Under the Radar: A Survey of Students’ Experiences of Discrimination in the German University Context

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for the opportunity to review this paper. I read it with interest.

I am not sure what is meant by this statement on line 51: “depending on the permeability of the respective society”. Similarly, what is a “biographical transition into adulthood” (lines 52-53)? The term “biographical” appears again on line 56. May the authors please clarify?

What are examples of “acts of discrimination in the course of exams” (line 58)?

I am very surprised that the AGG excludes university students. I am equally surprised that employees are protected. Why this discrepancy if education is not a federal matter in Germany? Could the authors elaborate on the remarkable omission of students from the law?

The authors state that the “majority of articles…” (line 90) but offer no objective measure for this. It may be their (subjective) impression, but without a systematic review, the claim is not necessarily accurate. Similar claims appear in Section 1.2 and are problematic without a description of how the review was conducted.

In Section 2, I would encourage the authors to incorporate insights from the literature to support their selected methodology. The participatory nature of the questionnaire design is very interesting, for example. What inspired this and what does the corresponding literature say about this method?

Curiously, the authors do not define ‘discrimination’ or ground it in any scholarly discourse. Yes, the general sentiment around it is acknowledged, but what does it mean in the context of this study specifically, as well as in the broader literature? It may have different and intersecting connotations, across fields of social interaction, emotional and mental wellbeing, political ideology, etc. Are the authors referring to single incidents/acts of discrimination, and what kind of acts would these entail? We see some references to different forms on line 264, but this is insufficient in terms of the overall study. Was discrimination defined in the questionnaire or how did students understand the concept? Did the authors explore any forms of systemic or institutionalised discrimination? I find it interesting that the authors did not seek to determine how students understood the concept (and ground this in a body of literature).

I am not an expert in descriptive statistics, but the authors did not unpack the validity or reliability of their data. Read in isolation, the statistics are only descriptive, leaving both the writer and reader to interpret their meaning. But how reliable are these statistics and can we form any meaningful interpretations without knowing whether the data is statistically significant and whether the data collection/analysis was sufficiently rigorous? What is the total student population of this university? What proportion of respondents were under- and postgraduates in relation to the total number? Table 1 is not sufficiently detailed in this regard.

Regarding Is queer people the appropriate catch-all ‘category’ for LGBTQIA+ individuals? Why do the authors not categorise “Jews” under religious minorities (assuming that this is the case)? Why do the authors use a category like “fat people” when this could be described as “People with particular physical characteristics” or similar?

As a non-German reader, I am not familiar with the roman numerals as Faculty names, e.g., Faculty of Philosophy I, II or III. Does it refer to the year of study? What about postgraduates?

Overall, the paper is well structured and soundly argued, but requires substantive revision for it to be publishable. I encourage the authors to revise the methodology and to ground their findings in scholarly discourse.

-Comments on the Quality of English Language-

The paper can benefit from being proofread by a professional language editor.

Author Response

Dear Sir or Madam,

We thank the editor for the opportunity to revise our manuscript and we thank the reviewers for their detailed and insightful comments and suggestions. We think that this revision substantially improved the manuscript and that the resulting article presents our study in a more comprehensive manner.

Below, we explain in detail how we incorporated each of the reviewers’ suggestions. In the manuscript, the changes are marked in green. 

Review 1

1. I am not sure what is meant by this statement on line 51: “depending on the permeability of the respective society”. Similarly, what is a “biographical transition into adulthood” (lines 52-53)? The term “biographical” appears again on line 56. May the authors please clarify?

We used the term „permeability“ to qualify societies’ openness for upward social mobility. Since the term seems to be not as understandable as we thought, we omitted the explanation in brackets. Similarly, we initially qualified adolescents’ transition into adulthood as “biographical”, but we understand that this is a pleonasm here and deleted the term “biographical”. In the next sentence, we changed “biographical and learning trajectories” to “personal development and learning trajectories” to make the wording more precise. 

2. What are examples of “acts of discrimination in the course of exams” (line 58)?

We added an example of discrimination during exams to illustrate this point.

3. I am very surprised that the AGG excludes university students. I am equally surprised that employees are protected. Why this discrepancy if education is not a federal matter in Germany? Could the authors elaborate on the remarkable omission of students from the law?

We were very surprised as well when we first encountered this twisted legal situation. It can be explained by the fact that in Germany education is under the legal authority of the federal states, which includes the provision of everything related to university education including students’ protection against discrimination. In contrast, employees’ rights are a matter under the authority of the central government, therefore the AGG is effective for the protection of university employees.

We added more explanation on this to the manuscript.  

4. The authors state that the “majority of articles…” (line 90) but offer no objective measure for this. It may be their (subjective) impression, but without a systematic review, the claim is not necessarily accurate. Similar claims appear in Section 1.2 and are problematic without a description of how the review was conducted.

We agree that quantitative statements about the literature should only be made on the basis of a systematic literature review. As of now, no such review is available for our topic and it was beyond the scope of this study to systematically and comprehensively review the literature. We therefore changed the wording and now refer to the fact that the respective scientific discourse has a strong focus on the situation of US universities. This statement is backed not only by our knowledge of the literature but also the marginality of the topic at scientific conferences and in public discussions related to university policies. To further support this statement, we added a reference to a paper by the German Federal Antidiscrimination Agency where this government organization also deplores the research gap that we describe.     

5. In Section 2, I would encourage the authors to incorporate insights from the literature to support their selected methodology. The participatory nature of the questionnaire design is very interesting, for example. What inspired this and what does the corresponding literature say about this method?

We thank the reviewer for this suggestion. Overall, the participatory elements in our study design followed established procedures, which tempted us to neglect their detailed description. We are therefore grateful for the opportunity to provide more detail and revised the text accordingly.  

6. Curiously, the authors do not define ‘discrimination’ or ground it in any scholarly discourse. Yes, the general sentiment around it is acknowledged, but what does it mean in the context of this study specifically, as well as in the broader literature? It may have different and intersecting connotations, across fields of social interaction, emotional and mental wellbeing, political ideology, etc. Are the authors referring to single incidents/acts of discrimination, and what kind of acts would these entail? We see some references to different forms on line 264, but this is insufficient in terms of the overall study. Was discrimination defined in the questionnaire or how did students understand the concept? Did the authors explore any forms of systemic or institutionalised discrimination? I find it interesting that the authors did not seek to determine how students understoodthe concept (and ground this in a body of literature).

We thank the reviewer for this important comment. Indeed the article missed an explicit definition of our understanding of discrimination.

In designing the study, we mostly relied on a notion of discrimination that understands discrimination as the process of unfair treatment due to (possibly only alleged) group memberships, whereby the groups relate to each other in a relation of subordination that exists beyond the concrete discriminatory situation and reflects overall patterns of society. This definition has been popularized within public health by Nancy Krieger, to which we now refer at the beginning of the introduction. Also, we added a reference to the different forms of discrimination and come back to this classification in the discussion.

Trying to foster a shared understanding of discrimination when answering the questionnaire, we provided respondents with a definition of discrimination and used the legal definition used in Germany for that. This definition reads: “Discrimination in the legal sense is unequal treatment of a person on the basis of one (or more) legally protected categories of discrimination without an objective reason justifying the unequal treatment. Discrimination can be expressed, for example through the behavior of a person, a regulation or a measure.”

In line with this definition, the questionnaire prompted both interpersonal and institutional discrimination and explicitly asked about discrimination constituted by “physical barriers at university” or “university’s rules and regulation”.

We added information on this aspect of the questionnaire to the methods section.

Lastly, we are very much interested in students’ understanding of discrimination and the way they make sense of their experiences. Still, we felt that a quantitative survey would not be a suitable methodological approach to investigate this topic. Therefore, we conducted semi-structured interviews in parallel to the survey, which aimed to gain a deeper understanding of students’ experiences. Since the analysis of those interviews is not yet completed, we are unable to refer to those findings yet, but we aim to publish this qualitative part of the overall study as soon as possible.

7. I am not an expert in descriptive statistics, but the authors did not unpack the validity or reliability of their data. Read in isolation, the statistics are only descriptive, leaving both the writer and reader to interpret their meaning. But how reliable are these statistics and can we form any meaningful interpretations without knowing whether the data is statistically significant and whether the data collection/analysis was sufficiently rigorous? What is the total student population of this university? What proportion of respondents were under- and postgraduates in relation to the total number? Table 1 is not sufficiently detailed in this regard.

We would like to answer to this comment in five parts:

1. a) Generalizability

This comment touches on one of the major limitations of this study: Since our university denied us access to the registry of students’ email addresses, we have no way to ascertain how many students actually received our invitation. Thus, we cannot calculate the response of the survey and had to abstain from calculating prevalence estimates for the occurrence of discrimination. This also entails that we cannot generalize from our findings to the overall student population with respect to the prevalence of discrimination and therefore decided against reporting confidence intervals (which might falsely encourage the inference of generalizing conclusions). We revised our limitations section to make this point clearer.

In addition, we take this limitation into account in our formulation of the research question, that outlines our aim to “assess in which situations discrimination happens, who discriminates, and for which reasons”. The answer to this question does not rely on generalizations but can be given by understanding the particulars of the discrimination experiences of our participants.

1. b) Validity

While we agree that it would be preferable to validate the descriptions of participants’ experiences of discrimination, we cannot conceive any feasible methodology that would allow us to do that: While lacking any gold standard for the measurement of discrimination experiences, we see no way how our findings could be validated. If we misunderstood this part of the comment and the reviewer actually refers to something else, we will be happy to take this into account.  

1. c) Rigor of data collection and analysis

Rigor in data collection and analysis is key for any scientific work. We assured a reliable and rigorous data collection by following all required steps in survey research (participatory development of the questionnaire based on existing questionnaires, pretesting, working in teams etc.); we now added more information on these steps in the methods section.

Concerning data analysis, the issue of reliability mostly concerns the categorization of the text variables. Hereby we followed established procedure and used a collaborative analysis by two researchers that then discussed any discrepancies with the bigger research team. We hereby used an inductive Thematic analysis and aimed to stay as close to emic terms as possible. In cases where the text data was ambiguous, we resorted to broader categorizations. The categorizations derived in the Thematic analysis were discussed with the team of experts and adapted according to their feedback. We added more explanations concerning this procedure to the methods section.   

1. d) Statistical significance

This study was designed as an exploratory approach, so no power and sample size calculations were done beforehand. In line with current guidelines (e.g. Wasserstein and Lazar (doi: 10.1080/00031305.2016.1154108) or Wasserstein, Schirm and Lazar (10.1080/00031305.2019.1583913)) this excludes the use of statistical tests, since they would not be meaningful. We added this information to the methods section.

1. e) Presentation of findings

We agree that as of now it is not possible for readers to assess how the sociodemographic composition of our sample relates to the overall student population. We therefore added further columns to Table 1 that display the distribution of the various sociodemographic properties in the overall student population. Since university administration routinely does not collect data on many of the characteristics of interest for our study, for some of the characteristics no comparison is feasible. We denoted those in the Table as “no data available”.   

We hope that this way of presenting the data meets the reviewers concerns. At the same time, we want to highlight once more that our study design did not aim at the recruitment of a representative sample (since we did not aim to estimate prevalences) but rather aimed to cover all faculties and a wide range of student subgroups, so that we can understand the different experiences that students make in the context of university.  

8. Regarding Is queer people the appropriate catch-all ‘category’ for LGBTQIA+ individuals? Why do the authors not categorise “Jews” under religious minorities (assuming that this is the case)? Why do the authors use a category like “fat people” when this could be described as “People with particular physical characteristics” or similar?

We believe that this comment refers to the qualitative analyses that were used to analyze text variables. To improve the transparency of our approach, we added more details to the methods section that will help readers to understand how we derived the categories in question.

Overall, we constructed categories as specific as possible and only resorted to broader categories for characteristics that were vaguely defined in the data or that represented heterogeneous manifestations of one broader underlying construct. E. g., we categorized descriptions referring to respondents’ body weight or body shape under the self-imposed term “fat people” as distinct from the broader (and less well-defined) category “people with particular physical characteristics”, with the latter category serving as a residual category for color of the eyes, hair color etc.

Also, we did not used the term “queer” as an umbrella term for LGBTQIA+, but distinguished these terms whenever possible. For instance, in reference to respondents’ gender we followed their self-designated expressions and distinguished “queer” from “homosexual” and “bisexual”. Similarly, in terms of respondents’ gender we distinguished “queer” from “non-binary”, “non-binary trans”, and “transmale”.

The categorization of “Jewish” as a category in its own right follows established standards in antisemitism studies that argue that Judaism is not merely a religion but can be an ethnic and cultural identity as well (e.g. Webber (doi: 10.1080/01419870.1997.9993961)). The majority of Jews currently living in Germany have a migratory background from the former USSR, where Jewish religious practice was severely restricted (Chervyakov et al. (doi:10.1080/01419870.1997.9993962)). Consequently, many German Jews do not identify as religious (Körber (2019)). Since we have no information on the religiosity or ethnicity of the respondents, we cannot subsume Jews within the category of religious minorities, leaving the use of “Jewish” as a category in its own right as the best solution. (We do list “Jewish” as a religious affiliation though, if it was given as an answer to the question of religious affiliation.)

9. As a non-German reader, I am not familiar with the roman numerals as Faculty names, e.g., Faculty of Philosophy I, II or III. Does it refer to the year of study? What about postgraduates?

We thank the reviewer for drawing our attention to this potentially misleading form of presentation! The Roman numerals denote the different faculties (in their official names) and have nothing to do with the year of study. To avoid confusion with non-German readers, we removed the Roman numerals and use only the descriptive names for the faculties.  

10. Overall, the paper is well structured and soundly argued, but requires substantive revision for it to be publishable. I encourage the authors to revise the methodology and to ground their findings in scholarly discourse.

We sincerely thank the reviewer for his or her detailed comments and suggestions! We hope that our changes meet the reviewer’s expectations and that all ambiguities and shortcomings are sufficiently addressed now.  

Reviewer 2 Report

Comments and Suggestions for Authors

“little is known about the extent and characteristics of university students’ experiences of discrimination in Germany” in the abstract might not be an accurate description.

The literature review on the discrimination in German university could be further updated (line 103-121). Some source e.g. (Jendretzky et al., 2020) could be adopted there

There are some reference available to this matter as well:

https://yerun.eu/2023/03/discrimination-at-german-universities-how-widespread-is-it/#:~:text=The%20results%3A%20About%20a%20quarter,harassment%20and%20violence%20than%20men.

https://www.justiceinitiative.org/voices/german-schools-quiet-deep-discrimination-problem

https://www.nature.com/articles/d41586-023-00955-9.pdf

In short, the authors might have to clearly demonstrate in what way the understanding of students’ experiences of discrimination in Germany is important.

-Line 133 “The high frequency of discrimination in the university context is in stark contrast to the scarcity of scientific engagement with this topic”, is that this research more scientific then the previous?

- the scale “often” and “regularly” is not very scientific indeed.

-Perhaps section 3.5 could come after 3.1 for better understanding.

-“Queer” appears in three different occasions in the questionnaire which is quite confusing. Please confirm whether it is the only correct word.

-Section 4.2 – is there any reason / implication for the varied discrimination across different academic fields?

-Perhaps the author could suggest some changes to university’s rules which account for 25% of discrimination.

Other minor comments:

Not sure if the in-text citation format meets the journal requirement.

Wrong spelling for “physical” in table 4?

Line 174 on the limitation on sampling is covered in later part and perhaps could be deleted.

Comments on the Quality of English Language

language standard is acceptable.

Author Response

Dear Sir or Madam,

We thank the editor for the opportunity to revise our manuscript and we thank the reviewers for their detailed and insightful comments and suggestions. We think that this revision substantially improved the manuscript and that the resulting article presents our study in a more comprehensive manner.

Below, we explain in detail how we incorporated each of the reviewers’ suggestions. In the manuscript, the changes are marked in green. 

Review 2

1. “little is known about the extent and characteristics of university students’ experiences of discrimination in Germany” in the abstract might not be an accurate description.

We agree that our initial wording was a bit too vague and thank the reviewer for drawing our attention to this. We changed the wording in the abstract (and also in the manuscript) and are now more precise about the research gap that we address.

2. The literature review on the discrimination in German university could be further updated (line 103-121). Some source e.g. (Jendretzky et al., 2020) could be adopted there.

There are some reference available to this matter as well: …

In short, the authors might have to clearly demonstrate in what way the understanding of students’ experiences of discrimination in Germany is important.

We thank the reviewer for drawing our attention to these articles and the research that they report. Most of them were published after we completed our data collection and must have slipped our attention later on... We now included references to the three Germany-wide surveys conducted in the last years and added a separate paragraph for that to the introduction. (The reference to the study by Jendretzky has already been in our manuscript, in the discussion. We now also refer to this study in the introduction.)

With this new paragraph, the introduction became quite long. We therefore removed the section on the research gap, which was anyway a bit redundant with the sections “state of the literature” and “aims and objectives”. With a few more explanations in the “aims and objectives” section, we think that the particular focus of this article becomes more apparent now.  

3. -Line 133 “The high frequency of discrimination in the university context is in stark contrast to the scarcity of scientific engagement with this topic”, is that this research more scientific then the previous?

We did not mean to imply that previous research was not scientific enough. To avoid any ambiguity here, we completely changed this sentence and focus on the (lack of) reception of scientific findings in policy-making now. 

4. - the scale “often” and “regularly” is not very scientific indeed.

We assume that this comment refers to the frequency of discrimination experiences. The scale we used here encompassed the categories “never – once – repeatedly – regularly” (in the German original: nie, einmal, mehrmals, regelmäßig). We initially used “often” as a translation for “mehrmals” but upon further reflection now believe that “repeatedly” is a more accurate translation.

Beyond that, we are not sure what the criticism against this scale refers to: In constructing the question we followed Dillman’s (1978 and 2014) advice to use four categories for unipolar scales. Furthermore, we followed recommendations common in the literature (e.g. Christian, Parsons et al. 2009) not to combine numeric labels with vague quantifiers if the question does not aim to measure precise quantities. Beyond that, we pretested the questionnaire and did not encounter problems with this question in the course of the pretest.  

5. -Perhaps section 3.5 could come after 3.1 for better understanding.

Our manuscript does not contain numbered sections, maybe this is something the editorial manager added and we can’t see it? We are therefore unsure to which sections this comment refers.

6. -“Queer” appears in three different occasions in the questionnaire which is quite confusing. Please confirm whether it is the only correct word.

As outlined in more detail above (and as reflected in the new explanations added to the methods section), we tried to stay as close as possible to emic terms in our categorization of the information from text variables. As a result, the term “queer” came up as a description of respondents’ gender identity, sexual orientation, and self-identified group membership.

We acknowledge that this multiple usage of the term for (indeed different) purposes is slightly confusing. On the other hand, this usage reflects the category system of our respondents and is in congruence with the use of the term “queer” in their every-day discourse. We therefore prefer this approach over its alternative, which would entail to impose our own terms on our respondents’ categories. This might avoid the multiple use of one term, but would necessarily imply a loss of information.

If the reviewer sees a third option that we were unable to identify so far, we will be more than happy to change the manuscript accordingly.    

7. -Section 4.2 – is there any reason / implication for the varied discrimination across different academic fields?

We assume that the differences between the different faculties might be explained by three reasons: First, the faculties differ in the sociodemographic composition of the students, which might influence the absolute frequency of discrimination. E.g., in a mostly male faculty discrimination on the basis of sex might be less prevalent, simply because there are few women and some classes might even be all-male.

Second, the sensibility for discrimination might differ between faculties leading to an underreporting of discrimination in fields where discrimination is usually not paid attention to.

Finally, there might be differences in the actual occurrence of discrimination due to each faculty’s culture.

Unfortunately, our study design does not allow for conclusions concerning the relevance of each of these potential influences. Also, the existing literature does not provide ready answers here. We therefore decided not to discuss this topic in depth, but added an explanation to the limitations section.  

8. -Perhaps the author could suggest some changes to university’s rules which account for 25% of discrimination.

Some of those needed changes are outlined in the guideline by the German Anti-Discrimination Agency. We added a paragraph elaborating on some of those measures in the conclusions section.

9. Other minor comments:

Not sure if the in-text citation format meets the journal requirement.

Thank you for this comment! In our initial draft, we must have selected the wrong citation format in our source manager and corrected this mistake now.

Wrong spelling for “physical” in table 4?

Thank you, that was a typo. We corrected it.

Line 174 on the limitation on sampling is covered in later part and perhaps could be deleted.

We agree that this information was redundant and removed it from the section in the methods.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

approved after authors' last revisions

Author Response

Dear Sir or Madam,

We thank you very much for your detailed and constructive feedback! 

Best regards,

the authors 

Reviewer 2 Report

Comments and Suggestions for Authors

Thanks the authors in addressing much of my concerns, some observations based on my last review for the author’s consideration:

Wrong spelling for “physical violence” in table 4 is not fixed

The problem for In-text citation is still there occasionally starting from the discussion section.

Line 169-172 on the limitation on sampling is still there. It is up to authors’ decision to keep these words or not.

Author Response

Dear Sir or Madam,

thank you very mch for your detailed and constructive feedback in the first round of review. The issues that you mention below got lost between two version of the revised manuscript. We now made sure that they are addressed as you suggested it. 

Best regards,

the authors