On-the-Job Ethics – Proximity Morality Forming in Medical School: A grounded theory analysis using survey data

Hans O. Thulesius, MD, Ph.D.

Abstract

On-the-job-ethics exist in all businesses and can also be called
proximity morality forming. In this paper we propose that
medical students take a proximity morality stance towards ethics
education at medical school. This means that they want to form
physician morality “on the job” instead of being taught ethics like
any other subject. On-the-job-ethics for medical students involves
learning ethics that is used when practicing ethics. Learning
ethics includes comprehensive ethics courses in which quality
lectures provide ethics grammar useful for the ethics practicing
in attitude exercises and vignette reflections in tutored group
discussions. On-the-job-ethics develops professional identity,
handles diversity of religious and existential worldviews, trains
students described as ethically naive, processes difficult clinical
experiences, and desists negative role modeling from physicians
in clinical or teaching situations. This grounded theory analysis
was made from a questionnaire survey on attitudes to ethics
education with 409 Swedish medical students participating. We
analyzed over 8000 words of open-ended responses and multiplechoice
questions using classic grounded theory procedures, but
also compared questionnaire data using statistics such as
multiple regression models. The paper gives an example of how
grounded theory can be used with a limited amount of survey
data.

Background

Medical ethics is different from other subjects taught at
medical school and the importance of formal ethics courses has
been questioned (Hafferty & Franks, 1994). Some medical schools
combine instruction in bioethical principles with teaching of
humanities programs (Andre, Brody, Fleck, Thomason &
Tomlinson, 2003). The teaching of ethics varies in Swedish
medical schools from interspersed lectures to formal ethics
courses. We designed a questionnaire survey in order to elucidate
how Swedish medical students view the ethics education in
medical schools (Thulesius, Sallin, Lynöe & Löfmark, 2007;
Lynöe, Löfmark & Thulesius, 2008). Many students gave input to
the ethics course curriculum: Should ethics be taught in lectures
or learned through group discussions? Should the ethics course be
a separate course among others, or should it be part of other
courses with lectures and group discussions interspersed? Should
it come early or late in the medical school curriculum? Should the
literature be specific ethics literature or novels and short stories
with relevant ethical content? From multiple-choice responses we
found that strong ethics interest was associated with frequent
experiences of physician teachers as good role models and an
absence of poor role models (Lynöe et al., 2008). In the present
study we wanted to explore what was going on in medical schools
regarding the medical ethics education by analyzing open-ended
survey responses together with response data from multiplechoice
items.

Method

We constructed a survey on attitudes towards the medical
ethics education during 2005 as a request from the delegation of
medical ethics of the Swedish Society of Medicine. Swedish
medical students from the 1st, 5th and 11th (last) term
participated. The survey consisted of 14 items, of which 10 had a
total of 59 multiple-choice response options and generous space
for open-ended comments, and 4 items were open-ended only, see
Table 1.

The overall response rate to the questionnaire survey was
36%, and varied between different centers from 13% to 83%, with
a total of 409 respondents, 308 women (75%) and 101 men (25%).
More than half (220/409) of the respondents gave one or more
written open comments amounting to >8000 words. These
comments were transcribed into Word from handwritten text.
“Walking survey” data from informal interviews with four
physicians, of which two has been teaching medical ethics at
medical school for many years, were also analyzed (Glaser, 1998,
p 214).

At some centers a whole term would drop out since the
responsible teacher failed to hand out the survey. Yet, the
response patterns of the different questionnaire items did not
differ significantly between schools with low and high response
rates when different logistic regression models were applied to
the data (Lynöe et al., 2008). The most comprehensive open
responses came from last term students. Thus the most
experienced students gave the biggest input to the analysis of the
qualitative data – the main data source for this study.

We analyzed open-ended comments and multiple-choice
responses by classic grounded theory (GT) procedures according
to Glaser (1978; 1992; 1998; 2001; 2003; 2005; 2007). The GT
dictum “all is data” was taken ad notam in this study. We thus
compared both qualitative responses and quantified multiplechoice
items in the same analysis. Multiple-choice results were
dichotomized, analyzed in logistic regression models, and
compared with open-ended responses. The GT analysis began
with open coding trying to answer the questions “what is going
on?” and “what concept does this data represent“ or “what concept
that explains what is going on catches the latent pattern in this
data?” and most important: “what are the participants main
concern and how are they continually trying to resolve it?”
Theoretical memos were written, typed, or drawn in the
comparative process as soon as open coding started. This paper
was sorted and written up from more than 4000 words and many
dozens of pages of typed and handwritten memos.

“Memos are the theorizing write-up of ideas about
substantive codes and their theoretically coded relationships as
they emerge during coding, collecting and analyzing data, and
during memoing” (Glaser, 1998). Memoing is “the core stage of
grounded theory methodology” (Glaser, 1998), and should be done
at any time and place in order to capture creative ideas. The
analytic procedures were done with experience from earlier GT
studies (Thulesius, Håkansson & Petersson, 2001, 2004;
Sandgren, Thulesius, Fridlund & Petersson, 2006; Thulesius &
Grahn, 2007).

Discovery of Grounded Theory by Glaser & Strauss (1967) is
the most quoted reference for any single method analyzing
qualitative data according to Google Scholar search (12830
citations December 2008). GT has the inductive approach to
generate hypotheses explaining how participants in a studied
substantive area resolve their main concern. Thus, GT
conceptualizes “what is going on” in the field of study by the
“constant comparative method”, another name for GT. This
indicates a constant comparison of data during an iterative
research process, which involves open coding, memoing,
theoretical sampling (data collection based on hypotheses from
the ongoing analysis), selective coding (recoding data based on
concepts from the ongoing analysis), sorting and writing up
(sorting memos in piles based on concepts in the theory and then
writing up the sorted piles into a paper or book). Classic GT
analysis aims at conceptual theories abstract of time, place and
people and differs from most methods using qualitative data by
presenting explanatory concepts instead of descriptions. Many
clinical research methods consider persons or patients as units of
analysis, whereas in GT the unit of analysis is the incident not
the person(s) involved (incident = a distinct piece of action, or an
episode, as in a story or play). The number of incidents being
coded and compared typically amounts to several hundred in a
GT study since every participant often reports many incidents.
When comparing many incidents in a certain field, the emerging
concepts and the relationship between them are in reality
probability statements and therefore GT should not be considered
a qualitative method but a general method that can use any type
of data. The results of GT are not reports of facts but an
integrated set of conceptual hypotheses. Validity in its traditional
sense is consequently not an issue in GT research, which instead
should be judged by fit, relevance, workability, and modifiability
(Glaser, 1978; 1998). Fit has to do with how close concepts fit
with the incidents they are representing, and this is related to
how thorough the constant comparison of incidents to concepts
was done. A relevant study deals with the real concern of
participants and captures attention. The theory works when it
explains how the problem is being solved with much variation. A
modifiable theory can be altered when new relevant data is
compared to existing data. A GT is never right or wrong, it just
has more or less fit, relevance, workability and modifiability, and
readers of this paper are asked to try its quality according to
these principles.

Proximity Morality Forming by On-the-job Ethics

In this study we analyzed student attitudes and “what was
going on” in the medical ethics education and found that students
learn ethics on the job. This can also be conceptualized as
proximity morality forming since students practice medical
ethics in close connection with colleagues and patients. This
proximity morality forming also includes comprehensive ethics
courses with tutored small groups. Proximity morality forming
involves learning ethics where “ethics grammar” comes from
selected high quality lectures. Practicing ethicsis done when
patient cases and clinical issues are discussed in interactive
groups and in the clinical setting. This can also help students to
deal with emotionally difficult situations. Attitude exercises using
vignette reflections are done in “ethics labs”. To desist negative
role modeling is a function of the ethics courses where reflected
professionalism is developed for diverse medical students in a
heterogeneous world.

On-the-job-ethics in medical school – How? Forming
physician morality by learning ethics takes place in quality
lectures on ethics, preferably given by professional ethicists.
These lectures provide students with a basic “ethics grammar”
about ethical principles and concepts. This feeds the interactive
group discussions and improves their quality concerning ethical
issues.

“Professional lecturers from the faculty of arts (are wanted)”
(first term student). Forming physician morality by practicing
ethics is done in the interactive discussion groups, but also in the
“ethics lab” where students work with practical, sometimes
challenging attitude exercises and vignette reflections. These
stimulate critical thinking about current ethical problems in
clinical training. It requires that the participants position
themselves ideologically, and for some attitude exercises also
physically. Attitude exercises are often done in case studies.

“A case is presented and different opinions (re the case)
represented by four different corners. One can go to any corner
and argue against the other corners and eventually change
corners” (last term student).

On the job morality forming in medical school is typically
done in interactive discussion groups. In these groups the
learning and practicing go hand in hand. The discussion groups
also have a support network function allowing professional role
growth within a permissive context where ethical and valueladen
issues are discussed and tried. The structure ideally
consists of tutored groups that repeatedly work with case study
approaches, discuss ethical principles, and continue during
internship (i.e. in Sweden this is the paid physician work that
starts after medical school at the University). Within a frame
resembling the clinical setting students grow their own ethical
attitudes and shape their individual physician morality. Group
discussions provide good training for handling ethical difficulties
since real world medical ethics consist of unique complex
situations often involving several people. One goal of interactive
ethics group discussions is to understand what appropriate
physician behavior is.

“(we need) group discussions with teachers making sure that
everyone develops decent ethical values as physicians” (fifth term
student)

“ethics discussion forums should be based on tutored small
groups (to prevent people with strong views from dominating)”
(last term student)

“every section could end with ethical discussions related to
the specific subject, psychiatry/internal medicine/surgery” (last
term student)

On-the-job-ethics in medical school – Why?

Why would medical students want to form physician
morality on the job? The deliberate forming of a physician
morality seems necessary for various reasons, and several
student responses dealt with arguments for ethics education in
general and forming physician morality on the job in particular:

Professionalizing

Since professional identity requires moral reflection this is
an important argument for on-the-job-ethics.

“An open discussion forum on difficult issues and
professional identity conflicts would make us better physicians.”
(last term student)

“Small groups during clinical training – discussing the
professional physician role and work issues” (last term student on
suggested ethics education during internship)

Diversity handling

We live in a society with increasing diversity and multiple
religious views and this is dealt with in everyday on-the-job
ethics.

“What is it really like in our secularized country? How can
we say something is right when we don’t share the same values”
(fifth term student)

Medical students are different. Some are ethically naive, or
not interested in ethics, and others even described as socially
“autistic”. The importance of ethics education is obvious for these
groups.

“Only autistic people need ethics education” (last term
student)

Processing emotional difficulties

Medical school can be both emotionally and ethically
difficult with life and death issues pressing on. On-the-job-ethics
discussions involve processing tough experiences from the clinical
part of the education.

“Good with special ethics courses when we deal with
sensitive issues” (last term student)

”We underestimate the power of what we can do for each
other… An open forum for discussing difficult issues and identity
crises during the education would make us better physicians”
(last term student on importance of small discussion groups)

“Small groups discussing everyday problems and ethical
issues in the workplace” (first term student on suggested ethics
education during later internship)

Desisting negative role modeling

By defying ethics suppression and politically corrected ethics
the influences of physicians/teachers as poor role models may be
addressed and negative role modeling dealt with in the
interactive groups. Some teachers and physicians were described
as being “masters of opinion control” trying to neutralize
discussions about ethically sensitive topics by putting the lid on
discussions, and defending politically correct opinions.

“I prefer a good (neutral) clinician instead of zealous,
ideologically motivated people” (fifth term student)

“Teachers gave too little space for own views – there was a
correct key for the discussion” (last term student)

In a statistical analysis of the survey presented elsewhere
(Lynöe et al., 2008) we saw a significant relationship between a
low interest in ethics and frequent experiences of poor role
models and the absence of good ones in all three terms. For last
term students, there was a significant association between a high
interest in ethics and experiences of good role models and a
preference for discussions in small groups.

“Personally I’m always ready to learn, although I do not
always like being taught.” ~ Winston Churchill (1874 – 1965)

The quote illustrates the students’ attitudes towards medical
ethics education in this study. They want to form their own
physician morality on-the-job rather than being taught ethics.
This is an example of the proximity ethics that influences the
health professions today where “personal relationships and
partiality override impartialist and universalist ethical
considerations” (Nortvedt & Nordhaug, 2008). The present
analysis suggests proximity morality forming as a name for
what is going on when medical students learn ethics while
becoming physicians. Another informal name for this concept is
“on-the-job” ethics. This ideally takes place in comprehensive
ethics courses where tutored groups openly discuss and reflect on
difficult ethical topics and moral dilemmas. Learning ethics is
done through high quality lectures supplying an ethics grammar
that provide default ethical principles. These are used when
practicing ethicsin group discussions together with attitude
exercises and vignette reflections in ethics labs. These interactive
discussion groups also have a support network function. Here
students process ethical problems in an environment where
physician morality is allowed to form and grow on the job. Hence,
rather than being served ideologically stained opinions students
prefer to reflect and discuss different ethical attitudes.

To summarize its consequences proximity morality forming,
or “on-the-job” ethics develops professionalism, deals with
diversity issues, helps in processing difficulties, and desists
negative role modelling in clinical teaching.

The students hoped for more interaction between students
and teachers in a British study of university students’
expectations of teaching (Sander, Stevenson, King & Coates,
2000). They also suggested that groups provide effective learning,
and this view was most prominent among medical students.
Those findings resemble the present study when it comes to
preferences for teaching structures. In a Swedish study the
authors suggested that interactive lecturing was a stimulant to a
problem-based learning (PBL) program (Fyrenius, Bergdahl &
Silen, 2005). This is in line with the need for good quality lectures
to feed ethical discussions with ethics grammar and input from
ethics labs in the present study. In a review of medical ethics
teaching (Hafferty & Franks, 1994) the authors were nihilistic
about its effects and suggested that critical determinants of
physician identity operate not within the formal curriculum but
in a subtler, less officially recognized “hidden curriculum”. Also,
medical education could be seen as a form of moral training of
which formal instruction in ethics constitutes only a small piece.
In a study investigating the effect of ethics education on
physician morality it was concluded that moral development and
ethical confidence were unaffected by ethics education (Gross,
1999). The goals of ethics education was conceptualized as having
cognitive, behavior and attitudinal dimensions. Ethics was
supposedly studied for its own sake contributing to “one’s all
around character”. We agree with this author’s conclusions, and
our analysis suggests that instead of an emphasis on teaching,
ethics and morality has to be learned on the job as discovered in a
neonatal unit study of proximity ethics (Brinchmann & Nortvedt,
2001) As a reference to one’s own morality, Levinas (1969) talks
about “the other”. Similarly, “the others” (fellow students and
teachers/physicians) are necessary for understanding the
suggested “on the job” morality development in our study.

Most data used for the GT analysis in this study are limited
to written open comments to survey items and multiple-choice
survey responses. We did not theoretically sample data outside of
the survey apart from data from our own experience, both as
medical students, physicians and teachers (all four authors of the
paper by Thulesius et al. (2007) are physicians and two authors
have experience of teaching medical ethics at medical school).
Thus the constant comparison was done mainly with cross
sectional written data, though “walking survey” data were also
used (Glaser, 1998, p 214). Yet we conceptualized a tentative
explanatory model of how 220 medical students want their
education in medical ethics. This suggests relevance enough for
generating a preliminary core variable GT. This theory is,
according to the GT paradigm, not right or wrong. It is just a set
of probability statements from which hypotheses are generated
by constantly comparing available data. When presenting this
proximity morality forming model of on-the-job-ethics to
physician colleagues and ethics teachers (both in Sweden and in
the USA) the reactions have been positive with some exceptions.
The model makes sense and seems to fit with experience. This
indicates a certain workability, at least for Swedish and North
American contexts.

Limitations

This paper proposes a model showing how medical students
want their ethics education in medical school, but does not take
into account their teachers’ views. Also, our study is limited by
the qualitative data being mostly written comments (O’Cathain &
Thomas, 2004) in an otherwise multiple-choice survey with a
partial response rate. As for the low response rates, the centers
with the highest response rates (83%) had the same attitude
pattern as those with low response rates (13%) (Lynöe et al.,
2008). Thus the data seems generalisable enough to fit the
requirements for an inductive study. The 11th term students gave
the largest quantitative input of qualitative data and thus had a
comparatively larger impact on theory generation. Whether this
was a limitation is questionable. In our view it gave us more
valuable longitudinal data. To use interview data by theoretically
sampling outside of the survey might improve the model. We
tried to compensate for this by also sampling dichotomized
multiple-choice survey data analyzed by different statistical
methods including multiple regression models (Lynöe et al.,
2008). Thus, we also used quantitative data according to the GT
maxim “all is data”. Furthermore, we used as data four
physicians’ experience as “walking surveys”. For possible future
application in medical schools we intend to refine and modify the
model and develop it through interaction with medical students
and teachers.

Acknowledgements

I wish to thank the Delegation for Medical Ethics of the
Swedish Society of Medicine and the participating medical
students. Dr Barney G. Glaser gave valuable support and input
for naming the core variable. This article is a modified version of
a study published in BMC Medical Education
http://www.biomedcentral.com/1472-6920/7/27 (Thulesius et.al.,
2007), and my co-authors of that paper are of course to be
acknowledged for their input also as “walking surveys”.

Author

Hans O. Thulesius1,2
1Department of Clinical Sciences Malmö,
Division of Family Medicine
Lund University, Sweden
2Research and Development Centre
Kronoberg County Council
Växjö, Sweden
Email: hans.thulesius@ltkronoberg.se

References

Andre, J., Brody, H., Fleck, L., Thomason, C.L., & Tomlinson, T.
(2003). Ethics, professionalism, and humanities at
Michigan State University College of Human Medicine.
Academic Medicine, 78, 968-72.

Brinchmann, B.S., & Nortvedt, P. (2001). Ethical decision making
in neonatal units–the normative significance of vitality.
Medicine, Health Care & Philosophy, 4, 193-200.

Fyrenius, A., Bergdahl, B., & Silen, C. (2005). Lectures in
problem-based learning–why, when and how? An
example of interactive lecturing that stimulates
meaningful learning. Medical Teacher, 27, 61-5.

Glaser, B.G. (1978). Theoretical sensitivity: Advances in the
methodology of grounded theory. Mill Valley, CA:
Sociology Press.

Glaser, B.G. (1998). Doing grounded theory: Issues and
discussions. Mill Valley, CA: Sociology Press.

Glaser, B.G. (2001). The grounded theory perspective I:
Conceptualization contrasted with description. Mill
Valley, CA: Sociology Press.

Glaser, B.G. (2003). The grounded theory perspective II:
Description’s remodeling of grounded theory. Mill Valley,
CA: Sociology Press.

Glaser, B.G. (2005). The grounded theory perspective III:
Theoretical coding. Mill Valley, CA: Sociology Press.

Glaser, B.G., & Strauss, A.L. (1967). Discovery of grounded
theory: Strategies for qualitative research. Chicago, IL:
Aldine.

Gross, M.L.(1999). Ethics Education and Physician Morality.
Social Science & Medicine 49:329-42.

Hafferty, F.W. & Franks, R. (1994). The hidden curriculum,
ethics teaching, and the structure of medical education.
Academic Medicine, 69, 861-71.

Levinas, E. (1969). Totality and infinity. Pittsburg, PA: Duquesne
University Press.

Lynöe, N., Löfmark, R.,& Thulesius, H. (2008). Teaching medical
ethics: What is the impact of role models? Some
experiences from Swedish medical schools. Journal of
Medical Ethics, 34, 315-6.

Nortvedt, P., & Nordhaug, M. (2008). The principle and problem
of proximity in ethics. Journal of Medical Ethics, 34, 156-
61.

O’Cathain, A., & Thomas, K.J. (2004). “Any other comments?”
Open questions on questionnaires – a bane or a bonus to
research? BMC Medical Research Methodology, 4, 25.

Sander, P., Stevenson, K., King, M., & Coates, D. (2000).
University students’ expectations of teaching. Studies in
Higher Education, 25, 309-323.

Sandgren, A., Thulesius, H., Fridlund, B., & Petersson, K. (2006).
Striving for emotional survival in palliative cancer
nursing. Qualitative Health Research, 16, 79-96.

Thulesius, H., & Grahn, B. (2007). Reincentivizing – a new theory
of work and work absence. BMC Health Services
Research, 7,:100

Thulesius, H., Håkansson, A., & Petersson, K. (2003). Balancing:
a basic process in end-of-life cancer care. Qualitative
Health Research, 10, 1353-77.

Thulesius, H., Sallin, K., Lynöe, N., & Löfmark, R. (2007).
Proximity morality in medical school–medical students
forming physician morality “on the job”: grounded theory
analysis of a student survey. BMC Medical Education, 7,
27.