A guest post from Helen Colley from The University of Huddersfield.

I just got an enquiry from a colleague about whether the university has guidelines for research theses in relation to formatting quotations, whether from the literature or primary data from respondents. At first sight, this might look like a fairly mundane technical question – but there is far more to it than that! The question rang all my bells about how technical questions such as this can only be answered by going back to the philosophical underpinnings of our research, and our own quest for an appropriate authorly voice.

I’m always a bit bewildered why this particular question about formatting quotations comes up so often (and I get to see so much weird and wonderful formatting!). Doctoral researchers have presumably read scores if not hundreds of journal articles and books. Almost all of these use exactly the same protocols, so the answer is easy: ‘When in Rome, do as the Romans do’, i.e. just copy the same protocol you find everywhere else in the literature. You are becoming a researcher, so emulate the practices expected of researchers by publishers! This relates to the issue of learning to write by reading as a writer – analysing how other writers have constructed their texts and how we can emulate them, rather than just focusing on the content of their findings. Anyway, that’s a whole other issue… So let’s start off with some technical basics and how I approach them in line with publishers’ (and readers’) expectations.

Quotes from other authors of 3 lines or more should be in an indented paragraph, separated from the main text by a line space above and below, NO quotation marks, NO italics, citation comes after the final full stop of the quote and must include page number(s). This looks like this:

[M]uch of that literature fails to account sufficiently for the socio-cultural and socio-political context characterised today by the crisis of neo-liberal capitalism. The neoliberal belief system focuses on the individual as consumer in deregulated markets, including the labour market…(Benozzo and Colley, 2012, p.305)

In the original text, the quotation starts with a clause that is not necessary to reproduce when quoting it, therefore ‘much’ is not capitalised in the original – that is why it appears as ‘[M]uch’ in my quotation here: I’ve altered it to be grammatically correct in the context of my own text, and the square brackets show that I’ve changed the original. Also the original text has a colon and continues further after ‘labour market’ – but again, that is not relevant to my use of the quotation here, so I have an ellipsis of three full stops in a row ‘…’ to show this. If I had left out something in the middle of the quotation, I would show the ellipsis in square brackets like this: […]

Quotes from primary data should usually be in an indented paragraph, separated from the main text by a line space above and below, NO quotation marks. (Short quotes can be incorporated into the main text in inverted commas, but do not have to be.) Italics should NOT normally be used unless there is an exceptional reason why, and if they are, this should be explained in the methodology. Citation comes after the final full stop of the quote, like this:

I get lots of enquiries from students and supervisors about formatting formal texts such as theses and progress reports. (Helen, Director of Graduate Education)

There is no common rubric about any citation format for research respondents – this is something that the researcher should think through in relation to their overall methodology and how they, as an author, think it is best to present the quotes to their reader: (a) to engage them with the overall text; and (b) to ensure the reader has sufficient information about where the quote has come from to make a judgment about its internal validity and the strength of the evidence it presents as a warrant for the findings. For example, in an in-depth narrative study with just one or few respondents, it is usual to use a pseudonym to give a sense of the person, and often to add a descriptor(s) depending on the design of the study and what the reader needs to know about how that quote relates to the overall data set in the study. The example below might be appropriate for a very small narrative study, where the reader has had and is very likely to remember who the respondent is.

So let’s imagine we are doing a study of PGR students and their perceptions of writing. A quote from a very small-scale narrative study with four respondents, whom you have already described in detail, might look like this:

I read lots of texts and saw how they were formatting the quotes, but I just didn’t feel confident to use the same formatting myself without asking first. (Aisha)

If the study were still qualitative, but a bit larger, and included e.g. full-time and part-time EdD, PhD and MA by Research students, and those distinctions were important for the study, this would be appropriate:

I read lots of texts and saw how they were formatting the quotes, but I just didn’t feel confident to use the same formatting myself without asking first. (Aisha, MA by Research student, FT)

If a key issue in the study is the perceptions of students on different programmes, in different disciplines at different stages of their study:

I read lots of texts and saw how they were formatting the quotes, but I just didn’t feel confident to use the same formatting myself without asking first. (Aisha, Humanities, PT MA by Res, Year 1)

If the study included multiple interviews with research students because it was tracing change over time, then it would be helpful also to note which of the interviews it was e.g.

I read lots of texts and saw how they were formatting the quotes, but I just didn’t feel confident to use the same formatting myself without asking first. (Aisha, MA by Research student, interview 2)

OR

I read lots of texts and saw how they were formatting the quotes, but I just didn’t feel confident to use the same formatting myself without asking first. (Aisha, MA by Research student, 3rd year interview)

If there are many respondents, e.g. from a survey, and the reader is therefore unlikely to remember pseudonyms, or you may not have names attached to the responses, then it would be better to replace the ascription with a numeral one:

I read lots of texts and saw how they were formatting the quotes, but I just didn’t feel confident to use the same formatting myself without asking first. (PGR45)

Again, depending on the research questions, you might need to add to ‘PGR45’ which programme they were on, which year they were in, and/or which of multiple interviews with that student it was, so it will look more like a code, and instead of ‘PGR’ you could use M for MA by Research, P for PhD, E for EdD e.g. (M45 Y1, i/v1). The reader doesn’t need a sense of the individual person, but might be looking to see that the data comes from a good spread of respondents rather than from just the same ones over and over again.

All these decisions are things to be noted, with your justifications as to why you have made these choices, in your research journal – and you need a paragraph about it in your methodology chapter, so your reader understands what format you will be using (especially important if it appears as a code) and why.

There is also, of course, a big issue in terms of presenting qualitative data, and whether we present it ‘raw’ (with all the ums, ers, false starts, ungrammatical speech etc) or ‘cooked’ (tidied up a bit to read more coherently). This partly depends on research design – if you are doing discourse analysis, you’ll definitely need all the ums and ers and timed pauses etc. But if not, it can just make people look inarticulate on the written page when in fact they are just speaking the way that all of us do in everyday talk. On the other hand, sometimes we have a piece of data where the uncertainties or contradictions of what someone is saying can only be conveyed by including the ums and ers and false starts etc., and the reader needs to see that. And sometimes, the only way we can convey the authenticity of respondents’ voices is by using their actual speech, warts and all (see Geoff Bright‘s work where he directly transcribes the strong dialect and vernacular of his respondents in Nottinghamshire’s former pit villages). Harry Wolcott discusses these kinds of choices in excellent detail in his book ‘Transforming Qualitative Data‘, to which I always refer students on these matters.

Like all seemingly technical questions, then, this one about how to format primary data quotations and cite them is an opportunity for supervisors to lead doctoral researchers back to think about the philosophical and authorly implications of their overall methodology, and the fact that there can’t be any single common rubric for how we present our data, including even the form of the citations we use – we have to think and make choices about it. Those choices should be coherent with our overall methodology but also to think about our reader and what they need to know for the data to be really meaningful to them and allow them to make judgements about the strength of the evidence presented in support of the research claims.