PhD Skills: Proper use of references

Summary: References are an important part of scientific writing. They inform the reader how your work is situated, from which sources you draw, and make it easier to understand where your work is novel. There are a number of conventions to make sure you use them in the right way. It is easy to avoid some common mistakes - in this text I give an overview how to use references in the right way.

Research builds on other work. It is proper practice to clearly state what part of your research is yours, and what is derived from others. References are used to clarify for the reader what the sources of these other works are. In scientific writing the goal is to write with the utmost clarity and transparency of your argument. The reader has your text at his/her hands, but all other information that you draw from should be delivered to the reader in the easiest possible way. This is done through references.

Golden Rule: The reader is not Sherlock Holmes - (s)he must find your source with the least effort possible, and not playing detective.

What references are used for

References in the main text are used for five different purposes:

• State of the art: The state of the art section describes related work to yours. The purpose is to outline the field in which you do your research. For non-specialist readers the section is a valuable introduction to your area, since they are not acquainted with it. For specialist readers it helps to see whom you are referring to. In many cases you cannot be completely exhaustive, or there are several different lines of research possible. By means of the state of the art section, the specialist reader therefore knows to which 'group' you belong. Since the state of the art is often quite specialized research, sources typically are journal articles and conference papers.
• Problem context: The problem context describes what kind of problem or main research question you are trying to resolve. It is important to realize that problems are not neutral: what you may see as evidently something being a problem, may not be the case at all for someone else. Thus, you should also devote space to describe why you are engaging with a particular question. Most problems have some bias - they are linked to a particular ideology, theoretical framework, or research direction. They can also be quite new or particularly urgent - in those cases it is difficult to find substantial sources that may argue for a particular problem setting. Thus, sources may include not only journal articles and conference papers for the more established research directions, but also text books, and even common media such as newspapers or Internet links.
• Theoretical framework: The theoretical framework describes your network of concepts, definitions, relations between concepts, causal principles, and basic assumptions that you have in your research. In short, the theoretical framework is your story why your research makes sense. It is very likely that you work in some research tradition of other people who share the same set of assumptions as you do. Needless to say, the theoretical framework is critically important to your own research. Especially when readers can question the consistency or logic of your framework, it severely limits the credibility of your text. References are used to embed your framework in the work done by others. Since frameworks usually take a long time to develop or mature, sources typically are text books and journal articles.
• Method: The method section describes the step-wise procedure how you derive valid results from your research work. Research can have many different flavors: it can be theoretical, empirical experimental, quantitative/qualitative measuring, through design, and so on. Each of these different flavors of research encourages particular working methods to get results. Making a good method is not an easy task; methods are improved by many people over many years. Therefore, in most cases you will follow a particular proven method in your research work. Since every research has it own particularities, you need to document carefully how those methods are implemented in your work. Especially small deviations or variations of method can be very valuable contributions to the research community. References in most cases are to text books to lay the general foundation of your method, and typically to journal articles that show the latest and most up to date applications of those methods.
• Earlier work by yourself and your team: as you progress in your research work, you will increase the number of results and your insights in your research problem. Thus, you can refer to earlier work and demonstrate how things have developed or improved over time. For the reader, seeing the earlier work references helps to understand the amount of publications and also the relative quality based on the output channels (prestigious or lesser known conferences, or journals with high or low impact). Earlier work references can be to any work produced earlier.

What should be in a reference in the text?

Whenever you publish in a journal, for a conference, or any other kind of publication, there is a template available or guidelines how references should be used in the text. There are many kinds of systems available, and every journal, conference, or publisher uses a particular variant of one of those.

Follow the guidelines of the source you want to publish in

References in the text come in two main different styles: numbered or named:

1. Numbered example 1: As Charles [1] already noticed, referencing is not a trivial task. Variation 1a: As [1] already noticed, referencing is not a trivial task. The references are numbered. They can be listed at the end of the text in the reference section (alphabetically or in order of appearance in the text), or they can appear as footnotes on the same page in the text. Numbered references save you the trouble of precisely formatting the named reference in the text. Unless you use a reference manager with your word processor however, it means manually updating the numbers whenever you change the amount of references though.
2. Named example 1: Referencing is not a trivial task (Charles 1990, pp. 3). The reference gives the name of the author, year of publication, and page number(s) where the particular text can be found - all between brackets. The particular way of writing this down is always defined in the guidelines of the journal, conference, or other. Variations are: Charles (1990), pp. 3; (Charles 1990:3); or simply Charles (1990). Not having the page number in the body text reference however, would require putting the exact page number in your reference list. If the source is mentioned only once in your text, this is a viable option - but it becomes less handy when you refer to the same source and different page numbers multiple times. In those cases it is better to have the page number mentioned right after the reference.

Common mistakes with references

References serve many purposes. They not only document for yourself your area of research, but also inform the reader about your work and give important background information. For this purpose, references are an important part of proper scientific writing. Making mistakes means that the reader will not have complete information, or be confused what you are referring to. Handling references in the right way, simply is a matter of being precise and methodical - since this is rather repetitive work, doing the references right is also boring and at times uninspiring.

There are a number of common and less common mistakes with references. Here are a number of them:

• Ignore the guidelines: The guidelines are always very precise how references should be formulated both in the main text, and in the reference list. This goes down to small details like including the author's initials or full names, punctuation marks between initials or not, the year of publication between brackets or not, the precise order of the elements of a publication, and so on. This is the easiest mistake to make, and one that will upset the editor of the journal, conference, or publication quite a lot. Personally I do not use reference managers, because I find that at the end of the day, I still have to check whether it got the punctuation all right. So a simple trick I use, is first to assemble for each type of publication (journal, book, conference, etc.) the first reference correct, and keep a copy of it as baseline that I can use for new references. For the new references then it is simply a matter of replacing the copy with the information of the new references, and I can be sure not to forget anything of the formatting. In many cases the guidelines also give such examples - when they are available, then use those examples as base to copy from.
• Overstatement: Making claims like "many people say," "it is well known," or "it is generally assumed," and then referring to only one source, does not give sufficient proof for your claim. Try to find at least three correct sources to substantiate claims like this.
• Incompleteness: Make sure that all the required elements of a reference are indeed present. See the table below for the most common elements that should be in each type of reference. The DOI is a so-called Digital Object Identifier, which will link the reference to the exact source if available on the internet. This is most typically the case with Open Access articles (free to access for anyone), but also most recent publications behind a publisher's paywall have them. For more detailed information, check out for example this website: https://library.uic.edu/help/article/1966/what-is-a-doi-and-how-do-i-use-them-in-citations.

Typical elements that should be in a reference to a specific publication. Notice that the guidelines may always deviate from this list.

• Wrong reference: Making mistakes is human, making mistakes with references is More Than Human; it happens to the best. What sometimes happens, is that you attribute the wrong reference to a particular source (you meant to refer to Smith, but ended up referring to Jones). The best way to avoid this, is to check your publication after it is finished. Leave it a day alone, because at some point you become insensitive to the text anyway when you have seen it so many times. Then carefully follow each reference and see if they check up in the list.
• Missing reference: When you have a lot of references, it may happen that you forget to include one in your list. Use the same advice as with the wrong reference above.
• Superfluous reference: Superfluous references are references that you actually do not use in the text. This happens mostly after a number of revisions when you delete a source, but forgot to do so in the reference list. Another motivation to have superfluous references is to artificially inflate the literature list by adding references that actually do not add to the text. Such references should be removed from the reference list since they are not represented in the research text.
• Inappropriate self-references: To outline a development in your research, you need to refer to earlier publications of your own. However, you should do so only within reason of your argument. Overly citing yourself just to have your earlier publications cited in a publication is unethical behaviour, and you should avoid that at all cost.
• Wrong scope or balance: Try to be as comprehensive as possible and balance out the references evenly between topic areas. It usually appears rather strange for a reader if a certain topic of your publication is under-represented in the reference list, or another topic has the majority of references.
• Must I read the whole book? A quite common mistake is to take one specific sentence or section from a publication, and then in your reference refer to the whole volume. This means that you oblige the reader to read the whole book, in order to find your reference.
• I love my dictionary and encyclopedia: Unless your research is on language, getting your definitions from dictionaries and encyclopedia is not advisable. The reason is that most definitions from dictionaries and encyclopedia are targeted to the most common and generic form to be useful for all kinds of readers, but they often lack the specificity and nuance of research definitions. Use them by all means to be inspired, but when it comes to forming or referring your own concepts, it is better to stick to scientific sources like journals, conference proceedings, and text books.
• The internet is my library: The Internet is a great source to find information quickly. You do not even have to leave your computer and go to a library! However, information on the Internet also has the tendency to be changing, is less reliable, and impermanent - thus sources or information found on the Internet should be treated very carefully. Follow the guidelines below on the reference list for Internet sources, but it is better to avoid them in general. They can be a good first source for finding papers and textbooks, but always try to get then that publication in your library, or if your institution has a subscription to a publisher's database of books, from that source.

What reference lists are used for

In most cases, all your references are summarized at the end of your text. Another option, which is less common, is to have the references as footnote or side-note on the same page where the reference appears. As mentioned above, reference lists come in two major kinds: numbered, and based on author name.

Numbered reference lists look something like this (fictitious examples):

[1] Smith, J.J. (1990). My First Publication in a Journal. Journal of First-Time Authors, vol. 13, issue 13, pp. 1-13.

[2] Jones, A.A. and Carly, C. (1985). Our Paper in A Berlin Conference. In Schmidt, H. (editor). Proceedings of the International Conference. 12-15 July, University of Berlin, Berlin, pp. 408-499.

Note that with numbered reference lists, the order of the publications does not have to be alphabetical nor chronological.

Named reference lists look something like this (fictitious examples):

Ahmed, B.C. (2005). Publication Title. Journal of Publications, vol. 1, issue 1, pp. 1-6.

Coles, M. (2018). A Cool Conference Paper. In Lang, L. (editor). Proceedings of the 88th International Cool Conference. 1-3 January 2018, Helsinki University of Technology, Helsinki, pp. 1-5.

Coles, M. and Haines, J. (2018). A Hot Text Book. Cooking Publishers Ltd., Death Valley, CA, USA, pp. 120-150.

Named reference lists usually are organized alphabetically, on the family name of the first author. Authors with multiple publications are then organized chronologically.

Internet references to publications:

Achten (2019). Interaction Narratives for Responsive Architecture. Buildings, Vol. 9, Issue 3, 66. https://www.mdpi.com/2075-5309/9/3/66 [accessed 25-2-2020].

Kormaníková, L.; Achten, H.; Kop?iva, M. and Kme?, S. (2018). Parametric Wind Design. Frontiers of Architectural Research. Vol. 7. Issue 3, pp. 383-394. DOI: 10.1016/j.foar.2018.06.005 [accessed 25-2-2020].

Nováková, K.; ?eps, K.; Achten, H. (2017). Experimental development of a plastic bottle usable as a construction building block created out of polyethylene terephthalate: Testing PET(b)rick 1.0. Journal of Building Engineering. Vol. 12. Issue 12, pp. 239-247. DOI: 10.1016/j.jobe.2017.05.015 [accessed 25-2-2020].

Publications that are available through Open Access publishing need to have their Internet link stated with it. This either is a direct link provided through the publisher, or by means of the more general DOI. Additionally, it is necessary to note the date when you last successfully accessed the publication.

Internet references to all other kinds of topics (not recommended!):

Category Theory: Abstraction in Mathematics. Encyclopaedia Britannica. URL: https://www.britannica.com/science/foundations-of-mathematics/Category-theory#ref412235 [accessed 25-2-2020].

Coyne, R. (posted 8. 2. 2020). The future of prediction. URL: https://richardcoyne.com/2020/02/08/the-future-of-prediction/ [accessed 25-2-2020].

Global Warming (from Wikipedia, the free encyclopedia). URL: https://en.wikipedia.org/wiki/Global_warming [accessed 25-2-2020].

As stated before, internet references to general topics should be avoided. If it is necessary to use it, because you cannot access any other more reliant source, then make sure that an internet reference always needs a date when you last successfully accessed that source.

If the source has an author, then organize the url on the family name of the author. Organize the other source alphabetically by their topic/title name. Another option is to use numbered references for internet sources.

Common mistakes with reference lists

The most common mistake with the numbered reference list is mixing up the reference number. The most common mistake with the named reference list is not following the alphabetical order. Both cases usually happen after a couple of revisions. As a rule, after some main revisions of your text, take your time to carefully check each of the references.

For both types applies, that it is very easy not to follow the guidelines of the references. Carefully check the following:

1. Format of authors names and their initials (in particular punctuation marks between initials).
2. Year of the publication (sometimes in brackets, sometimes preceded with a comma, semi-colon, or colon).
3. Title of the publication and volume in which it is published (tricky is the use of Caps On All First Letters In The Title - or not; and the use of italics or not).
4. Volume and issue numbers of journals (they usually have a very specific style, like Vol. 1., Issue 2 or 1:2, and so on).
5. Page numbers (preceded by abbreviation p., pages, or pp.).

After a major revision, give the publication a day rest, then carefully check the references.

Summary: Getting references right is a tedious job. Putting all the points, commas, brackets, italics, and so on as they are requested in the guidelines can get very boring. However, getting your references right is a crucial part of your publication. Also make sure that your scope of references is well-balanced. Enjoy writing!