|APA Publication Style (6th Edition)
To start with, let’s deal with some of those tricky questions:
Why do we need to know about this?
Most of your work is supposed to conform to APA publication guidelines, as set out in the APA Publication Manual (APA, 2009). Also many journals and other organizations require that reports are set out according to APA guidelines.
Erm…excuse my ignorance, but…what, or who, is the APA?
The American Psychological Association. Not to be confused with the American Psychological Society, which is now the Association for Psychological Science. Or the Advertising Producers Association, or the Albanian Peoples -
Goddammit, we’re British - why should we be interested in American publication style?
Well, the Swiss Psychological Society recommends it, so we mustn’t be too proud. And, for that matter, the Journal of Ecotourism, among many non-psychological publications. (Yes, I know that’s technically American…)
OK then… so how does APA style differ from, say, BPS style?
To tell the truth, there’s not a great deal of difference. Minor formatting foibles; whether you put the volume number in bold in your references, that kind of thing. In practice, once you’ve mastered APA style it’s pretty easy to adapt to other systems. And there are plenty of them. For example, Sage and Blackwell (who both publish psychology books and journals) have their own house style that differs (in places) from BPS and APA. Generally speaking, the more important issues in APA style – the reporting of statistical results – are fairly consistent across the board.
I’m doing an IPA study, what do the APA say about that?
Guidelines for preparing reports using qualitative research are mysteriously lacking in the APA manual. This may be related to the mysterious absence of qualitative research from APA journals. As a rule of thumb, if the APA don’t advise on something, it’s probably safest to follow the format of a BPS journal or a reliable (psychology) journal from one of the leading UK publishers. The APA do, however, provide guidance on the layout of block quotes. This will be covered later in the Quotations section.
Key Aspects Of APA Publishing Style
The APA manual has advice on everything from preparing your manuscript for publication down to where to use [brackets] as opposed to (parentheses) and where to put full stops in references. In this document we’ve tried to boil it down to just a few highlights based on our own experiences of reading trainee work.
Structure Of The Paper And Headings
Papers will generally be structured in sections denoted by headings. APA style has five levels of headings, the formatting of which is described below. This formatting is new in the 6th edition of the Publication Manual. These levels of headings must be used consecutively starting at Level 1.
Most research papers will be structured in the traditional Introduction, Method, Results and Discussion sections. These sections are often split up into subsections, some of which are optional. The type and level of headings generally used for each section in a one-study paper are shown below.
Sections and headings of a one-study paper
Heading description & level (where applicable)
Running head: CAPITALISED
Title Of The Paper
Heading = title text
The Title Of My Paper
(Other optional subsections)
Data Analysis (2)
NB: Some times Results and Discussion are combined. If so, a Summary and Conclusions section is often added
Quotations, Citation and Referencing
APA’s referencing system is based on the Harvard system, although with some departures. The main common feature is the citation of authors and dates in the text that link to an entry in an alphabetical list of sources at the end of the work. Other referencing systems require that you put such information in a footnote, and simply ask for a bibliography at the end of the report, or book chapter. Footnotes are permissible in APA, but only “if they strengthen the discussion” (APA, 2001, p. 202). Bibliographies, however, are redundant under anything based on the Harvard system. Of course you should all be highly familiar with the Harvard/APA approach from psychology undergraduate days. However there are still a few things to watch out for.
Quotations (p. 170-4)
Direct quotes can be from other published work or from interview transcripts. Quotes from published work must be referenced including page numbers. The APA isn’t explicit about interview transcripts, but a similar style of reference to participant and line number is also appropriate. Both types of quotations, of less than 40 words, can be embedded within the text. For example:
Interpreting these results, Msetfi (2009) suggested that the “therapists in dropout cases may have inadvertently validated parental negativity” (p. 792).
The study used “a rather unusual and regrettably cumbersome” method (Roen & Skirrow, 2006, p. 546).
However, if the quote is over 40 words, then this is a block quotation and should be displayed as a freestanding block of text that is indented ½ inch from the rest of the text.
Another participant said:
I think maybe if it wasn’t a strange situation. You know, like when you feel like you’re here – do you know what I mean? You’re here and you can’t come back. And right now it’s like, maybe next year, but right now I’m, I’ve just come here and it’s really strange. It’s like, I don’t want to make that change now. And they’re like, come on now. And it’s like, what do you know? (Rachel, line 22)
Note that these block quotations don’t include quotation marks and should be double spaced like the rest of the text. Finally if there are paragraphs within the block quote, you should indicate this by indenting within the block quote. Note also that that the reference (Rachel, line 22) comes after the final punctuation.
Citations in the Text (p. 174)
Usually it is sufficient just to cite a single source as evidence to support a claim, or as information for the reader.
A full review of the literature can be found in Simpson and Done (2002).
You might want to cite that source as an example (note, by the way, how the ‘and’ becomes an ampersand when put inside parentheses):
Studies that have used this technique (e.g., Simpson & Done, 2002)
However you might want to elaborate and cite some of those other studies:
Studies that have used this technique (Simpson & Done, 2002, 2004)
When you cite several studies, use semi-colons to separate them, arrange them alphabetically according to the first author names, and arrange works by the same first author chronologically). So:
Rachel Msetfi and colleagues have conducted several studies (Msetfi, Murphy, Simpson & Kornbrot, 2005; Msetfi, Murphy & Simpson, 2007).
N.B. The BPS Style Guide (BPS, 2004) recommends that you never cite more than three studies at a time in parentheses. It’s always a good idea to avoid gratuitous ‘stacking’ of unnecessary sources. A statement like:
A full review of the literature can be found in Murphy (2003)
will usually suffice if you have lots of examples that simply repeat the same type of evidence or information.
Sometimes awkward academics publish more than one work a year under the same name(s). If you need to cite both works, then you need to refer to them by letters, e.g.
One year, Jane Simpson clearly had too much time on her hands (Simpson, 2002a, 2002b).
Don’t forget to remember which letters refer to which sources when you draw up your reference list!
Where three or more authors are named on a publication, the standard practice is to include them all in the first citation.
First mention: Daiches, Verduyn and Mercer (2005) found that…
Second mention: Daiches et al. (2005) argued that
Subsequent mention in the same paragraph: Daiches et al. also noted… (NB no date required).
If there are six or more authors, use the et al. construction throughout unless there are two six plus author papers, with a similar authorship. In this case you would use as many authors as necessary to distinguish between the two papers. So:
Golding, Brown, Atkins, Slinger, Amor, Morgan and Munks (2010) suggested that….
Golding, Brown, Slinger, Atkins, Amor, Morgan and Munks (2010) argued that….
Because of the similarity between these two six plus author papers, they should be referred to in the text as:
Golding, Brown, Atkins at al. (2010) suggested that….
Golding, Brown, Slinger et al. (2010) argued that….
There is a handy table on p. 177 of the APA guide that helpfully summarises these in-text citations.
The Reference Section (p. 193).
The following formats should be familiar to you. However, there have been a number of changes, including the inclusion of something called a DOI number. DOI numbers are ‘digital object identifiers’ which provide a persistent link to that content on the internet. This allows readers to access the content in your reference list. DOI numbers are added to the end of the usual journal reference format in the form: doi: xx.xxxxxxxxxxx. This is shown in the first example here. If you do not have the DOI number, you are then required to provide retrieval information as in the second example shown below. Note that issue numbers are only necessary if each issue of the journal starts at p. 1. You would then reference as in the first example here.
NOTE: Both of the two following rules regarding DOI numbers and retrieval information will be integrated into DClinPsy academic submissions from 1st September 2010.
Journal article (if you have the DOI number)
Msetfi, R., Murphy, R., Kornbrot, D., & Simpson, J. (2008). Impaired context maintenance in mild to moderately depressed students. The Quarterly Journal of Experimental Psychology, 62(4), 653-662. doi:10.1080/17470210802486092.
Journal article (if you do NOT have the DOI number)
Wilde, D., & Murray, C. (2010). Interpreting the anomalous: Finding meaning in out-of-body and near-death experiences. Qualitative Research in Psychology, 7(1), 57-72. Retrieved from PsycINFO database.
Giles, D.C. (2003). Media Psychology. Mahwah, NJ: Lawrence Erlbaum Associates.
Daiches, A., Verduyn, C., & Mercer, A. (2005). Continuing professional development and the transition from ‘A' Grade to ‘B' Grade. In L. Golding & I. Gray (Eds.), Continuing professional development for clinical psychologists: A practical handbook (pp. 102-124). Oxford: Blackwell.
Note that if DOIs are assigned to books or book chapters, these should be added to the end of the reference in the same way as for journal articles.
Other sources include: dissertations, newspaper and magazine articles, conference papers, non-academic reports (e.g., to funders). The format varies according to the nature of the source, but for a typical example of a conference paper:
Giles, D.C., Pietrzykowski, S., & Clark, K.E. (2005, July). “The joy of ownership”: The psychological meaning of personal record collections. Paper presented at the Ninth European Congress of Psychology, Granada, Spain.
Online sources are much more complicated but most are covered in the APA 6th edition manual. So for example, explicit instructions are given about how to reference a blog post or a message posted to a discussion board. However it should be noted that very often the material you cite has landed on the Web via an offline source, or lacks a named author or publication date. The main rule-of-thumb is that you give as much information as necessary for a reader to locate the material, so typically you would add the date of retrieval after the author’s name (could be a screen name) and conclude with the words “Retrieved from” followed by the URL information in the address bar. Here are some specific examples though:
Nathan. (2010, April 30). Depressive Realism - A More Accurate Worldview But Not Necessarily a Healthy One. Retrieved from http://tryinghardtobreakfree.blogspot.com/2010/02/depressive-realism-more-accurate.html
Adelson, R. (2010, April). Probing the puzzling workings of depressive realism. Monitor on Psychology, 36(4). Retrieved from http://www.apa.org/monitor/apr05/realism.aspx
Don’t forget to keep a record of the content and references (save a copy of the webpage) before you submit your work, otherwise you may find that they have changed – the site itself may no longer even exist!
Reporting Statistical Information (p. 116)
The APA is very particular about the use of punctuation, abbreviation and typeface when it comes to statistical material. In the Appendix we have listed the most common symbols in their correct typeface. However most of the conventions are universal so they will apply to any scientific publication.
First, authors should include enough information to enable readers to understand the analysis. For an ANOVA, for example, you would expect to include means and standard deviations, F ratios, degrees of freedom, p-values, and a measure of effect size.
Note also which symbols are in italics, which material appears in parentheses, and where spaces occur in the text. For example, there’s no space between F and the left hand parenthesis denoting the degrees of freedom, but there is a space either side of the equals sign and between the p and its value.
The test of the main effect of condition was statistically significant, F(4, 191) = 4.28, p = .003, η2 = .07.
Also note how the information is incorporated into the sentence as if it were a verbal statement and is not surrounded by brackets, as it used to be in earlier versions of APA style.
Just a note about p values: some of you may have been threatened with dire consequences by psychology teachers for citing anything other than p < .05 or p > .05, the argument being that the statement of significance should simply relate to the set alpha level. However it is standard practice in these days of SPSS to record the precise p value, e.g. p = .02, if it is available. It seems silly not to. Also, note that when values of any statistic are less than zero, it is conventional not to cite the zero (i.e., .05 rather than 0.05). In spite of this preference towards citing exact p-values, there is an exception to this rule. If SPSS gives a p = .000 value, this should never be cited! SPSS produces zero values because the probability is so small and there is a lack of decimal places. Logically, there can never be a zero probability of type 1 error. Therefore, this should be reported as p < .001.
For other statistical procedures, notably regression, there will be different requirements. For multiple regression you will need to cite the overall R2, along with relevant beta coefficients (ß) of the specific predictors, and with hierarchical models you will need to cite the change in R2 at each level (represented as ΔR2). You may also need to cite beta coefficients across levels. You will almost certainly need, therefore, to construct a table for these data.
Tables and figures
The most important thing about tables and figures is that they are illustrations rather than analyses. Think of them as pictures accompanying the story. Don’t rely on them to tell the story for you! You should already have described the most important results in the text, so generally tables contain data that are too unwieldy to describe in words. There are no guidelines as to which content should go in a table and which into text: you need to decide this according to the size of your data set and the number of analyses. The guide gives examples of effective table layout for all types of analyses including regression, factor analysis, and basic descriptives amongst others. This is particularly useful if you are not sure on the best way to construct your table.
However there are guidelines on how to format tables and figures. Examples of tables and instructions for their appropriate constructions are given from p. 128 of the guide. Note that APA style tables do not contain vertical lines. However, it is still possible to organise your table in a coherent informative manner. There are suggestions on p. 129 of how to use horizontal lines to split your table appropriately.
The table title goes above the table. Note also, that when a table (or a figure) is referred to in the text, say as Table 3, it is capitalised. Another key thing to remember about figures is that the labels and titles go underneath the illustration. Also, for journal submissions in APA format, tables and figures are included in a separate section after the References, indicated in the text by the following type of entry:
Insert Table 5 here
Use of English
Much of the APA guide is devoted to minutiae relating to wording, grammar, punctuation and general editorial style. For example, there are sections on Continuity on Presentation of Ideas, Smoothness of Expression, Tone and Economy of Expression, which contain useful examples of how to avoid common writing errors.
A few general tips though:
It is important to use tense consistently but correctly. So for example, the past tense is appropriate for a literature review or for the description of procedures that have already taken place:
“The researchers have shown…”
Use of the past tense would also be appropriate in the results section:
“Levels of depression decreased significantly following….”
However, in the discussion it is more usual to use the present tense to discuss the implications of findings and conclusions:
“These findings indicate that …”
Numbers and words.
Never use 10 words where 1 will do! Or should that be ‘ten’ and ‘one’? In fact the rule of thumb is that any number consisting of more than one digit (i.e. 10 and above) should be written as a number and single-digit numbers as words. The same is true of ordinal numbers (23rd but third).
The greengrocer’s apostrophe.
Is it really apple’s and pear’s on sale today? Or might you ask ‘apple’s what?’
Indeed, there are only two occasions on which you need to use ’s in English.
1) To denote possession.
I need the client’s address. John’s address, not Gladys’s
I need the clients’ addresses. Not the women’s, only the men’s.
But not with possessive pronouns (his, hers, etc).
I need the list in its entirety.
2) To denote omission.
Where’s the book? It’s on the table.
So the greengrocer got it wrong: you never need to use ’s to denote a plural! This extends even to the use of acronyms (SDs, not SD’s).
What’s the difference between commas, full stops, semicolons and colons? It’s all a matter of degree.
Commas are used to denote the end of a clause, at the point where you might pause slightly in speech. They also help structure your sentences for the reader, so use them particularly if there is any danger of ambiguity.
The following day, the participants repeated the measure.
Semicolons and colons are like big commas, denoting longer pauses in spoken delivery. We often use commas instead of semicolons, which is not technically correct.
The participants in the first study were paid; those in the second were unpaid.
Technically, you can use a full stop in place of a colon, although it is usually best to avoid lots of short, snappy sentences that break up the flow of the prose.
There were two studies: the first involved payment, but not the second.
A sentence is not defined by its length. It depends entirely on your general prose style. I’ve been keeping it short and snappy up to now, but I’m liable to lapse into long-windedness at any moment and who knows, out come the semicolons and colons and all other kinds of paraphernalia, and then…
Bias in language
There are detailed guidelines on p. 70 – 77 on how to reduce bias in language. Common areas for bias include labels such as ‘borderline’ or ‘at risk’. These terms should be more carefully explained in the following way, such as ‘people with borderline personality disorder’ or ‘children at risk from school exclusion’.
APA (2001). Publication manual of the American Psychological Association (5th ed.). Washington, DC: American Psychological Association.
APA (2009). Publication manual of the American Psychological Association (6th ed.). Washington, DC: American Psychological Association.
For general tips on writing style, see also Robert Sternberg’s The Psychologist’s Companion: A Guide to Scientific Writing for Students and Researchers (3rd edition, Cambridge University Press, 2003).
Statistical abbreviations and styles (from APA, 2001)
α alpha, the probability of making a Type I error, Cronbachs alpha for internal consistency (reliability)
β Beta, probability of making a Type II error, also values of regression coefficients.
δ Cohen’s measure of effect size
Δ Delta or difference, meaning change
df Degrees of freedom
F Fisher’s F ratio (for use with ANOVA etc.)
H0 Null hypothesis
H1 Alternative hypothesis
M Mean (arithmetic average)
MS Mean square
MSE Mean square error
n Number in a subsample
N Total number in a sample
pr Partial correlation
r Pearson product-moment correlation
r2 Pearson product-moment correlation squared
rs Spearman rank correlation coefficient
R Multiple correlation
R2 Multiple correlation squared
SD Standard deviation
SE Standard error
SEM Standard error of mean
SS Sum of squares
sr Semipartial correlation
t Computed value of t test
T Computed value of Wilcoxon test
U Computed value of Mann-Whitney test
x Abscissa (horizontal axis in graph)
y Ordinate (vertical axis in graph)
z A standard score; difference between one value in a distribution and the mean of the distribution divided by the SD
α Alpha; Cronbach’s index of internal consistency
β Beta; standardized multiple regression coefficient
Δ Delta; increment of change
κ Cohen’s estimate of effect size
η2 Eta squared; measure of strength of relationship
Λ Lambda; Wilks’s multivariate criterion
χ2 Computed value of a chi-square test
ω2 Omega squared; measure of strength of relationship
As a rule of thumb abbreviations like, ANOVA, ANCOVA, MANOVA are not put in italics.