Study settings and sample. The study was
approved by the: Research Ethics Committee of the
Faculty of Nursing, Chiang Mai University; Hospital
Directors of the two hospitals, used in Phase I; and,
Research Ethics Committees of each of the ten
hospitals, used in Phase II. Hospitals used in Phase
II were selected to obtain a heterogeneous sample of
nurses.20 All hospitals used as data gathering sites
were located in Northern Thailand. None of the
hospitals used in Phase I were used during Phase II.
The sample consisted of two types of subjects
(patients and professional nurses). Subjects were
informed: about the purpose of the study; what their
particular involvement would entail; their participation
was voluntary; they could withdraw from the study
at any time without negative repercussions; and, their
anonymity and confidentiality would be maintained.
In the event they had questions, the names and
addresses of the investigators were provided. Those
consenting to participate were asked to sign a consent
form.
Phase I included 37 patients from medical,
surgical, orthopedic, obstetrical and gynecological
units at one regional hospital, who participated in
defining the meaning and dimensions of caring
behavior. Selection criteria included being a patient
who was: over 18 years of age; fully conscious; able
to provide experiential information; and hospitalized
on a medical, surgical, orthopedic, obstetrical or
gynecological clinical unit for at least one day.
Head nurses of the clinical units provided names of
potential patient subjects, who then were purposively
selected by the primary researcher. In addition, during
Phase I, 60 nurses, from one regional hospital and
one community hospital, participated in testing the
instrument’s internal consistency and stability. Ten
of these nurses examined the face validity of the
instrument. Criteria for selection included being
employed in one of the two hospitals used as a data
gathering site, and having, at least, one year of
nursing experience.
Names of potential nurse subjects also were
provided by head nurses of each clinical unit. The
number of nurses selected from the two hospitals
was proportionate to the number of nurses working in
each hospital. The regional hospital had 800 nurses
employed, while the community hospital had 60.
Thus, 50 nurses were obtained from the regional
hospital, while 10 were obtained from the community
hospital. During this phase, seven nurses failed to
successfully complete the second administration of
the instrument. Thus, data from only 53 nurses were
usable. Of the seven nurses who were unavailable
for the second administration of the instrument, 5
were from the regional hospital and 2 were from the
community hospital. Both patients and nurses, in
Phase I, who met the selection criteria, were invited
by the primary investigator to participate in the study.
For Phase II, 810 nurses from 10 institutions
(one university, two regional, three general and four
community hospitals) were asked to participate in
examining psychometric properties of the instrument.
A total of 758 (96.06%) provided complete and
usable responses. The number of professional nurses
sampled for Phase II was proportional to the number
of professional nurses’ employed within each hospital21
(see Table 1). In addition, the number of the
participants involved in Phase II was determined via
the minimum observation-to-variable ratio, 10:1,22
with 30% added for sample attrition. Potential nurse
subjects meeting the selection criteria were contacted
by one of 10 research assistants who were trained by
the primary investigator regarding the data collection
procedure. The recruiting process ended when the
number of subjects from each selected hospital met
the requirement. Hence, the sampling procedure
applied both convenience and quota methods.