Reference Accuracy Among Research Articles Published in Research on Social Work Practice

Background

According to the Publication Manual of the American Psychological Association (APA; 2009, p. 9), “Authors of scientific articles contribute most to the literature when they communicate clearly and concisely.” This clarity includes correctness of content in the reference list. The significance of reference accuracy in published work is reflected in a number of conditions. It reveals the rigor of quality writing, at least in the editing phase, in a profession’s scholarly output (e.g., journal publications and public reports). It allows the author to acknowledge another’s work in a professional and ethical manner. In turn, it allows the reader to search for sources embedded in an article without confusion or delay. Conversely, the lack of accuracy in reference reporting potentially undermines all of the aforementioned respects.

In the first study of its kind among social work journals, ¹ Spivey and Wilks (2004, p. 281) assessed reference accuracy and highlighted a number of studies from other disciplines to emphasize the following point: “[I]mplications of citation errors are serious.” They examined the top five circulated journals (at that time) in social work and noted a substantial amount of errors among a random sample of published references; the number of errors paralleled results previously reported in other fields. Approximately a decade later, a follow-up study was conducted with similar methodology, noting similar rates of errors among the previously examined social work journals (Mitchell-Williams, Skipper, Alexander, & Wilks, 2015). Research on social work practice (RSWP) was a journal omitted from these two studies because, at the time of Spivey and Wilks’ (2004) work, RSWP was not among the top five circulated journals in social work. The current study examined RSWP articles primarily for two reasons. First, RSWP exhibits a strong 5-year impact factor compared to its peers of social work and related journals (many of which have no impact factors)—1.580 as noted in a 2014 comparative list of said journals (Leung & Cheung, 2014). Second, RSWP has promoted the importance of reference accuracy in social work, evidenced by two publications related to such (Mitchell-Williams et al., 2015; Spivey & Wilks, 2004).

Purpose and Hypotheses

The purpose of this study was to assess reference accuracy in research articles specifically housed in RSWP. Following the template set forth by Spivey and Wilks (2004), two of their three hypotheses were adopted for this study:

Hypothesis 1: The age of the reference is relevant to reference accuracy; that is, we expect a significant relationship between the age of the reference—pre-2010 versus post-2010—and the likelihood of reference errors.

Unlike the Spivey and Wilks’ study, the current study was unable to include a third hypothesis because the single journal sampling (see Method section) disallowed multiple journal analyses.

Method

Results

Descriptive Statistics

Among the entirety of the sample (N = 500 references in RSWP), approximately 27% (n = 134) contained errors, totaling 173 errors within the entire sample. Looking deeper, almost 21% of references (n = 104) contained one error, and 6% of references (n = 30) contained two or more errors. Within the total sample, the mean number of errors per reference was 0.35.

The most common errors were identified in the article title field (13.2%), followed by author field (11.8%), journal title field (3.4%), pagination (3.2%), publication year (1.6%), and the volume number (1.4%). More specifically, the most commonly made error was article title misspellings (5.6%). Frequencies and valid percentages of error types within the references containing one or more errors are illustrated in Table 1.

OPEN IN VIEWER

Table 1.

Frequencies and Percentages of Error Types Within RSWP Published References.

Error Type	Frequency	Valid (%)	Total	Percentage of Total
Article title			66	13.2
Misspelling	28	5.6
Omitted word(s)	18	3.6
Omitted grammatical notation	20	4
Author field			59	11.8
Misspelling	25	5
Omitted name(s)	12	2.4
Inaccurate/omitted initial(s)	20	4
Order of authors	2	0.04
Journal title			17	3.4
Misspelling	8	1.6
Omitted word(s)	6	1.2
Omitted grammatical notation	4	0.8
Pagination			16	3.2
Publication year			8	1.6
Volume number			7	1.4

Hypothesis 1

Following the original hypotheses established by Spivey and Wilks (2004), a Pearson’s χ² test was conducted to evaluate the relationship, if any, between the age of the reference and the existence of reference errors. The results of this test were significant, χ²(1, N = 500) = .099, p < .05. Details from the two-way contingency table analysis of this test are displayed in Table 2.

OPEN IN VIEWER

Table 2.

Cross Tabulation: Age of Reference and Existence of Reference Error.

	Error
	No	Yes	Total
Age
Pre-2010	133	59	192
Post-2010	233	75	308
Total	366	134	500

Note. χ²(1, N = 500) = .099, p < .05.

Hypothesis 2

To evaluate the relationship, if any, between the number of authors in a given article (single vs. multiple) and the existence of errors, a Pearson χ² test was conducted. Results of the test were significant, χ²(1, N = 500) = .227, p < .001. Details from the two-way contingency table analysis data of this test are displayed in Table 3.

OPEN IN VIEWER

Table 3.

Cross Tabulation: Number of Authors in Reference and Existence of Reference Error.

	Error
	No	Yes	Total
Author number
Single	144	24	168
Multiple	222	110	332
Total	366	134	500

Note. χ²(1, N = 500) = .227, p < .001.

Discussion

APA (2009, p. 180) affirmed the notion that the accuracy of references is linked with the “credibility as a careful researcher.” In order to improve reference accuracy, thereby enhancing the credibility of scholarly inquiry, researchers have offered solutions to the problem of reference list inaccuracy. In medical literature, Fenton, Brazier, De Souza, Hughes, and McShane (2000) suggested that the understanding of shared responsibility among authors and editors may lead to increased accuracy within reference content. APA (2009) appeared to concur with Fenton and his colleagues with the following account: “The author, editor, and publisher share responsibility for the ethical and efficient handling of a manuscript. This responsibility begins when the editor receives the manuscript and extends through the life of the published article” (p. 225).

Spivey and Wilks’ (2004) original work cited two sources that continue to hold poignancy decades later regarding the responsibilities of reference list accuracy among authors and journal editors:

Foreman and Kirchoff (1987) suggested that authors verify references against the original source and that authors check with each amended draft to address errors that may be embedded within the article.

Another suggestion is for authors to abide by, and for editors to promote vigorously, a fewer number of references unless a larger number is unequivocally warranted. Literature (e.g., Adair & Vohra, 2003; Safer & Tang, 2009.) in this area supports the notion that manuscripts contain a plethora of preventable citations/references because of the specious assumption of a more scientific appearance, thereby making it appear more presentable to journal reviewers. Safer and Tang (2009, p. 52) highlighted this point, stating that authors are encouraged “to include more references, particularly in the introduction, for the sake of credibility and appearance, rather than for specific substance.”

In addition to the above implications for editors and authors, implications exist within the educational setting. A 1991 study by Peden (as cited by Spivey & Wilks, 2004) attested that precision in reference preparation should fall on the shoulders of instructors and students. As authors of the current study, we suggest that social work educators adequately prepare students for scholarly writing (and by extension, scholarly publication) by holding students accountable. We recommend more emphatic instruction in the editing phase of scholarly writing, which would allow justly for more rigorous grading penalties for reference and other grammatical/structural errors. Accordingly, we assert that the APA Publication Manual (APA, 2009) should be a required text, rather than recommended, in every social work class. To ease the problem of reference inaccuracy, Greasley (2011) offered a solution that may be helpful for students—the utilization of bibliographic software. For example, Word software (Microsoft, 2015) enables one to create a database of references, which can subsequently be inserted as a bibliography or reference list.

Keeping these educational sentiments in mind, increased errors among published references in the past few years may reflect technological advancements. For example, writers can search peer-reviewed articles from Internet research databases (e.g., ESBSCOhost, Google Scholar, and ScienceDirect). Such easily accessible and user-friendly search conduits provide often APA formatted versions of references, allowing authors to copy or import these citations directly into their reference lists or into citation managers such as EndNote. While allowing authors to quickly identify relevant research, research databases oftentimes have incorrect formatting or citation elements which may not be corrected, even when imported into a citation manager, if left unchecked by authors. If researchers transcribe references by hand, they may be more likely to check citations for accuracy rather than import them without checking for editing errors. The relatively high rate of reference list errors in the current study indicates that many researchers may not be checking or transcribing citations by hand. At worst, the results perhaps point to an even deeper problem of authors citing without reading or possessing the cited article. Indeed, this problem has been studied in the past, with estimates that less than 25% of authors actually read the original source when citing a study (Simkin & Roychowdhury, 2006). Ease of access through these technological advances may further exacerbate the problem of authors copying references without consulting the original study by allowing authors to easily copy and paste the improper citations of other authors, thereby hindering other scholars from finding the original source. Recognizing that technology influences the efficiency of information gathered, we assert the importance for educators and journal editors to encourage scholars to take caution when utilizing these online reference importation methods.

Conclusion

APA (2009, p. 180) informs us that reference accuracy helps establish one’s “credibility as a careful researcher.” To some, this credibility of research carefulness among all disciplines has markedly declined. In reviewing literature for this study, we found at least one journal editor who intimated critical mass level in the carefulness of scholarly writing (Heatwole, 2008, p. 159): “The standard of writing in current scientific journals has reached an all-time low, in terms of both poor grammar and imprecise communication.” The current study examined and revealed the carefulness, or lack thereof, among published research articles—specifically among reference content—in RSWP. We appreciate the commitment from RSWP over the past decade in disclosing reference error rates among top social work journals, including itself. The risk in doing such is exposure of writing carelessness embedded within the profession’s scholarly work. Conversely, the potential reward of exposing reference errors in social work research, we feel, is greater: increased awareness of the problem that leads to genuine self-reflection and corresponding enhanced editing skills from social work scholars.