Term papers in corpus linguistic seminars should follow the so-called IMRaD structure, a standard form of empirical research papers consisting of four sections: Introduction, Methods, Results, and Discussion. To learn more about this standard and its use in general, consult the Wikipedia article on IMRaD. Here, we have adapted it specifically to corpus linguistic research, specifying the questions you should answer in writing each section of your paper. A more detailed version of the questions you should answer in the core part of such a paper and the way you should keep track of them when conducting your research can be found here. Note that it is also useful to use the structure for taking notes while /reading/ corpus linguistic research papers — you can download a document template for such notes as a docx file (Word) or an odt file (Libre Office) (if your browser does not allow the direct download of files, you need to right-click on these links and select “Save Link As” from the context menu.

The introduction should simply be titled Introduction. It must answer the following questions:

• What is the topic of the study, What is the linguistic phenomenon under investigation (give an example!). What aspect of the phenomenon is under investigation, what factors are expected to play a role in the description and explanation?
• What is the specific research question or hypothesis?
• What is the current state of knowledge about the phenomenon in the field? Are there current publications that deal with the same question, how is the phenomenon described in current reference works (grammars, dictionaries), if at all?

In addition, it may answer the following questions:

• What is the broader purpose of the study?
• Why (or for whom) is this research question relevant?

The methods section can simply be titled Methods, but depending on the focus that you give different aspects of the research design, it can also be titled Methods and Materials, Methods and Data, Data Extraction and Annotation or something similar. It must answer the following questions:

• What corpus or corpora were selected for the study and why?
• How were the data extracted? What query was used? Were the data checked manually? Were specific cases excluded individually? Did you include the entire data set in your analysis or did you draw a random sample?
• How were the data categorized? What categories did you use and why? How did you assign data to your categories (annotation)?

The results section should simply be titled Results. It must contain a compact presentation of your data (in the form of tables and/or diagrams) that answer the following questions:

• What is the answer to your research question?
• Is your hypothesis supported by your results, or is it falsified?
• In addition, it may contain a brief assessment of the quality of your data, for example, if the data do not provide a clear or complete answer.

The discussion section can simply be titled Discussion, but depending on your focus you can also use combinations of the words Summary, Conclusion, and Outlook (if you have a long discussion section, you can use these as subheadings). The section must answer the following questions:

• How can your study be summarized, i.e., what was your research question and what is the answer (someone reading only the discussion section should be able to have a rough idea of your study).
• How does this relate to the current state of knowledge in the field? Does your result confirm what others have found? Have you added new insights?

In addition, it may contain answers to the following questions:

• What follows from the answer? How is it relevant in terms of theory or application?
• What are the limits of your study? How could it be expanded in future research?