and complexity of reasoning Research ethics in

Downloaded from jme.bmj.com on August 9, 2010 - Published by group.bmj.com

Research ethics in dissertations: ethical issues and complexity of reasoning S Kjellström, S N Ross and B Fridlund J Med Ethics 2010 36: 425-430

doi: 10.1136/jme.2009.034561

Updated information and services can be found at: http://jme.bmj.com/content/36/7/425.full.html

These include:

Supplemental Material References

http://jme.bmj.com/content/suppl/2010/07/26/36.7.425.DC1.html This article cites 20 articles, 9 of which can be accessed free at: http://jme.bmj.com/content/36/7/425.full.html#ref-list-1

Email alerting service

Receive free email alerts when new articles cite this article. Sign up in the box at the top right corner of the online article.

Notes

To order reprints of this article go to: http://jme.bmj.com/cgi/reprintform

To subscribe to Journal of Medical Ethics go to: http://jme.bmj.com/subscriptions


Research ethics

Research ethics in dissertations: ethical issues and complexity of reasoning S Kjellstro¨m,1 S N Ross,2,3 B Fridlund4 < Additional data are published

online only. To view these files please visit the journal online (http://jme.bmj.com). 1

Institute of Gerontology, School of Health Sciences, Jo¨nko¨ping University, Jo¨nko¨ping, Sweden 2 Antioch University Midwest, Yellow Springs, Ohio, USA 3 ARINA, Inc., Cincinnati, Ohio, USA 4 Department of Nursing, School of Health Sciences, Jo¨nko¨ping University, Jo¨nko¨ping, Sweden Correspondence to Sofia Kjellstro¨m, Institute of Gerontology, School of Health Sciences, Jo¨nko¨ping University, PO Box 1026, SE-551 11 Jo¨nko¨ping, Sweden; [email protected] Received 20 November 2009 Revised 22 February 2010 Accepted 2 March 2010

ABSTRACT Background Conducting ethically sound research is a fundamental principle of scientific inquiry. Recent research has indicated that ethical concerns are insufficiently dealt with in dissertations. Purpose To examine which research ethical topics were addressed and how these were presented in terms of complexity of reasoning in Swedish nurses’ dissertations. Methods Analyses of ethical content and complexity of ethical reasoning were performed on 64 Swedish nurses’ PhD dissertations dated 2007. Results A total of seven ethical topics were identified: ethical approval (94% of the dissertations), information and informed consent (86%), confidentiality (67%), ethical aspects of methods (61%), use of ethical principles and regulations (39%), rationale for the study (20%) and fair participant selection (14%). Four of those of topics were most frequently addressed: the majority of dissertations (72%) included 3e5 issues. While many ethical concerns, by their nature, involve systematic concepts or metasystematic principles, ethical reasoning scored predominantly at lesser levels of complexity: abstract (6% of the dissertations), formal (84%) and systematic (10%). Conclusions Research ethics are inadequately covered in most dissertations by nurses in Sweden. Important ethical concerns are missing, and the complexity of reasoning on ethical principles, motives and implications is insufficient. This is partly due to traditions and norms that discount ethical concerns but is probably also a reflection of the ability of PhD students and supervisors to handle complexity in general. It is suggested that the importance of ethical considerations should be emphasised in graduate and post-graduate studies and that individuals with capacity to deal with systematic and metasystematic concepts are recruited to senior research positions.

Research has a potential to encroach on people’s lives, autonomy and integrity. To prevent or mitigate the potential for such effects, the research community has created ethics codes and regulations, institutionalised ethics review boards and formalised ethics requirements in scientific journals.1e3 However, how do we know whether the formalisations of research ethics actually result in researchers’ ability to operationalise ethics in the ways intended? One way is to analyse how they write about research ethics. Including a well-written section about research ethics in a dissertation is important for several reasons. Compared to protocols written for research ethics committees, this section allows a comparison of the expected and actual research J Med Ethics 2010;36:425e430. doi:10.1136/jme.2009.034561

ethics as reflected in the entire research process. Scientific journals increasingly require that ethical considerations are elucidated, but most journals severely limit space for elaboration.4 Since studies have questioned the ethical skills of doctoral students, dissertations provide a forum for students to expound on ethics and enable an assessment of acquired proficiencies. One purpose of graduate school is to train doctoral students in skills necessary for future research careers, including more critical thinking and more complex reasoning. The quality and depth of the research ethics section is essential to examine whether a researcher has acquired necessary skills to reflect and report on ethics. Despite an increasing interest in research ethics, surprisingly little is known about the quality of research ethics in dissertations, particularly in nursing research. Research on written materials focuses primarily on research review boards5e9 and journalsdfor example, ethics guidelines10 and research ethics in articles.4 Research on Turkish nursing dissertations showed deficiencies in informing participants and protecting privacy.11 A study on Swedish nurses’ dissertations from 1987 to 2007 showed that an increase in occurrence and proportions of reported ethical considerationsand that the texts were short, had few references and covered a narrow range of topics.12 We found no other studies that address the design of the research ethics section and how different topics were combined. The study’s purpose was to examine which research ethical topics were addressed and how these were presented in terms of complexity of reasoning in Swedish nurses’ dissertations approved in 2007. The research questions were: Which research ethics issues are reported? How is the research ethics section organized around different ethical issues? How is the information coordinated in terms of the complexity of reasoning that structures the text? What is the relationship between ethical issues and complexity of reasoning in the text?

METHODS Design and methodological approaches The study used a mixed-methods approach to address the four research questions.13 We performed a qualitative content analysis and a quantitative analysis of the hierarchical complexity of ethics-related content. The quantification method was the Hierarchical Complexity Scoring System (HCSS) (Commons, et al, unpublished manual), which derives from the Model of Hierarchical Complexity, a mathematicsbased, formal general theory applicable to all actions in which information is organised.14 15All reasoning 425


Research ethics involves organising information. The theory and validated scoring method enable reliable measures of discrete stages of reasoning complexity.16e20 In accord with Swedish law, ethical approval was not obtained for this study,21 but ethical principles were used and issues were addressed in ongoing reflective processes.

Data collection The sample consisted of 64 dissertations from Swedish universities in 2007 (Appendix 1). The primary inclusion criteria were that the dissertation was written by a nurse and that it was a PhD dissertation (4 years of full-time studies). Suitable dissertations were identified from the Swedish Society of Nursing’s list of self-reported dissertations (n¼65) followed by a systematic comparative analysis with the Swedish National Library (n¼1). One of the self-reported dissertations discussed no research ethics and one was by an unsuccessful doctoral candidate: they were not included in the sample. Dissertation languages were English (n¼48), Swedish (n¼15) and Norwegian (n¼1). Dissertations were retrieved via full-text online access or as books from the university library.

Data analysis The dissertations were examined to identify research ethics sections, often under the subheadings “Ethical considerations” or “Ethical approval”. The texts were analysed for the topics addressed and how they were reported. An unstructured matrix of research ethics issues was created and grounded in the data. The coded texts were further analysed for subcategories through an inductive process. Descriptions of meanings of quantitative and qualitative character, that is manifest and latent content analysis, were sought. The analysis was performed by SK with BFdwith extensive experience in qualitative methods. In hierarchical complexity scoring, such content is “seen through” to examine its underlying structure. The method measures the levels of abstraction and how information is coordinated. Each section and subsection of a research ethics discussion was assessed on stage of hierarchical complexity. The overall discussion was scored based on the highest stage of performance the text demonstrated. The correlation of content and its complexity indicated which topics were addressed at different stages of complexity. Scoring was performed independently by SK and SR, then discussed to reach consensus. Both authors scored the English texts, and SK scored the ones in Scandinavian languages and discussed with SR. SR is an expert HCSS stageand-transition scorer while SK is a qualified HCSS scorer of stages 8 through 11. See table 1 for stage complexity information.22

RESULTS Research ethics issues in dissertations Dissertations contained one to seven research ethics topics: approval of research ethics board (94%); information process and informed consent (86%); confidentiality (67%); ethical aspects of methods (61%); use of ethical principles and regulations (39%), rationale for the study (20%) and fair participant selection method (14%; table 2). All but three of the dissertations involved direct interaction with study participants; three were registerbased studies.

Table 1 Common range of stages of performance in adult tasks’ hierarchical complexity Stage

Stage name

General descriptions of tasks performed

8

Concrete

9

Abstract

10

Formal

11

Systematic

12

Metasystematic

Uses simple deductions; time sequences are used to describe actual instances that occur in past or present time, and deal with specific things, specific incidents or events, specific actors, specific actions, specific places. Uses dualistic stereotypes (eg, good/bad, right/wrong); forms abstract variables out of concrete classes of objects; makes and quantifies propositions; uses variable time, place, act, actor, state, type; uses quantifiersd for example, all, none, some and categorical assertionsdfor example, “We all die”. Attempts to justify an assertion take the form of another assertion without specifying a logical connection; circular reasoning can appear. Uses empirical or logical evidence; logic is linear, onedimensional (ie, one input variable); solves problems with one unknown using algebra; forms logical relationships between abstract variables. Uses terms such as if.then, thus, therefore, and because in justifying assertions. Constructs multivariate systems and matrices, coordinating more than one variable as input; situates events and ideas in a larger context, that is, considers relationships in contexts; forms systems out of multiple formal relations. Integrates systems to construct metasystems, including principles, out of disparate systems; compare systems and perspectives in a systematic way (across multiple domains); reflect on systems, that is, is metalogical, metaanalytic; name properties of systems (eg, homomorphic, isomorphic, complete, consistent, commensurable).

Note: From ‘Hierarchical Complexity Scoring System (HCSS) How to Score Anything’ (pp. 62; 92e93), by M. L. Commons, P. M. Miller, E. A. Goodheart, and D. Danaher-Gilpin, 2005, Cambridge, MA: Dare Association, Inc. Copyright 1991e2005 by Dare Association, Inc. Adapted with permission.

parency concern was that several sections included no name of the ethics board and/or registration number (n¼13). A minority related the issue of ethics approval to ethical codes, the Helsinki declaration or current national research ethics laws (n¼14) by either stating that studies were performed in accordance with ethics regulations (n¼8) or by arguing against the need for an ethics approval due to national laws (n¼6).

Information and informed consent We broadened the traditional informed consent category to accommodate information-giving processes discussed but not always expressed in terms of informed consent. Most dissertations discussed information-giving and informed consent (n¼56). A third of these explicitly mentioned the concept of informed consent (n¼19). A substantial amount of space was typically used to detail the informing phase of research, including the information’s form (written and/or verbal) (n¼41) and type. The most often-given information was freedom to withdraw from the study (n¼33) and a declaration of voluntariness (n¼30). Other information included confidentiality (n¼22), withdrawals’ non-interference with further treatment (n¼7), the right to not answer questions (n¼4), aim of the study (n¼2), risks and benefits (n¼2) and feedback of results (n¼1). Those responsible for providing information as well as those receiving the information were described. Some informed consent discussions included an ethical rationale for the information process by referring to principles, codes or laws (n¼16).

Ethics approval The ethics approval category included descriptions of whether the dissertation has been vetted by an ethics review board. Almost all dissertations included a discussion of ethics approval (n¼60), and a majority stated they had been approved by a research ethics review board (n¼55). A quality and trans426

Confidentiality Items coded in the confidentiality category reported that information was accessible to only authorised persons. Confidentiality procedures were succinctly reported (n¼43). Besides describing confidentiality as something that participants were J Med Ethics 2010;36:425e430. doi:10.1136/jme.2009.034561


Research ethics Table 2

Design of research ethics sections in Swedish nurses’ dissertations

Number of topics Hierarchical complexity on research ethics Dissertations in Topics on research ethics of reasoning in the dissertation the category Ethical Informed Ethical aspects of Principles, Fair participation approval consent methods Confidentiality regulations Rationale selection Score 9 Score 10 Score 11 1 Sum

4 4

4 4

2

4 2 2 1 9

4 2 2 8

4

5 5 1 1 1 1 14

5 5 1

5

11

5 5 1 1 1 1 14

8 4 3 1 1 17

8 4 3 1 1 17

8 4 3 1 1 17

9 2 1 1 1 1 15

9 2 1 1 1 1 15

9 2 1 1 1 14

1 11

3 1 1 5

3 1 1 5

3 1 1 5

3 1

Sum

3 1 1 5

7 Sum

1 1

1 1

1 1

64 100

60 94

55 86

Sum 3

Sum 4

Sum 5

Sum 6

Total Per cent

4

2 2

2 2

0 0

1

2

3 2 1 1 8

0 0 0 0 0

0 0 0 0 0 0 0

5 5 0 1 1 0 12

0 0 1 0 0 1 2

0 0 0 0 0 0

7 4 2 1 1 15

1 0 1 0 0 2

0 0 0 0 0 0 0

8 2 0 1 1 1 13

1 0 1 0 0 0 2

3 1 1 5

0 0 0 0

2 1 3

2 1 3

1

5 1 1

1 1

1

6

7

2

8

8 4

4

3 1 12 9 1

1 1

1 1

3 1 1 12 9 2 1 1 1 1 15

5 9 2

1 4

2 1 1

1

1 1 1 3

12

1 5

4

3 1 1 5

1 1 2

1 4

0 0 0 0

1 1

1 1

1 1

1 1

1 1

0 0

1 0

0 0

39 61

43 67

25 39

13 20

9 14

4 6

54 84

6 10

guaranteed and informed about, some researchers identified how confidentiality had been handled: data were safely stored protecting participant’s identity (n¼12); data were analysed and reported without identifying participants (n¼19) and participants in focus group interviews were counselled in ways to promote freedom of expression and confidentiality (n¼2).

Ethical aspect of the methods The category for ethical aspect of the methods included the research ethics issues in collecting data, except for questions regarding informing participants. Ethical aspects of study methods were comprised of descriptions of interviews and questionnaires (n¼37). Explanations of why interviews were ethically problematic were done by referring to principles or risks of harm (n¼17). The negative aspects stated (n¼24) were physical and psychological with an emphasis on emotional. J Med Ethics 2010;36:425e430. doi:10.1136/jme.2009.034561

3

Strategies to impede negative consequences were depicted (n¼20): adopt a sensitive attitude, adapt to the physical and mental status of the interviewee, reduce questions, provide time to reflect on the interview and arrange for a contact person. Sometimes, statements about how the participants seemed to enjoy the interview experience were included (n¼14). A few sections described problems that appeared during the research interview (n¼14)dfor example, interviewees who cried or did not answer all questions. The most comprehensive sections covered all these issues, but the most common strategy was to mention the potential laboriousness of the interview yet argue that participants benefited from practical solutions that were provided in the interview situation or by claiming that research participants appreciated the opportunity to tell their stories. The reported ethical problems with questionnaires were primarily the tedium of answering questions and how researchers adjusted 427


Research ethics the number of requests for completion out of respect and concern for participants’ possible fatigue.

Use of ethical principles and regulations Discussions that included the usage of principles and ethical regulations like laws and research ethics codes were coded to the category of ethical principles and regulations. This category was analytically different from others because it revealed how ethics were applied in the research sections. Explicit report of laws, ethics codes and principles occurred in fewer than half of the dissertations (n¼25). Principles were employed but performed in qualitatively different ways (n¼17). The simplest form was to state that the study had been performed in accordance with a research ethics declaration, code or rules outlined in a research ethics book. The most elaborate ones integrated the principles and described how they were used as compasses for research procedures (n¼8).

Rationale for the study To provide an ethical rationale for the study means to justify why the study is important in a wider perspective. Thirteen dissertations featured an ethical rationale for the study, and when included, it was framed in terms of risks and benefits. The need for new and valuable knowledge that could potentially improve conditions for other people weighed heavier than the extra demand and little direct gain that the research subjects gained from participating. Some reported that the value of pursuing the research outweighed the disadvantages but entailed the necessity of protecting the autonomy of the research participants.

Fair participant selection Fair selection of participants signifies reflections on a justified choice of participants. The reason to include vulnerable groups and groups that previously has been excluded from research was sometimes given (n¼9). A few sections justified the choice of participants (n¼8). The importance of including important and vulnerable groups so their voices would be heard was the main reason reported.

Design of the research ethics section The topics of the research ethics sections are outlined in table 2. Most frequent was to report four ethics issues (n¼16), followed by three (n¼14) or five issues (n¼ 14). The majority (72%) included 3e5 issues. Four sections stated one topic and only one dissertation section reported seven issues. The most common composition of a section about research ethics discussed five topics: the approval from a research review board, information and informed consent, ethical aspects of the methods, confidentiality and principles.

Complexity of reasoning The analysed texts demonstrated three stages of performance as measured by hierarchical complexity: abstract (n¼4), formal (n¼54) and systematic (n¼6). Abstract stage text performances consisted of declarative statements (table 3). Unsupported categorical assertions were made and justified by invoking another assertion. Generalisations were created by quantifying people and events. Often-used quantifications in the sample were “all participants” and “all studies”. Research ethics sections included mainly generalisations about actions that had been performed. Reasoning at the formal stage of performance used empirical or logical evidence (table 3). Assertions were supported by explicit 428

logic or evidence to justify the assertiondfor example, by providing a logical explanationdfor example, using such terms as because, in order to, since, if, then, therefore. Descriptions of hypothetical or alternative options in the future were sometimes included. The logic was linear. Such linear logic took the form of ifethen constructions or chains of logic. Some used principles as logical reasons for actions. Systematic stage performances were characterised by the ability to coordinate at least two logical relations into a system (table 3); in other words, they demonstrated reasoning about complex causation and ability to understand a system of logical relationships. For example, one researcher described procedures for finding the “right people” by invoking a multivariate system that required the coordination of multiple variables. Systemic stage performances were characterised by more fluid reasoning than the linear, logical performances.

Comparing content and complexity Few dissertations demonstrated abstract reasoning and systematic reasoning, four and six, respectively, but showed interesting patterns. The texts with abstract stage reasoning reported either one or two topics. All four mentioned approval; information and methodological issues were raised by only two. Texts with systematic reasoning introduced three to five ethical issues. Half of them discussed principles (as compared to merely citing a principle as the reason for an action), and the other three reported the rationale for the study, indicating that the topic and study could perhaps be viewed in a wider context. Among the majority of texts demonstrating formal reasoning, the topics varied from one to seven, meaning at least formal reasoning was needed to explain all conceivable aspects. Formal reasoning is required to report such tasks as fair selection of participants, rationale for the study and principles, ethics codes and laws.

DISCUSSION Our study demonstrates that research ethics are insufficiently reported and inadequately described in many nursing dissertations. Few ethical topics are considered, and they are not discussed in a thorough way. While most note official approval and describe informed consent issues, other issues like the rationale for the study and how the participants were selected are infrequently reported. The level of complexity of reasoning was inadequate in most dissertations. The majority of the dissertations used formal reasoning, although by their nature, the ethical issues introduced in them require more complex reasoning to be satisfactorily addressed. A methodological strength of our study is its inclusion of a large number of dissertations, which are likely representative of dissertations by Swedish nurses. A major advantage of our method is that the analytical approach permits assessments and comparisons of the coverage of ethical issues and the complexity of reasoning. A methodological shortcoming is that the analysis was primarily focused on the section denoted “Ethical considerations/approval”, thus some ethics topics and reasoning might have passed undetected if they were treated in other parts of the dissertation. The analysis is thus limited to what the authors define as belonging to ethics sections. Our analysis identified the most complex stage of reasoning as a criterion for analysis because ethical considerations are complex matters. A more extensive analysis could have also analysed the entire low to high range of reasoning demonstrated in each ethics section. An implication of the language analyses is that we do not know which and how the ethical issues were applied in reality. Some J Med Ethics 2010;36:425e430. doi:10.1136/jme.2009.034561


Research ethics Table 3

Representative examples of reasoning in research ethics at three stages of complexity

Topic

Abstract reasoning

Formal reasoning

Systematic reasoning

Who informed

As a second step, they were contacted by a nurse over the phone and asked whether or not they were willing to participate. (Dissertation 6)

A coordination nurse at first approached the parents in order to let them decide about participation, and to ask whether the researcher could take contact with them. (Dissertation 4)

Information

All participant were given written (I, II, IV), or written and oral information (III) about the study and that participation was voluntary. (Dissertation 58)

Risk with interviews

No example.

Supported by the principle of beneficence, information about the research was therefore given verbally and in writing, including information about voluntary participation. (Dissertation 53) Since the patients were interviewed about their experiences that could have strong emotional meaning and could be distressful to talk about, they were offered an opportunity to contact the interviewer after the interviews to discuss any subsequent reflections. (Dissertation 18)

By getting help from a gate-opener in finding suitable participants in the research studies the risk of exposing sensitive and too vulnerable individuals was eliminated, which upholds the principle of non-malfeasance (Beauchamp and Childress, 1994). (Dissertation 53) No example.

issues could have been omitted from the dissertation text even though the issue was dealt with in practice and vice versa. The consistency between writing about ethics and ethical behaviour in the fielddfor example, in contact with research subjects and patients, should be investigated in future studies. The first main finding is the incompleteness of the elaboration of topics and details in several dissertations, which is consistent with several studies in the domain of research ethics. A previous study showed a high level of errors in research ethics committee letters; that is, procedural violations, missing information, slipups and discrepancies.8 Earlier research on Swedish nurses’ dissertations demonstrate the questionable quality due to short length, few references and a narrow range of topics.12 In our study, few topics were addressed. Emanuel et al argued for seven requirements to be considered and met in the conduct of ethical research: scientific value, validity, fair subject selection, favourable riskebenefit ratio, independent review, informed consent and respect for potential and enrolled subjects.23 Applied to our findings, some requirements may be treated in other parts of a dissertation, but several dissertations leave out topics that are necessary for judging their ethical quality. Informing potential participants and pursuing informed consent was reported in almost 90% of the dissertations’ ethics sections. This frequency is higher than that reported in a study of Turkish nurses’ dissertations where subjects were not informed about the study (72.7%) and the researchers had not obtained permission from the subjects (73.6%).11 The second main finding is the insufficient level of complexity of reasoning, with which research ethics are handled. Findings from a discourse analysis of research ethics committee letters showed that there was “the lack of formal reasoning” (p 258) and ethical argumentsdfor example, informed consent are described as procedural norms rather than an ethics principle possible to dispute.9 This is consistent with our findings, because a significant number invoked research ethics principles to justify procedures taken, rather than to use principles to support ethical arguments for and against certain procedures. However, our findings also showed that the great majority used at least some formal reasoning, as measured by hierarchical complexity. Unfortunately, formal reasoning is necessary but not sufficient for adequacy in ethical matters. The analysis showed that formal reasoning and systematic reasoning were needed to elaborate on topics, and the comparison of complexity reasoning and content indicated that higher levels of reasoning involved J Med Ethics 2010;36:425e430. doi:10.1136/jme.2009.034561

Despite the risk that interview questions might cause feelings of uneasiness, there was deemed to be a greater potential for beneficence that outweighed the risks by providing participants with the opportunity to tell their stories in an open and encouraging way. (Dissertation 13)

more elaborated use of ethics principles. Very few used systematic reasoning, and none used metasystematic, which would be preferable because several of the research ethics concepts are metasystematic stage principles. For example, informed consent is a metasystematic stage concept because it coordinates the system of informing a research subject and the system of obtaining consent from the person.24 This means that metasystematic reasoning is needed for a full understanding and use of these concepts. What are possible explanations for the low levels of reasoning on research ethics? One possibility is that ethical issues are dealt with at a sufficiently high level of complexity in practice, whereas the text of the dissertation merely reflects a research tradition that discounts the importance of performing and explaining ethical reasoning. Disciplinary norms for terse writing styles are presumably promoted by supervisors and department guidelines. For example, nurses’ dissertations in social science use more references to methods, ethics and philosophy of science than dissertation in the medical science tradition.23 In addition, poor writing may occur because researchers mimic previous dissertations or regard ethical considerations as bureaucratic hurdles rather than moral requirements to protect participants. The supervisor role is an important factor since they sometimes acknowledge a considerable lack of knowledge about research ethics.25 Another conceivable explanation is that the level of ethical reasoning corresponds rather accurately to the level of complexity the doctoral students and their supervisors use to handle complex issues in general. In other words, they are arguing on ethical issues at their highest complexity level. In that case, the scientists’ (PhD students’ and supervisors’) ability to discuss at more complex levels must be improved for ethical issues to be sufficiently managed in the future. All these possibilities suggest further research is needed to account for our findings, since ethics have long been an important part of nurses’ education and occupation. There are several implications of insufficient ethical reasoning. Integrity of the research subjects and patients are at risk, and patients, if they participate, may be informed without understanding the implications. From the perspective of the readers of the scientific literature, it is impossible to assess how and why the authors dealt with various ethical issues. A crucial implication is the consequences of selection of research questions, methods and participants/sample. Scientists performing at abstract or formal stages are less likely to integrate relevant ethical aspects into their research aims than scientists at higher 429


Research ethics complexity levels. This is because such integration, by its nature, is multivariate at minimum. They will differ quite dramatically in the way they understand principles as principles, “risks” and “benefits”, rationale of the investigation, etc. Researchers with systemic or metasystematic stage reasoning are able to ask more complex questions, juggle ethics, research questions, and methods and design more complex research projects.26 Our conclusion is that if the established praxis to include discussion of research ethics in Swedish nurses’ dissertations is going to be valuable, and if its purpose is to indicate that the research complied with expected ethics, then the reporting must exhibit a certain quality, comprehensiveness and sufficiently significant treatment of ethics. Our study illustrates that factors that improve the quality include: appropriately thorough consideration of several ethical issues while avoiding minutiae; use of ethical principles in appropriate contexts to justify choices and reasons to support actions taken and use of at least formal and systematic reasoning. In addition, we would like to see more reflection and a critical stance to what has been done in the dissertation work. In order to accomplish the intent of reporting research ethics, several improvements are needed. The most straightforward solution is to enhance the research ethics teaching in graduate education. Students must learn how to perform ethically sound research from the first steps of planning and performing to writing up the results and their potential and ability to report and reflect on ethical aspects of the research process must be enhanced. A more profound resolution is to emphasise metasystematic thinking in post-graduate studies and recruit senior researcher and post-graduate students who already have developed a systematic or metasystematic way of reasoning. This longer-term solution will also constitute the foundation for further development of complexity in handling ethics issues in the future.

3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

16. 17. 18. 19. 20.

Acknowledgements We would like to thank professor Per Sjo¨lander for valuable comments on the discussion. Competing interests None. Provenance and peer review Not commissioned; externally peer reviewed.

REFERENCES 1. 2.

430

Emanuel EJ, Wendler D, Grady C. What makes clinical research ethical. JAMA 2000;283:2701e10. Dixon-Woods M, Ashcroft RE. Regulation and the social license for medical research. Med Health Care Philos 2008;11:381e91.

21. 22. 23. 24. 25. 26.

Fischer BA. A summary of important documents in the field of research ethics. Schizophr Bull 2006;32:69e80. Finlay KA, Fernandez CV. Failure to report and provide commentary on research ethics board approval and informed consent in medical journals. J Med Ethics 2008;34:761e4. Dixon-Woods M, Angell EL. Research involving adults who lack capacity: how have research ethics committees interpreted the requirements? J Med Ethics 2009;35:377e81. Angell EL, Jackson CJ, Ashcroft RE, et al. Is ‘inconsistency’ in research ethics committee decision-making really a problem? An empirical investigation and reflection. Clin Ethics 2007;2:92e9. Angell EL, Bryman A, Ashcroft RE, et al. An analysis of decision letters by research ethics committees: the ethics/scientific quality boundary examined. Qual Saf Health Care 2008;17:131e6. Angell E, Dixon-Woods M. Do research ethics committees identify process errors in applications for ethical approval? J Med Ethics 2009;35:130e2. O’Reilly M, Dixon-Woods M, Angell E, et al. Doing accountability: a discourse analysis of research ethics committee letters. Sociol Health Illn 2009; 31:246e61. Rowan-Legg A, Weijer C, Fernandez CV. A comparison of journal instructions regarding institutional review board approval and conflict-of-interest disclosure between 1995e2005 J Med Ethics 2009;35:74e8. Ulusoy MF, Ucar H. An ethical insight into nursing research in Turkey. Nurs Ethics 2000;7:285e95. Kjellstro¨m S, Fridlund B. Status and trends of research ethics in Swedish nurses’ dissertations. J Nurs Ethics 2010;17:382e3. Creswell JW. Research design: qualitative, quantitative, and mixed methods approaches. Thousand Oaks (CA): Sage, 2009. Commons ML, Pekker A. Presenting the formal theory of hierarchical complexity. World Futures 2008;64:375e82. Commons ML, Goodheart EA, Pekker A, et al. Using Rasch scaled stage scores to validate orders of hierarchical complexity of balance beam task sequences. In: Smith JEV, Smith RM, eds. Rasch measurement: advanced and specialized applications. Maple Grove (MN): JAM Press, 2007;121e47. Dawson TL. A comparision of three developmental stage scoring systems. J Appl Meas 2002;3:146e89. Dawson TL. A stage is a stage is a stage: a direct somparison of two scoring systems. J Genet Psychol 2003;164:335e64. Dawson TL, Xie Y, Wilson M. Domain-general and domain-specific developmental assessments: do they measure the same thing? Cogn Dev 2003;18:61e78. Dawson TL. The meaning and measurement of conceptual development in adulthood. In: Hoare CH, ed. Handbook of adult development and learning. Oxford: Oxford University Press, 2006:433e54. Dawson TL. Assessing intellectual development: three approaches, one sequence. J Adult Dev 2004;11:71e85. Swedish law. Act on Ethical Review of Research Involving Humans, 460. 2003. Commons ML. Introduction to the model of hierarchical complexity and its relationship to postformal action. World Futures, The Journal of General Evolution 2008;64:305e20. Heyman I. Ga˚nge hatt till: omva˚rdnadsforskningens framva¨xt i Sverigedsjuksko¨terskors avhandlingar (The emergence of nursing research in Swedendnurses’ dissertations). Go¨teborg: Daidalos; 1995. Commons ML, Rodriguez JA, Adams KM, et al. Informed consent: do you know it when you see it? Psychiatr Ann 2006;36:430e5. Szirony TA, Price JH, Wolfe E, et al. Perceptions of nursing faculty regarding ethical issues in nursing research. J Nurs Educ 2004;43:270e9. Commons ML, Richards FA. Four postformal stages. In: Demick J, ed. Handbook of adult development. New York: Plenum, 2002:199e219.

J Med Ethics 2010;36:425e430. doi:10.1136/jme.2009.034561