Help for scientific work

To make writing a scientific paper easier, you will find information and assistance for writing a successful scientific paper here.

Interested visitors can find further helpful references under Research and Further links.

It is worth starting as early as possible in your studies to set up your own literature database with the appropriate software. You can enter everything you read there over the course of the semester. This gives you an overview of the texts you have studied. If you store notes or excerpts on the texts and index everything, you can always access them again later. A well-maintained literature database makes academic work much easier.

Citavi software is available free of charge for students via the university, but can only be used free of charge for the duration of your studies. A good alternative is the free and open software Zotero.

A large proportion of academic literature is now available on the internet and can be read online and/or downloaded as a PDF file. However, it is still essential to make use of the printed holdings of libraries, as older texts in particular are often not available online.

While introductory and basic texts can often be found in monographs and anthologies, current empirical findings and theoretical debates are often published in specialist journals or as research reports. When working on current topics, it is therefore essential to also search for articles from relevant journals and research reports from relevant institutions (ministries, offices, institutes, foundations, etc.). The website of the Urban and Regional Sociology Section of the German Sociological Association provides a list of journals, national institutes and organizations, as well as international institutes and organizations.

When you start working on a topic (e.g. for a term paper, thesis, etc.), it makes sense to read basic texts first. This will give you an overview of theoretical terms and concepts as well as the current state of research. The literature cited can be used to identify important authors and publications for the subject area. If these are also relevant to your own research question, you should continue your research here.

Many books and journals are licensed as e-books or electronic journals via the university library. To gain access to the full texts, you must either be online via the university network (WLAN on campus, educampus) or be connected to the university network from home via VPN. When researching with KARLA, full texts can now be accessed, depending on availability. On other sites, such as SpringerLink or JSTOR, you will also be identified as a member of the University of Kassel and given access to the licensed texts.

The following sites, among others, are suitable for literature research

• KARLA: With KARLA you can search the literature available through the university library. Under "Articles & More" you will also find an extensive list of generally available texts for the keyword entered. The results are provided by the meta search service EBSCO, which searches numerous catalogs and databases worldwide simultaneously. Some of these texts are also available as full text; a corresponding link is displayed.

• Google Scholar: Google Scholar searches the Internet for scientific publications. Here, too, the search results are quite extensive, but they also include term papers and PowerPoint presentations. Here you should check carefully whether the texts meet scientific standards. Some PDFs are available directly via a link. Otherwise, you can use KARLA to check whether the relevant text is available as a full text from the university.

• GESIS: The Leibniz Institute for the Social Sciences (GESIS) offers various services on research data, survey instruments, publications and other information.

• JSTOR: Here you can find older articles (partly from digitized print editions), mostly from journals in the humanities and social sciences. The database is also integrated into KARLA (under Articles & more).

• BASE: A search engine of Bielefeld University Library, which searches various scientific document servers with mostly freely accessible texts. Titles from BASE are integrated into KARLA (under Articles & more).

Further information on specialist databases in the field of architecture, urban planning and landscape planning and on correct citation in academic papers can be found in the Faculty's guide to academic writing.

In an exposé, the idea for an academic paper is briefly outlined using important key points. It should be formulated before you start writing your term paper or BA/MA thesis. An exposé helps to concretize your own idea, to define a procedure and common thread for your work and serves as a basis for consultation with the lecturers. Therefore, an exposé should be prepared before writing any academic work (term papers as well as BA/MA theses). This should already demonstrate competence in scientific work and it is therefore essential to ensure good language and correct citation. The length should be approximately two pages for a term paper and three pages for BPS, BA or Master's theses.

An exposé should be structured as follows:
1. Topic
2. State of research
3.Research question & hypotheses
4. Procedure
5. Outline
6. Literature

First, the general topic of the paper is presented and why it is relevant to deal with it is emphasized. This is followed by key (preferably current) research findings and any unanswered questions in the field. The relevance of the topic and possible research gaps are then used to derive the specific research question for the thesis. If hypotheses on the research question can be formulated from the literature already reviewed, these can also be presented. Finally, a description is given of how to proceed in order to answer the research question. This includes explaining and justifying which literature will be used and - if planned - which empirical data will be collected and analyzed and how.

A rough, preliminary structure (which may well change) and a list of the literature already researched and used in the synopsis form the appendix.

(Applies equally to term papers, student research projects, BA and MA theses)

1. every academic paper needs a concrete question.
It is not enough to define an approximate topic to be dealt with. Every academic paper has a specific question that the paper attempts to answer. The question should be specific enough to be answered within the maximum number of pages allowed. Answering a question does not mean finding "the truth", but finding evidence for answers with the help of scientific literature and (external or own) empirical data. Possible contradictions should be pointed out and discussed.

2. a term paper must contain an original contribution and not just reproduce existing information.
Simply reproducing existing information and theses from other scientific texts is not a sufficient personal contribution. Rather, it is necessary to link existing theories and data within the framework of your own research question. For example, a theory could be tested on the basis of a specific case study. An own empirical survey to answer a question is also a personal contribution.

3. thorough literature research is essential for scientific work!
Scientific knowledge builds on one another. In order to arrive at your own research question, you should therefore first research the literature on the topic of interest. It is advisable to first read introductory and general literature on a topic to gain an overview. Sub-topics and research directions can then be explored in greater depth. When researching, attention should also be paid to the topicality of the sources.

More detailed information on literature research can be found on this page of the homepage of the Urban and Regional Sociology research group.

4. every piece of information must be supported by a scientific text! Newspaper articles and internet sources are not scientific texts.
Assertions that are made and information that is reproduced must be substantiated by scientific texts. Of course, this does not apply to generally accepted facts (the earth revolves around the sun). Newspaper articles are not scientific sources; they should only be cited to illustrate a media/public discussion. The same applies to websites. They can be cited, for example, to illustrate the self-portrayal of a company or the plans of a municipal building authority. Your own assertions and theses do not have to be substantiated. However, they should be derived from the information presented. In order to prove that extensive literature research has been carried out, it is advisable to cite several sources.

Information on correct citation and references can be found in the guidelines for scientific work

5 To avoid plagiarism, any information taken from texts must be identified as a citation!
Foreign ideas and words must be distinguished from your own. For this reason, verbatim quotations from other texts, as well as rephrased or analogous quotations, must be identified and documented as such. Otherwise it is plagiarism. This is not only bad academic practice, but is also regarded as intent to deceive in the context of term papers and theses.

6. only information and arguments relevant to answering the research question should be presented and these should be sensibly structured.
All contents of the thesis should be subordinate to the research question, i.e. contribute to answering it. Information that does not serve to answer the research question should be omitted. The paper is also structured sensibly by logically arranging the various sections. Chapter introductions explain why the following information is important. Chapter excerpts summarize and lead to the next chapter. This creates a common thread and the reader retains an overview.

7. finished work should be proofread!
Correct language (spelling and punctuation) and appropriate expression (understandable and precise, but not in everyday language) are also part of academic work and are included in the assessment of papers. Often you can no longer see your own mistakes and shortcomings because of all the text. That is why it is important to have the finished work proofread by another person. In addition to linguistic correctness, this person should also check the comprehensibility of the content. Even a cursory proofreading can reveal the majority of typos and spelling mistakes.

8 Don't despair! Writing scientific papers can also be practiced.
What is a good question for a term paper? When is my literature research complete? How do I express myself appropriately? These are all questions that can only be answered in the abstract. But it can be practiced and with experience comes a feeling for good questions and literature research. Feedback and criticism from lecturers and fellow students is also important.

If quantitative research is to be carried out as part of a scientific paper, a question should be found that can also be answered empirically as part of the paper. As a general rule, the more general the question, the greater the amount of work involved in both the research and the writing. This can easily lead to the thesis getting out of hand and not being manageable within the given time and number of pages.

The aim of the thesis is to address the underlying question. This does not mean finding a general "truth", but finding evidence for arguments and answers with the help of theory and empirical research. Contradictions and problems should be identified and discussed during the process of finding an answer.

Example of a question: What is the connection between the crime rate in a neighborhood and the proportion of people without German citizenship ("foreigners") living there?

First of all, a thorough and comprehensive literature research to determine what other researchers have already found out about the research question and closely related topics. Contradictory findings can or should also be presented. The summary of the state of research serves to specify your own research question, to derive the research desideratum and to be able to better classify your own findings in the course of the work. For example, your own findings may contradict previous research (there can be a whole range of reasons for this: The context has changed over time, researchers have made mistakes, different methods have been used, etc.). In this section, however, only what has already been done will be determined. The results are only classified in the analysis and conclusion.

Example sentence from a state of researchFriedrichs (1985) shows that the proportion of people without German citizenship (proportion of foreigners) in cities and urban districts shows no correlation with the crime rate.

The state of research is intended to provide an overview of previous research on the topic, so it is usually important to use more than one author.

Quantitative scientific work (almost) always proceeds deductively. This means that the researcher begins with theory-based assumptions and checks these or attempts to refute them. Therefore, the next step in a quantitative paper is to research theoretical works relevant to the research question and summarize them in the paper in such a way that they provide a framework for the research question. For example, the socio-ecological approach of the sociology of crime (Chicago School) could be used for the example question.

In the case of non-deductive (descriptive) quantitative work, the focus is on the state of research and how your own work fits into this.

As part of the theory section, it is necessary to clearly define the technical terms used. Especially when terms are used colloquially or debated in science, the same term can mean different things. A short definition at this point saves the reader a lot of confusion and helps the authors to see the research object more clearly.

Example definitionIn this work, criminality in the German legal system is understood to mean deviant behavior that is considered punishable (see Härter 2013).

The researcher then forms hypothesesbased on the theories and the current state of research. These represent simple assumptions about aspects of the research question and are hypothetical answers to the research question. They should be formulated in such a way that they can either be rejected or retained. If a hypothesis contains several aspects, it makes sense to formulate sub-hypotheses, each of which can be tested separately.

Example hypothesisNeighborhoods with an above-average proportion of socially disadvantaged people have an above-average crime rate.

If it turns out that there is no difference between neighborhoods with an above-average proportion of.... with regard to... or that even the opposite is true, the hypothesis would have to be rejected and is thus falsified. It can only be retained if a clear correlation corresponding to the hypothesis can be seen.

It must be clarified and justified which data from which source is to be analyzed. In the simplest case, an existing data set (from other researchers) is used (e.g. the Socio-Economic Panel or the World Value Survey). These data sets are often freely available on the Internet, or can be requested from the research institutions. In the event that data has been collected, it must be explained which procedures were used to collect the data, what problems were encountered, to what extent the data may be unreliable (e.g. different definitions of migration background in different cities) and how the problem is dealt with. Data must generally be briefly characterized: Sample size, distribution of characteristics, etc.

Example of data basisThis paper uses data from the German Socio-Economic Panel (SOEP) provided by the German Institute for Economic Research (DIW Berlin).

The methods section explains which variables are used, how they are scaled, how they correspond to the terms used and which evaluation methods are applied. It is useful to first explain which variables are to be used in the analysis: dependent variables (the variables that are influenced according to the hypothesis, in the example crime), independent variables (variables that influence the dependent variable according to the hypothesis, in the example citizenship) and control variables (variables that have an influence and must therefore be taken into account in the model but are otherwise irrelevant for the hypotheses, e.g. gender). Here it is important to mention how the variables are scaled and how they were calculated (in case indices were used, extreme cases were excluded or the variables were otherwise transformed).

The next step is to explain which analysis procedures are to be carried out. These would be, for example, various descriptive methods (e.g. frequency counts, cross-tabulations, graphs, ANOVAs) or regression and correlation analyses. Here, the analysis methods should be roughly explained (e.g. comparison of two distributions - when is there a clear difference?), potential peculiarities and problems explained and own criteria (e.g. what is considered a low, medium or high correlation? (There are research standards, but these can differ, so it is important to explain the criteria used)) are explained and justified.

In the analysis section of a paper, the relationships stated in the hypotheses are examined. Descriptive or inductive methods are used to analyze whether the hypotheses are correct or must be rejected. The discussion of the results and their interpretation is sensibly the subject of the analysis section, but can also be presented separately.

If unexpected results occur (even when investigating control variables, as long as it does not go beyond the scope) that do not correspond to the expectations from the state of research or theory, plausible explanations and possible sources of error should be discussed. In doing so, it may be useful to refer to the literature (e.g. why does the increased presence of people of female gender have an increasing effect on crime according to the data). It may also be useful to take another critical look at the methodology.

At the end of longer quantitative analyses, it is useful to briefly summarize the most important results so that the reader has an overview.

Example analysisTable 3 shows a slight but significant correlation between the crime rate in the neighborhood and the citizenship characteristic. Accordingly, the hypothesis that non-Germans tend to live in neighborhoods with higher crime rates cannot be falsified. It can be assumed that this correlation is due to the correlation researched by Friedrichs and Triemer (2009) that foreigners primarily live in neighborhoods with generally socio-economically disadvantaged residents. Neighborhoods with high levels of poverty have higher crime rates, while no or negative effects can be observed for foreigners (Friedrichs 1985).

In the conclusion the results are briefly summarized once againto answer the research question and place them in the research context. Here it should also be mentioned what could be done better in the work (e.g. limitations of the method, limitations of the data used, theoretical deficits, etc.), how the topic can be developed further and which further research questions follow on from the results.

If the results show that the hypotheses are not confirmed or the question cannot be answered clearly, this is not a problem as long as the steps in the work have been well justified and carefully worked out. Great results are not part of the assessment criteria.