Dissemination of PhD Dissertation Research by Dissertation Format: A Retrospective Cohort Study

Abstract

Purpose:

The study purpose was to compare dissemination of PhD dissertation research by dissertation format: traditional (five-chapter document providing a complete and systematic account of the PhD research) versus an alternate (substudy [document containing three smaller studies but not written as stand-alone manuscripts] or publication [document containing three or more related manuscripts intended for submission or published in a peer-reviewed journal]) format.

Design:

A retrospective study of all PhD dissertations (1999–2019) from one research intensive school of nursing.

Methods:

Following identification of graduates via the school’s PhD database, we searched ProQuest and PubMed databases for the dissertation and first authored peer-reviewed publications of each graduate to determine dissertation format, study design, timing and number of dissertation research publications, and inclusion of dissertation sponsor in authorship. Data were analyzed using descriptive statistics and Wilcoxon rank sum tests.

Findings:

Of 113 graduates, 80 (70.8%) employed a traditional format, with the remaining graduates structuring dissertations using an alternate (substudy [n = 12], publication [n = 21]) format. Of those using the traditional format, 33 graduates (41.3%) never published dissertation research findings in a peer-reviewed journal. For those who published their dissertation research in a peer-reviewed journal, time to first publication was 1.4 ± 2.1 years (median 1.6 years) following degree conferral. In contrast, all graduates who utilized alternate formats published one or more components of their dissertation research with shorter time to first published manuscript (−0.6 ± 1.1 years; median −0.5 years; p < .001). Number of peer-reviewed publications was higher for those who utilized an alternate format compared to the traditional format (2.9 ± 1.5 [median 3.0] vs. 1.8 ± 1.1 [median 1.0], p = .001). Acknowledgment of the sponsor’s contribution via publication authorship was higher for those using an alternate format compared to the traditional format (100% vs. 70.2%).

Conclusions:

Number and timeliness of peer-reviewed publications stemming from dissertation research was higher for PhD graduates who utilized an alternate dissertation format. Alternate dissertation formats should be encouraged by PhD programs as one means to improve dissemination of PhD nursing research.

Clinical Relevance:

Dissemination of PhD research through peer-reviewed publications promotes the continued development of nursing science to inform nursing practice and advances the career trajectory of PhD graduates.

Dissemination of research via peer-reviewed publications is important to the profession and to the career development of the emerging nurse scientist. While the traditional five-chapter dissertation format containing an introduction to the problem (chapter 1), review of literature (chapter 2), methods and materials (chapter 3), research results (chapter 4), and discussion and conclusion (chapter 5) was once the sole option for students in research-focused graduate programs leading to a doctor of philosophy (PhD) degree, over the past decade there has been rapid uptake of the publication format dissertation both in nursing (Graves et al., 2018) and other fields (Stock & Siegfried, 2013). The publication format dissertation, also known as the three-paper format, differs from the traditional format dissertation in that chapters 2, 3, and 4 are reports of individual studies prepared as manuscripts consistent with the format specified by the targeted peer-reviewed journal. As recently as 10 years ago, fewer than 20 schools of nursing offered a publication format dissertation option to their PhD students (Baggs, 2011; Nolan et al., 2008; Robinson & Dracup, 2008). Since then, the number has more than doubled. Of 79 schools of nursing offering PhD programs who responded to a recent survey, 70% offered the publication dissertation format option. Both format options were offered by 59% of the schools, with less than one fourth of schools offering the traditional option only (Graves et al., 2018).

Integrated within a program of PhD study, the publication format dissertation has the potential to offer substantial benefit both to the student and faculty mentor. For the student it provides a first-hand experience in journal selection and discussion about authorship early in the process (Broome, 2018; Chyun & Henly, 2015). The student who writes a publication format dissertation benefits from the opportunity to identify journals appropriate for the research, develops skills in clear, concise scientific writing, avoids unintended plagiarism, benefits from external peer-review critique, and revises the manuscript based on reviewer feedback (Foster, 2009; Francis, Mills, Chapman, & Birks, 2009; Gross, Alhusen, & Jennings, 2012; Kearney, 2015). These experiences are less likely to be encountered during predoctoral education by those who write a traditional dissertation. One or more first authored peer-reviewed publications may provide a competitive edge for a postdoctoral fellowship or tenure track faculty position (Freeman, 2018). For faculty, co-authorship with their dissertation advisee is a demonstration of mentorship (Gross et al., 2012). On the other hand, concerns have been raised about potential copyright issues, need for additional help with writing by students, and faculty burden in providing that assistance when the publication option is pursued (Broome, 2018; Kearney, 2014; Robinson & Dracup, 2008).

Although there has been discourse about the potential pros and cons, little is known about the dissemination of research findings as peer-reviewed publications for those who have opted to structure their dissertation research as distinct publications. The purpose of this study was to examine the timing, number, and authorship of peer-reviewed publications stemming from dissertation research of PhD graduates at one school of nursing by dissertation format: traditional five-chapter versus an alternative (substudy or publication) format. The substudy and publication formats are similar since both contain dissertation research presented as three smaller studies. However, the formats differ in the sense that the publication format presents each study as a standalone manuscript, whereas in the substudy format the results chapter is limited to findings of each study with interpretation of findings from the three studies found in the last chapter of the dissertation.

Methods

We analyzed data from a retrospective cohort of 113 Columbia University School of Nursing PhD graduates (May 1999 to February 2019). The program began as a Doctor of Nursing Science (DNSc) program in the mid-1990s, with its first graduate earning the degree in May 1999. In 2008 the program transitioned from a DNSc to a PhD program. With this change, all students were required to maintain full time enrollment status as per Columbia University policy. First and second year students in good standing who had entered the DNSc program were offered the option to either continue the DNSc curriculum or to complete the additional requirements of the new PhD nursing curriculum (one additional two-credit course, Interdisciplinary Research Models). The rigor of the dissertation requirement was not altered with change in terminal degree. The first PhD degree in nursing was awarded in October 2009. The history of doctoral education for nurses in the United States (US), including similarities between the DNSc and PhD degrees, recent phasing out of DNSc programs, and their replacement with PhD programs in nursing, has been summarized by Reid, Ponte, and Nicholas (2015). Similar to other universities (e.g., University of California at San Francisco), in 2011 Columbia University retroactively awarded a PhD degree to 58 DNSc graduates without any additional course requirements. From program inception, while the majority of students employed a five-chapter traditional style dissertation format, select students (e.g., students who had successfully competed for National Research Service Award F31 predoctoral fellowships) were offered the opportunity to format their dissertation using an alternative format.

In 2014 the PhD program curriculum was modified with the expectation that PhD students would register for core, elective classes or dissertation advisement during fall, spring, and summer semesters with the goal of accelerating time to graduation, making it possible to complete the program in 3 years. The minimum number of credits required for graduation remained unchanged. Also, in 2014, a process for a publication option consisting of three publishable papers was formalized and included in the PhD student handbook as a choice for all Columbia Nursing PhD students. The decision regarding dissertation format is made by students in collaboration with their faculty mentor and declared at the time of the students’ dissertation proposal defense. Table 1 provides a summary of requirements for the dissertation proposal by format type. For those choosing a publication format option, the target journal and an alternate journal for each proposed paper as well as authorship are determined at the time of proposal defense. The first manuscript is a review of literature in the topic area of the research using systematic review, integrative review, scoping review, or qualitative meta-synthesis methods. Topics for the second and third manuscripts are determined by the student in collaboration with their dissertation committee and may include a methods paper, a quantitative study, or a qualitative study. For both dissertation formats, a fifth chapter will be written following completion of the research prior to the dissertation defense. Chapter 5 includes summarization and discussion of research findings within the context of what is known in the field and the implications of the research for policy, research, and practice.

Table 1.

Requirements for Written Dissertation Proposal Defense by Format Type

Dissertation chapter	Traditional format	Publication format
1	Introduction to overall topic, including significance, conceptual underpinnings, research questions, and study aims	Same as traditional formation with addition of plan for three manuscripts, logical link across studies, for each manuscript: targeted journal and 2 alternate journals identified, proposed authorship
2	Literature review	Manuscript 1: completed systematic or integrative review of literature either under review or ready for submission to targeted journal
3	Methods, including plans for data analysis fully described	Manuscript 2 (e.g., quantitative study): introduction, study purpose, and methods fully described
4	Findings of the dissertation research (not applicable at the time of dissertation proposal defense)	Manuscript 3: introduction, study purpose, and methods fully described (e.g., qualitative study

Note. Summarized from Appendices F1 and F2, PhD Program Student Handbook, 2018–2019 available at http://nursing.columbia.edu/sites/default/files/documents/phdstudenthandbookrevaugust2018.pdf

In the US, federally funded T32 training programs provide sources of funding to support the education of PhD students as well as to establish an infrastructure within the school for specific types of research. Currently, there are 21 National Institute of Nursing Research–funded T32 training programs in the US (National Institute of Nursing Research, n.d.). Since PhD program inception, the school has been awarded three distinct T32 training programs in informatics (T32 NR007969, 2002–2022), infection prevention (T32 NR013454, 2007–2018), and comparative effectiveness research (T32 NR014205 2013–2023). Many students have been supported by these training grants during their doctoral studies.

Data Sources and Variables

Data for this retrospective cohort study were collected from three sources: Columbia University School of Nursing PhD Microsoft Access database, ProQuest dissertations & theses at Columbia University, and PubMed (U.S. National Library of Medicine, National Institutes of Health). The PhD database was used to identify all program graduates from 1999 through February 2019. From the PhD program database, name, dissertation sponsor (faculty member who provided mentorship to the student in planning, implementing, and reporting the dissertation research), date of degree conferral, dissertation title, and degree(s) earned prior to entry to the PhD program (bachelor of science in nursing [BSN], master of science in nursing, or master of public health) were extracted for each graduate since program inception (J.K.). Each dissertation was retrieved using the ProQuest database. At Columbia University, dissertations are available via ProQuest within a few weeks of final submission of the dissertation. The abstract, table of contents, and acknowledgments sections were read for general sense of the research aims, study design (quantitative, qualitative, mixed methods), topical area of study, and whether the graduate had collected original data as part of dissertation research or used a secondary source such as national or state level data (E.H., K.J.). The dissertation format was categorized as either a traditional or alternate (substudy, publication) format based on the dissertation abstract and table of contents (A.S.). When the name of the graduate’s dissertation sponsor was unrecorded in the PhD database, the name was retrieved from the acknowledgments page of the dissertation. Finally, PubMed was searched in September 2017 and updated in March 2019 to identify the number of first authored peer-reviewed publications that were directly related to the dissertation research. We excluded publications that did not stem from the students’ dissertation research. The year and month of first publication and year and month of the last publication were retrieved, as were the name of the journal in which the manuscript was published and whether the dissertation sponsor was included as a co-author or senior (last) author on the publication. If the publication was available online ahead of print, this date was selected. Journals were categorized as nursing, interdisciplinary, or other discipline (E.H., K.J., A.S.). Using date of degree conferral and date of first publication, we estimated time to first publication. For those with more than one dissertation research publication, dates of first and last publication were used to estimate time from first to last publication. Time to graduation was estimated using date of degree conferral and date of program entry.

Data Analysis

Data were analyzed using descriptive statistics; dissemination of research via peer-reviewed publications and time to degree conferral by dissertation format (traditional vs. alternate [substudy and publication] formats) were compared using the Wilcoxon sign rank test. SAS 9.4 statistical software (SAS Institute, Inc., Cary, NC, USA) was used for all analyses. A p value of <.05 was considered statistically significant.

Results

Of 113 PhD graduates, the majority (70.8%) had formatted their dissertation as a traditional five-chapter document, with the remaining graduates using alternate (substudy [10.6%] or publication [18.6%]) formats. Figure 1 illustrates the shift from use of the traditional dissertation format to the publication format and time to graduation by dissertation format over the 21-year period. Across formats, median time to graduation decreased from 4.7 years during the period 1999–2004 to 3.5 years during 2015–2019 (data not shown). Across time intervals, use of the publication format option did not increase time to program completion.

Figure 1.

Dissertation format (A) and time (years) to degree conferral (B) over a 21-year period.

Table 2 provides details regarding the characteristics of graduates and their dissertations by format type. Across formats, the majority of graduates had entered the PhD program after achieving a master’s degree in nursing. For those who used the traditional dissertation format, quantitative study designs were most frequently employed (68.8%), with many using existing national or state level data sources. More than half of those who used an alternative format (substudy [58.3%], publication [61.9%]) employed mixed-methods study designs. Across dissertation formats, the majority of graduates had collected original data for all or part of their dissertation research. Overall, topic areas for dissertation research varied widely (Figure 2), with some (e.g., informatics [16.1%] and infection prevention [8%]) being very consistent with the T32 research training efforts of the school. Notably, approximately one third (35.7%) of dissertations were categorized as “other,” reflecting both the diversity of research interests by PhD graduates and the broad expertise of faculty mentors to guide the research.

Table 2.

Characteristics of Dissertations by Format Type

	Dissertation format
	Traditional		Alternate format
Variable	N = 80		Substudy N = 12		Publication N = 21
	n	%	n	%	n	%
Format type	80	70.8	12	10.6	21	18.6
Educational preparation at PhD program entry
BSN to PhD	3	3.8	3	25.0	3	14.3
MPH	5	6.2	3	25.0	4	19.0
Master’s degree in nursing	72	90.0	6	50.0	14	66.7
Study design
Qualitative	14	17.5	2	16.7	--	--
Quantitative	55	68.8	2	16.7	8	38.1
Mixed methods	8	10.0	7	58.3	13	61.9
Unable to categorize	3	3.7	1	8.3	--	--
Involved primary data collection
Yes	47	58.8	10	83.3	16	76.2
Unable to determine	2	2.5	1	8.3	--	--
Number of first authored publications
No publications	33	41.3	--	--	--	--
1–2 publications	36	45.0	3	25.0	11	52.4
≥3 publications	11	13.7	9	75.0	10	47.6
Sponsor as co-author or senior author	33	70.2	12	100.0	21	100.0
Journal audience
Number of publications	89		39		53
Nursing	37	41.6	4	10.3	19	32.8
Interdisciplinary	51	57.3	35	89.7	32	60.4
Other field	1	1.1	--	--	2	3.8

Note: BSN to PhD = a student entering a PhD program in nursing following completion of the Bachelor of Science degree in nursing; MPH = a student entering a PhD program in nursing following completion of one or more degrees in nursing and a master’s degree in public health.

Figure 2.

Dissertation topic areas by dissertation format.

Peer-Reviewed Publications

Of those who utilized the traditional dissertation format, 58.7% of graduates had published one or more peer-reviewed research articles, whereas all graduates using the alternate formats disseminated their research via peer-reviewed publications. In total, 181 peer-reviewed publications were retrieved. The number of first authored publications stemming from dissertation research was higher for those who utilized alternative formats versus the traditional dissertation format (2.9 ± 1.5 [median 3.0; range 1–7] vs. 1.8 ± 1.1 [median 1.0; range 0–6], p = .001). Across dissertation formats, graduates who had earned a master’s degree in public health prior to PhD program entry, on average, had disseminated more peer-reviewed publications (2.7 ± 2.0; median 2.0, range 1–7) compared to those who had entered the program either as BSN to PhD (2.3 ± 1.9; median 2.0, range 0–5) or with a master’s degree in nursing or another field (1.4 ± 1.4; median 1.0, range 0–6; p = .02). Of graduates who had disseminated their research via peer-reviewed publications, the dissertation sponsor was included as either a coauthor or senior author in 70.2% of publications stemming from traditional dissertation formats compared to 100% of publications using an alternate format. Across dissertation formats, graduates most frequently published their work in journals targeted to an interdisciplinary readership. Manuscripts were published in more than 90 journals.

Timing and Dissemination of Dissertation Research

Time to first publication was significantly shorter for those using alternate formats compared to the traditional dissertation format (−0.6 ± 1.1 [median −0.5] years vs. 1.4 ± 2.1 [median 1.6] years; p < .001), with no differences in duration between publication of the first and last dissertation publication (1.8 ± 1.0 [median 1.7] years vs. 2.0 ± 1.6 [median 1.3] years; p = .70). There were no differences in timing or number of publications by educational preparation of the graduate at time of program entry.

Discussion

In this sample of PhD graduates from one school of nursing, dissemination of dissertation research via peer-reviewed publications was both greater in number and the timing of the first peer-reviewed publication occurred approximately 2 years earlier for those with dissertations structured in an alternate format compared to those whose dissertations were written using the traditional five-chapter format. For more than half of the PhD graduates using an alternate format, publication of at least one manuscript published in a peer-reviewed journal occurred prior to defense of the dissertation. Acknowledgment of the dissertation sponsor’s contribution to the research via either co-authorship or senior authorship was more frequent when an alternate dissertation format was employed.

While the majority of graduates disseminated their research via peer-reviewed publications, approximately 40% of those who used the traditional dissertation format did not. This may be because the graduate either did not attempt to submit manuscripts of their work or was unsuccessful in doing so. However, the proportion of PhD graduates with unpublished dissertation research in our sample was significantly lower than what has been reported by others both in nursing (Kearney, 2017) and in other fields such as psychology (Evans, Amaro, Herbert, Blossom, & Roberts, 2018), where the proportions of unpublished dissertations were 76% and 74.4%, respectively. An unpublished dissertation is a lost opportunity for both the graduate and the scientific community at large because research findings reported in dissertations are less likely to be cited in academic journals (Thomas, 2015) or included as part of systematic reviews of literature (Moyer, Schneider, Knapp-Oliver, & Sohl, 2010). Envisioning dissertation research as publishable manuscripts at the time of proposal defense helps the PhD student to think about journal selection, negotiate authorship, and envision the end goal of dissemination right from the beginning.

Faculty support and mentorship is critical to success for all PhD students. Students using the publication dissertation option have additional mentorship needs relating to the publication process, including journal selection, criteria for authorship, keeping to strict word limits, the manuscript submission process, and responding to the feedback of peer reviewers. Evidenced by the publication rate of graduates who opted for the publication format in our sample, mentorship in writing and the publication process was both available and sufficient to foster PhD student success. However, this may not be universally true across all PhD programs in nursing. In a recent survey of PhD nursing programs, lack of PhD faculty support was the reason most frequently reported by schools who chose not to adopt a publication format dissertation option (Graves et al., 2018). Our T32 training grants have provided tuition and stipend support for many of our predoctoral students. Most of our PhD faculty currently have federally funded projects and are actively engaged in disseminating their research via peer-reviewed publications; this may not necessarily be true for all schools that have PhD programs. Further, in 2013 the school implemented a manuscript writing workshop for faculty in which manuscripts are “workshopped” and receive feedback from peer reviewers who are naive to the work. The workshop, now open to postdoctoral students and select PhD students with a completed manuscript ready for feedback, is an additional resource for PhD students (Kulage & Larson, 2016). Schools of nursing need to consider faculty and school resources when implementing an alternate format dissertation option.

Our study has several limitations. The sample included only one school of nursing and may not be broadly representative. We did not directly contact graduates regarding their peer-reviewed publications of dissertation research; instead, we relied on an electronic database search. Peer-reviewed manuscripts may have been published in journals that were not indexed in PubMed and therefore missed. Other forms of dissemination such as conference poster and oral presentations were not examined. It is also possible that the number of publications, particularly of more recent graduates, may be underestimated.

Conclusions

Findings of this study demonstrate outcomes from one research-intensive school of nursing. PhD graduates who opted to employ an alternate dissertation format had both earlier dissemination of their research and a greater number of peer-reviewed publications of work completed as part of their dissertation. This illustrates the potential for increased dissemination of nursing research with greater uptake of the publication dissertation format that is occurring nationwide in schools of nursing. Future research in multisite samples is needed both to confirm and expand upon these findings by examining other forms of research dissemination. Dissemination of PhD research through peer-reviewed publications promotes the continued development of nursing science to inform nursing practice and advances the career trajectory of PhD graduates. Alternate dissertation formats should be encouraged by PhD programs as one means to improve dissemination of PhD nursing research.

Please gray-box Clinical Resources.

Clinical Resources

• American Association of Colleges of Nursing. PhD Education. http://www.aacnnursing.org/Nursing-Education-Programs/PhD-Education

• Columbia University School of Nursing. PhD program student handbook. http://www.nursing.columbia.edu/sites/default/files/documents/phdstudenthandbookrevaug2018_0.pdf

• National Institute of Nursing Research. Institutional training programs (T32). https://www.ninr.nih.gov/researchandfunding/funded-ninr-grants-collaborative-activities/institutionalfile

• Reid Ponte, P., & Nicholas, P. K. (2015). Addressing the confusion related to DNS, DNSc, and DSN degrees, with lessons for the nursing profession. Journal of Nursing Scholarship, 47(4), 347–353. doi:10.1111/jnu.12148