparatory course in Dutch must pass a test of their Dutch language skills (DFL test) at the B2 level of the CEFR (Council of Europe 2001). This assessment consists of four subtests on writing, speaking, reading and listening developed by the University Language Center (ULC) in line with the State Examination in Dutch as a Foreign Language. Developing this set of tests by the ULC was necessary because the results of the State Examination would not be available before the start of the academic year.

In Zijlmans et al. (2016), we correlated the DFL test scores of 41 German students (2014 cohort) with their academic performances in their first year of study, in terms of the ECTS credits obtained and the grade point averages (GPA). The results of the DFL test correlated with the number of ECTS credits achieved in their first two terms (r = 0.459; p < 0.01; N = 41). Correlations were even stronger with their GPA in the first two terms (r = 0.625 and p < 0.01) and after one year (r = 0.502 and p < 0.01; N = 41). Our hypothesis that language played a role in the relatively high dropout rate was supported.

We repeated these calculations for the 2015–2016 cohort. This time 57 of the 151 German students on the 2015 summer course agreed to share information on their academic performance. Only seven students (12%) achieved enough ECTS credits by the middle of the year, only 12 (21.8%) the full 60 credits by the end of the year.

These results imply a substantial number of resits during the year and over the summer. Figure 1 shows the numbers of students who had obtained sufficient ECTS to transfer to the second academic year.

We found the same trend on the role of language for this cohort as we did in 2014–2015. There was a moderate but positive significant correlation between the DFL test scores and the ECTS credits at mid-year (r = 0.375; p < 0.01; N = 57), as well as a positive and higher correlation between scores on the DFL test and the GPA (r = 0.430; p < 0.01; N = 57). This result again confirms that DSL proficiency plays a role in academic success.

Another way to gauge the role of FL language proficiency is to use data on English proficiency for the German students. Dutch universities increasingly offer English mediated programmes (English-Medium Instruction, EMI). These programmes are very popular amongst German students nowadays. The students in our study took part in Dutch programmes, but English is nevertheless prominent in reading materials and international staff may teach in English. The level of English as a foreign language (EFL hereafter) may play a role in their study performance as well. We did not have formal information on our study participants’ EFL proficiency. When students have a European secondary school diploma with English as subject of their final examination, they are exempted from providing an English language certificate.^[1] It is assumed that their EFL is at the B2 level or higher. We measured English proficiency of the participating German students using the OOPT (Oxford Online Placement test, a standardized placement test from Oxford University Press). We found that 23% scored B1 or lower. We identified a significant positive correlation between EFL and GPA at mid-year (r = 0.333; p < 0.05).

Finally, we see an argument for language problems in the reports from the writing lab that substantial amounts of German students make use of this service. In 2015, the percentage of new German students was 5% (236/4067). The percentage of German students using the writing lab was 11% (117/1040). The overrepresentation of German students may indicate that they experience problems or insecurities in their writing performance.

1.2 Research questions

We investigated the academic success of FL German students, concentrating on FL proficiency. The analysis in Section 1.1 helps us to formulate our questions for the present research.

Learning Dutch is relatively easy for German people due to the small linguistic distance between the two languages (Schepens et al. 2016). In an intensive summer course, students manage to advance from zero to a CEFR level of B2 in a little less than five weeks. In 2015, this was the case for 80% of the participants. Despite this, German students seem to experience problems. A successful student, even in a Dutch-instructed programme, additionally needs a high level of proficiency in (academic) EFL as the relevant literature is in English and international staff members may teach in English. In our study, we therefore cannot ignore the role and impact of the level in EFL when assessing the academic performance of German students. The question arises to what extent academic performance and FL proficiency are related. The specific questions we want to answer are:

1. Do German students encounter language problems in their starting year at our university?

2. Which problems do German students encounter in their starting year at our university, due to the fact that they have to study in languages that are not their native language?

We sought to answer these questions more thoroughly by interviewing staff and students. Do they experience discrepancies between the requirements for admission and the actual requirements for study activities, and do these discrepancies affect academic performance as other studies (Deygers 2017; Meyer et al. 2013) suggested? Admission tests may not reflect the true requirements for academic tasks (Deygers 2017).

We will explain our methods of research in the following section. In Section 3, we will assemble the results, in factors, processes and motives experienced that influence study performance. A general discussion and conclusion follow in Sections 4 and 5.

2.1 Participants

We approached relevant stakeholders who could provide us with information covering all aspects of the research questions (Doorewaard et al. 2015). We collected information from three parties: the students themselves, their teachers, and members of the academic advisors. We contacted teachers and academic advisors for an interview by e-mail, in which the goal of the study was explained. All of them work in the two faculties where the majority of the German students are, the Faculty of Social Sciences (17% German first year students in 2015–16) and the Faculty of Natural Sciences (5% German first year students in 2015–16). We approached all German students who took part in the DFL prep course in the summer of 2015. In a letter the goal of the study and the required effort was explained. A consent form was attached. Students gave permission to obtain their study results (see Section 1.1), and/or indicated that they were willing to take part in interviews after the first semester and at the end of the first year. After transcribing the interviews, all participants’ names were replaced by reference codes.

Students can turn to academic advisors at the Student Information Desk for help with a variety of issues, including falling behind with their studies, doubts about their chosen subject or withdrawal. We selected two academic advisors in faculties with the largest numbers of German students, one from the Faculty of Social Sciences (FSS) and one from the Faculty of Natural Sciences (FNS). Quotes taken from these interviews are referred to as “Staff” in Section 3.

We approached teachers in the two faculties who were highly involved with German students. We contacted these teachers and asked for an interview. This resulted in positive responses from two teachers in the FSS and one teacher in the FNS. The respondents were mentors in addition to lecturers, and as such well-informed about the challenges new students face. As mentors, they conduct interviews with each student during their first year and discuss whether the students feel settled in their education and in the city, and whether the training they have received matches their interests, expectations and study skills. They also talk about the students’ grades and whether they are expected to pass the first year of the BA programme. Quotes from our interviews with these mentors/teachers are referred to with a “T”.

The start of our study was a pilot study in which six German students were interviewed. We integrated the findings of these students in our research.

When we sent out the invitation for the interview to the 47 students who had consented to take part, six students responded for the first interview at mid-year, and four also responded to the call for the second interview after the completion of their first year of study. We found out that five students had already dropped out by the middle of the year, while only seven had obtained the maximum number of 30 ECTS credits at that time. We did not succeed in interviewing students who had actually dropped out. After we discussed this with the teachers one of them put us in contact with two additional students who had struggled through the first year. We spoke to these two students once, at the start of their second year.

Ten of the 14 interviewed students were from the FSS, two were from the FNS, and two from other faculties. Quotes from the student interviews are marked with an “S” in the results section. Table 1 gives an overview of participants and interviews.

2.2 Construction of interview guidelines

After the pilot interviews, we continued the project with two one-on-one conversations with general informants from the FSS and FNS academic advisors on the academic progress of the German students. As researchers in FL acquisition, we were interested in the FL proficiency of the students, in DFL and the role of EFL. The central aim was to determine the explanation for the higher dropout rate among German students. Our questions were prompts for open conversations at this stage. We then constructed interview guides comprising major and follow-up questions (Rubin and Rubin 2011) on the topics submitted by the counsellors for the interviews with teaching staff and students. We also included a brief report of our previous quantitative study (Zijlmans et al. 2016) in the introduction to these interviews, as part of the explanation of the goals of the present study. The questions for the teacher interviews were related to the productive tasks in FL. For this, we translated the staff interview guides concerning academic writing by Leki (2007).

The interview guides for the first series of student interviews, held in the middle of the year, started with questions on their motivation for studying in the Netherlands followed by questions about their DFL proficiency in general and in each of the four language skills. In recognition of the importance of English, we added some questions on the role of this additional FL. There were questions on their personal experiences, such as housing, extracurricular activities and friends. Finally, there were questions on possible differences between academic life in Germany and in the Netherlands. Students were also asked if they knew fellow students who had had problems. The interview guide for the second set of student interviews, held after they had completed year one of their studies, contained quotes taken from the interviews with counselling and teaching staff. The interview protocols are provided as appendices to this paper.

2.3 Analysis of interviews

The interviews were recorded with the consent of the participants. An assistant transcribed them almost verbatim; only grammatically incorrect sentences were corrected, and hesitations, fillers and the like were left out. For the interviews of the pilot study, elaborate summaries were available. The first author checked the accuracy of each transcript. The analysis took place with the help of the software programme Atlas.ti in three steps (cf. Doorewaard et al. 2015). In the first step, quotes were marked and general keywords were assigned such as “reason to study in the Netherlands”, specific keywords on FL proficiency such as “language (not) sufficient at the start”, “(no) problems listening to lectures” and “(no) problems writing in the FL”. In the second step, frequency counts were conducted for all topics. This resulted in the identification of the most deeply discussed topics across all interviews. Our focus was on DFL proficiency, but other recurring topics in the interviews were what is called, in the literature on the study success of international students, “academic integration” or “academic adjustment” (Klatter-Folmer and Weltens 2017; Morrison et al. 2005; Rienties et al. 2012). We divided the frequently recurring topics into two groups, the first group being the topics related to FL proficiency, the central issue; the second group the more general topics on academic adjustment (Table 2). The interviewer brought up topics, but whether this resulted in many comments depended on the individual interviewees. Some topics were hardly discussed, such as “have you ever felt German students are an extra burden for you or the course?” (Leki 2007) and “do you think German students/you need more time than domestic students?”

The third step in the analysis was to summarize and interpret the information for each topic and between topics in search of the factors, processes and motives influencing the academic performance of the FL German students

3.1 Topics concerning the role of FL proficiency

Teachers and students alike reported that the DSL prep course was merely a basis and that the students had to continue learning DSL for study purposes to achieve the required level (all quotations are translated from the original Dutch):

S8: The real language skills I needed only came to life when I started studying.

T1: In the first year, they have the most communication problems; in the second year, they are reasonably okay, I think.

We asked questions in relation to the four traditional skills of reading, listening, speaking and writing. Reading in the FL did not seem to be problematic. None of the students claimed to have specific problems with the multiple choice questions, a topic brought up by staff members, although two students said they were not familiar with this form of assessment. One student explicitly mentioned she could derive the meaning of Dutch words by comparing them with German words. Teachers did not put forward reading skills as the source of problems. Four major language-related issues emerged, which we will report in the following subsections: listening and speaking in lectures and discussion groups (3.1.1), writing for examinations and assignments (3.1.2), the impact of language proficiency on grades for written assignments (3.1.3) and the use of English as an additional language (3.1.4).

3.2 General topics

Students and teachers mentioned the necessity to work systematically and hard as the important reason for early drop out. This is within the first few months. This appeared in 11 interviews. One of the interviewed students explained:

S13b: It turned out that many students have, eh, stopped their studies, so in the end I was the only [German student] left. … I think most [German] students I was with in the beginning; they stopped in the first semester, before January, something like that.

The students said the workload is “overwhelming” (S13b) in comparison with what their friends at German universities experience. An indicator of the difficult start was the additional number of resits taken by German students. This was an issue in six interviews. The results figures confirm this picture. For the 57 German students included in the study, the modal number of ECTS credits obtained by mid-year was 17; only seven had achieved the full 30 credits.

Although we only interviewed students who had not dropped out, most of them indicated that the first year had not been easy. They experienced it as being extremely stressful. One of the students expressed this situation for fellow German students in the following way:

S3a [about friends]: They said they really could not study more. They said, “I do so much already and I really can’t do more now. Then I really do not feel happy anymore, and my grades are too low. … If I have to resit everything and take more examinations, I simply could not cope.” Therefore, they stopped.

Students had to get used to different ways of teaching and assessing. In the section on speaking in discussion groups, one sign of this phenomenon was the observation of the teachers that students were reluctant to take part in the discussions, as they felt uneasy with this way of teaching. When asked whether reading in exams was a problem, two students declared they had no experience with multiple-choice questions in their earlier educational career.

One staff member introduced the topic of study habits and skills, which was confirmed in the teacher interviews; staff and teachers do not always consider the German students’ expectations of what they have to learn to be appropriate. They observe a mismatch between their study habits and the way academic knowledge is tested at a Dutch university:

T2: [For the German students] it is very much about facts; this is the concept and that is its meaning. They buy hundreds of cards, write the concept and its meaning on them, and they practice it. That is a lot of knowledge and a little bit of what we are testing. We want insights, we want them to apply their learning, and we want to test a lot more than just that factual knowledge.

One student noted that the Dutch grading system was not completely transparent for her at the start. In fact, 10 and 9 (the top grades in a 10-point scale) are hardly ever given, and an eight is considered “very good”. She was not aware of this and was discouraged when she received grade 8 or even 9. This caused extra stress, as it made her work harder.

When they are having problems, German students do not easily turn to staff members for additional help. When staff and teachers spoke of this, they attributed it to the perceived hierarchical relationship between staff and students in Germany. The Dutch system is less hierarchical. Faculties offered workshops to raise awareness about this, but have stopped running them because of a lack of interest. All the interviewed students stated that they had never heard of such workshops. One teacher mentioned the lack of interactivity as a possible cause for the distance that the German students maintain from the staff, which also restricts them from overcoming the language barrier:

T1: They are very quiet and do not ask questions; they just do what they have to do. All German students are very quiet and do not speak much to the teachers, so they do not ask much. I also think that perhaps those who I see as not so good might indeed be able to perform well if they would seek more contact and ask more questions, which I think they do not dare to do because of the language barrier.

The views of the interviewed students about seeking help were diverse; five students affirmed that they do not easily ask for help or that it is not the German nature to ask for help. Two of them stated that they would do so when necessary. On the other hand, six students explicitly mentioned that they appreciated the approachability of teachers and staff.

S12a: We Germans, somehow, we are like, “That is my own problem, I will do it myself,” and then we act like that too. We do not need that much [help from] staff.

4.1 FL problems

The small linguistic distance between the Dutch and German language is probably helpful to the students’ reading and listening skills. Their reading in Dutch does not seem problematic, nor is reading in EFL. PowerPoint presentations support listening to lectures. In speaking, FL users draw more on pragmatic skills than on linguistic proficiency to resolve disruptions in communication (Björkman 2013). The small linguistic distance also makes it relatively easy for German students to make themselves understood in Dutch, even when they use “Dutchified” German words or speak with a German accent.

Writing skills were by far the most important theme in the interviews. All German students had passed Dutch tests at the required CEFR B2 level before starting their studies; however, the experiences of the subject teachers in our study seem to suggest that the assessed level does not match the language skills required for academic tasks. Deygers (2017) also concluded this in his study. There seems to be an imbalance between the ideas of the language experts who develop and judge the FL tests and the requirements of the subject teachers. The test developers do incorporate the content aspect in the correction models of the assessments and award this with 24 points out of 70 (34.3% of the total possible score: CvTE 2016). The score on linguistic features adds up to 46 points (65.7%), and so linguistic features play a very important role. The reverse appears to be the case in the correction models for the study assignments: content is weighted at 80% of the final grade, form at 20%. However, if either component is assessed with a five or less, the final grade cannot be higher than five. Teachers intend to judge written products largely on content, with less of a focus on language proficiency; however, they are still hindered by syntactic and lexical errors. The teachers say that they ignore errors that do not interfere with the comprehensibility of the text, whereas major deficiencies in accuracy, such as serious deviations in spelling, meaning or grammatical form (Kuiken and Vedder 2012), will unavoidably influence their rating, as they cannot be clear of the student’s understanding of the subject. Teachers often return the text without constructive feedback, simply assigning “not sufficient” or low grades that require the student to re-write the assignment. Teachers refer students to additional DFL courses; however, students often ignore this assistance because they are not aware of their own deficiencies. Students feel that expressing their thoughts properly in the FL and the use of the different academic writing style are the most important obstacles for them and therefore turn to the writing lab for assistance. However, the tutors are not DFL experts and give no feedback on FL deficiencies.

German students nowadays prefer to attend the EMI tracks but, even in the Dutch mediated track (or bilingual track as it is called at present), they use English as an escape route and hand in their assignments in English. However, we have shown that a great number of German students also show deficiencies in their EFL skills: 23% of the German students in our study scored at or under the B1 level in the OOPT, while the study tasks are set at the higher B2–C1 skill level. The differences with domestic students in academic performance as described in Section 1.1 still existed for the 2016 cohort (De Vos 2019).

We are not the first to question the pass level in the admittance tests (see for example Deygers and Hamnes Carlsen 2014). It is generally assumed that international students will develop FL to a C1 level soon enough in their daily exposure to and experience in the academic register. We conclude that this is not true for all students. Baumgarten (2014) attributed this to the fact that most interactions within and outside the university are in the students’ L1. As there are many German students on our campus, this is the case for this group.

The B2 level should be sufficient for successful academic performance, based on the descriptors of the CEFR. The problem is that grammatical accuracy in the CEFR is only given in very general terms. For various languages, professionals started projects defining the CEFR levels of linguistic competence into Reference Level Descriptions (RLDs) (Hulstijn 2015); however, for DFL, there is no RLD that can be used by test developers.

We should also question the assessed tasks themselves. Deygers (2017) shows that various tasks in the DFL admission tests used at Belgian universities were never required during the course of study. Leki (2007) concluded that the tasks in the FL writing classes did not match the writing tasks in the study programme.

4.2 Problems with academic adjustment and its effect on FL development

We questioned possible problems due to FL proficiency, but the other recurring factor mentioned in the literature is the role of the academic and social integration of international students (Klatter-Folmer and Weltens 2017; Morrison et al. 2005; Rienties et al. 2012). Importantly, social contact and integration is directly relevant from a FL perspective as they foster the learning of FLs (Deygers 2017) and enhance confidence in using the language, which in turn supports further integration (An and Chiang 2015; Ferenz 2005; Hsieh 2009). Cultural differences can affect academic integration. For example, on the dimension uncertainty avoidance, which has to do with anxiety and distrust in the face of the unknown (Hofstede n.d.), Germany and the Netherlands score 65 and 53 points respectively (Claes and Gerritsen 2011). This suggests that Germans have a higher uncertainty avoidance and try to prevent this feeling. Here, we came across the German students’ unfamiliarity with teaching via discussion groups that hinders them from actively participating. Differences in the power distance, the different hierarchical levels of society, in the two countries (Claes and Gerritsen 2011) could also cause this attitude. The low power distance in the academic community of the Netherlands, where professors and students deal with each other in an informal way, seems to cause unease to the German students. As disciplinary discussions create an environment in which students develop and practice their academic language (Ferenz 2005), this attitude negatively influences the acquisition of academic disciplinary language when students maintain a high barrier between themselves and their professors. Having to work hard, unfamiliarity with the Dutch grading system and a focus on facts can cause unnecessarily low grades. Finally, reluctance to approach teachers and student advisors for help is another self-limiting academic behaviour, possibly caused by the above cultural differences.

We do not want to dismiss low levels of integration as a possible cause for negative academic performance; however, the previously reported correlations between EC and GPA scores were relatively low (Rienties et al. 2012). The high numbers of German students on campus means they need not feel lonely or socially excluded. It is possible that turning to familiar cultural and social structures helps students to overcome psychological and physical distress. However, the dominant role of the L1 might hinder their development of the FL (Baumgarten 2014).