bs_bs_banner

Nursing and Health Sciences (2014), 16, 298–306

Research Article

Psychometric properties of the Chinese-version Quality of

Nursing Work Life Scale
Ya-Wen Lee, RN, PhD,1,3 Yu-Tzu Dai, RN, PhD,1 Linda L. McCreary, RN, PhD,4 Grace Yao, PhD2 and
Beth A. Brooks, RN, PhD5
1
Department of Nursing, 2Department of Psychology, National Taiwan University, Taipei, 3Department of Nursing,
Changhua Christian Hospital, Changhua, Taiwan, 4Department of Health Systems Science, University of Illinois at
Chicago and 5Resurrection University & Fellow of the American College of Healthcare Executives, Chicago, Illinois, USA

Abstract In this study, we developed and tested the psychometric properties of the Chinese-version Quality of Nursing
Work Life Scale along seven subscales: supportive milieu with security and professional recognition, work
arrangement and workload, work/home life balance, head nurse’s/supervisor’s management style, teamwork
and communication, nursing staffing and patient care, and milieu of respect and autonomy. An instrument-
development procedure with three phases was conducted in seven hospitals in 2010–2011. Phase I comprised
translation and the cultural-adaptation process, phase II comprised a pilot study, and phase III comprised a
field-testing process. Purposive sampling was used in the pilot study (n = 150) and the large field study
(n = 1254). Five new items were added, and 85.7% of the original items were retained in the 41 item Chinese
version. Principal component analysis revealed that a model accounted for 56.6% of the variance with
acceptable internal consistency, concurrent validity, and discriminant validity. This study gave evidence of
reliability and validity of the 41 item Chinese-version Quality of Nursing Work Life Scale.

Key words exploratory factor analysis, instrument development, nursing work environment, psychometric testing, quality
of work life, reliability and validity, Taiwan.

INTRODUCTION percentage than in Portugal (77%), Brazil (75%), Canada

(71%), and the USA (68%). Inordinate workloads, low
Quality of work life (QWL) is the perception of an organi-
pay, benefits and incentives, and lack of recognition were
zation’s staff about the physical and psychological desirabil-
among the particular issues cited as contributing factors
ity of their work environment and working conditions (Sarah
(International Council of Nurses, 2011). In Taiwan, although
et al., 2012). It is an umbrella concept that covers salary,
nurses are protected by laws governing maximum working
working hours, work environment, career prospects, and
hours and mandated time off (Council of Labor Affairs,
interpersonal relationships among the workplace, the organi-
Executive Yuan Taiwan, 2008), they are nonetheless com-
zation, and society (Martel & Dupuis, 2006; Vagharseyyedin
monly asked to work overtime and are assigned excessive
et al., 2011). Past research has indicated that QWL is nega-
workloads (Council of Labor Affairs, Executive Yuan
tively associated with job stress (Mosadeghrad et al., 2011;
Taiwan, 2011). In 2008, the average turnover rate of Taiwan-
Bragard et al., 2012), turnover intention (Mosadeghrad et al.,
ese nurses was 19.9%, and the average vacancy rate reached
2011; Almalki et al., 2012), and depression symptoms (Wang,
nearly 7% in 515 hospitals nationwide (Chang & Yu, 2010),
2009), and positively with productivity (Nayeri et al., 2011)
higher values than the ideal standards of 10% and 5%,
and patient safety (Mitchell, 2012).
respectively (National Audit Office & Taiwan, 2013).
In a 2009 study by the International Council of Nurses that
A psychometrically-sound tool to measure our nurses’
that examined nurses’ expectations and needs globally, the
QWL is needed to further explore the relationship between
proportion of nurses in Taiwan who believed that nursing was
QWL and high turnover rate. We found an existing tool, the
“better today than five years ago” ranked ninth in the 11
Chinese version of the Nursing Practice Environment Scale
countries surveyed (International Council of Nurses, 2011).
(C-NPES; Chiang & Lin, 2008), which was translated and
Only 53% of Taiwanese nurses said they were “very likely” to
revised from the Practice Environment Scale of the Nursing
remain in the nursing profession in five years’ time, a lower
Work Index (PES-NWI; Lake, 2002). The C-NPES is a
Chinese language instrument and consists of 30 items that
Correspondence address: Yu-Tzu Dai, Department of Nursing, College of Medicine, fall under five subscales: management and leadership,
National Taiwan University, No. 1, Section 1, Jen-Ai Road, Taipei 10051, Taiwan.
nursing professional development, nursing quality, staffing
Email: [email protected]
Received 4 February 2013; revision received 12 September 2013; accepted 13 and resource adequacy, and participation in hospital affairs.
September 2013. However, the C-NPES does not include the items of working

© 2014 Wiley Publishing Asia Pty Ltd. doi: 10.1111/nhs.12099

Quality of Nursing Work Life Scale 299

conditions and work/home life balance, rendering the instru- Setting and participants
ment less than ideal for use with Taiwanese nurses in light of
Data were collected over seven acute care private hospitals,
their specific complaints of overtime and excessive work-
including a medical teaching center (1455 beds), two regional
loads (Council of Labor Affairs, Executive Yuan Taiwan,
teaching hospitals (202 & 262 beds), and four district non-
2011). This means there is a gap in the existing Chinese
teaching hospitals (93, 136, 157, & 190 beds) in central
version of the instrument: it fails to measure these working
Taiwan. A total of 1254 nurses were recruited for the field
conditions and work/home life balance issues in Taiwan.
study using purposive sampling. Inclusion criteria were
To address the lack of a culturally-sensitive questionnaire,
nurses who provided bedside care and worked rotating shifts;
we searched the MEDLINE, CINAHL, and PsycINFO elec-
exclusion criteria were nurses still in the probation period,
tronic databases for relevant articles published from 2005 to
those working in management, and those who had already
2011. Three tools related to QWL were identified: the Brooks
participated in our pilot study.
Quality of Nursing Work Life Survey (BQNW; Brooks &
A majority of the sample was female (98.4%), single
Anderson, 2005), the Work-Related Quality of Life Scale
(70.4%), and childless (77.3%). Nurses ranged in age from 21
(WRQoL; Van Laar et al., 2007), and the Self-Evaluation of
to 49 years (mean: 28.7 ± 4.5), and reported an average sen-
Working Life Quality questionnaire (SEQWL; Ventegodt
iority of 5.9 ± 4.6 years. Most had baccalaureate degrees as
et al., 2008).
their highest level of education (63.4%) and were non-
The BQNW assesses QWL for nurses, and is a self-report
religious (60.6%). Nurses were also asked to report their
questionnaire containing 42 Likert-type items scored from 1
nursing career ladder level, referring to a system guiding the
to 6 (1 = strongly disagree, 6 = strongly agree) along four
responsibilities and capacities of Taiwanese nurses, which is
subscales: work life/home life (i.e. work/home life balance),
overseen by the Taiwan Nurses Association. Level N desig-
work design (i.e. responsibilities and duties), work context
nates novice nurses; N1, performing general patient care; N2,
(i.e. practice environment), and work world (i.e. sociocul-
participating in critical patient care; N3, being responsible for
tural influences) (Brooks & Anderson, 2004). Cronbach’s
clinical educational activities and executing integrated
alpha for the subscales ranges from 0.56 to 0.88, and the
nursing care; and N4, nursing administration and conducting
BQNW shows high test–retest reliability (r = 0.90) (Brooks
nursing research. The most represented nursing career ladder
& Anderson, 2004). The WRQoL is also a self-report ques-
level was N2 (43.1%) (Table 1).
tionnaire, but it assesses QWL for healthcare workers in
general. It contains 23 Likert-type items scored from 1 to
5 (1 = strongly agree, 5 = strongly disagree) along five Ethical considerations
subscales: job and career satisfaction, general well-being,
home–work interface, stress at work, control at work, and This study, including the pilot study and the field test, was
working conditions (Van Laar et al., 2007). Cronbach’s approved by the institutional review boards of Changhua
alpha for the overall scale is 0.91, and those for the Christian Hospital, and authorization for use of all instru-
subscales range from 0.76 to 0.91 (Van Laar et al., 2007). ments was obtained from their authors or publishing
The SEQWL assesses QWL for the public: it is a self-report company.
questionnaire containing 108 Likert-type items scored from
1 to 5 (1 = very good, 5 = very bad) along four subscales – Procedure and data collection
quality of life, mastery, fellowship, and creation of real value
– and has high criterion-related validity (r = 0.69, The C-QNWL was developed in three phases over 11 months
P = 0.0001) (Ventegodt et al., 2008). We decided to select in 2010–2011 (Fig. 1).
the nursing-specific BQNW for evaluating nurse QWL in
this study.
Phase I: translation and cultural-adaptation process
The BQNW was published in 2005 (Brooks & Anderson,
2005), and has been employed in four studies to date (Brooks The translation and cultural-adaptation process followed
& Anderson, 2004; Brooks et al., 2007; Almalki et al., 2012; typical four step guidelines for the cross-cultural adaption of
Lee et al., 2013). The BQNW situates QWL within the self-report measures (Beaton et al., 2000; Kutlu et al., 2012).
context of O’Brien-Pallas and Baumann’s (1992) model of First, two bilingual translators, whose first language was
influences on the quality of nurses’ work life, and measures Chinese, completed independent forward translations of the
working conditions and work/home life balance. BQNW from English to Chinese. One translator was the first
In this study, we developed and tested the psychometric author, and was familiar with the concept of QWL; the other
properties of the Chinese-version Quality of Nursing Work was a Taiwanese physician and certified translator who was
Life Scale (C-QNWL). not familiar with the concept of QWL. We focused on trans-
lating the conceptual meaning of the items, rather than the
word-for-word literal translations.
METHODS
Second, back translations were performed by two bilingual
translators – a Hong Kong physician and a Filipino research
Study design
associate – whose first language was English. They created
Psychometric properties of the C-QNWL were estimated independent translations of the Chinese instrument into
using a descriptive and correlational design. English.

© 2014 Wiley Publishing Asia Pty Ltd.

300 Y-W. Lee et al.

Table 1. Participants’ demographic characteristics (n = 1254) (answer the phone) and 34 (standard operation procedure)
based on these interviews. Issues mentioned by at least three
Variables Mean (SD) Range % nurses during the interviews were considered significant: eight
such issues were observed. After the research team discussed
Age 28.7 ± 4.5 21–49 how relevant the content of these eight items were to the scale,
Seniority 5.9 ± 4.6 two were removed, and six new items, for which there were no
Sex
equivalents on the BQNW (items 43–48), were developed and
Female 98.4
Male 1.6
added at the end of the scale. These modifications produced
Marital status the 45 item C-QNWL (second version).
Single 70.4 We continued the instrument development by reviewing
Married 28.5 the characteristics of the 45 items and inter-item associations.
Divorced 1.0 Item 21 (friendship) had the highest mean (5.09) and rela-
Others 0.1 tively low variance (0.59), but we retained the item because
Have children of its importance to the QWL concept. Three items – items 4
Yes 22.7 (rotating schedules), 40 (find the same job), and 42 (impacts
No 77.3 patients/families) – had low inter-item (< 0.30) and item-total
Education level
correlations (< 0.20), indicating minimal contribution to the
Diploma 32.5
Bachelor 63.4
QWL concept, and so we deleted these three items. Items 45
Graduate 0.6 (take leave) and 46 (arrange holidays) had high inter-item
Others 3.5 correlation (0.83); the former was deleted to avoid redun-
Religion dancy. Four items were removed overall, and the resultant
None 60.6 41-item C-QNWL, the final version, was employed in the
Tao 21.4 remaining analyses.
Buddhist 10.5
Christian 5.1
Other 2.4 Phase II: pilot study
Nurses’ career ladder level
We conducted a pilot study using the 41-item C-QNWL, with
N 16.0
N1 32.0
data collected using purposive sampling at seven acute care
N2 43.1 hospitals in central Taiwan. Nurses who provided bedside
N3 8.7 care in rotating shifts and acted as opinion leaders within
N4 0.2 their nursing units were invited to evaluate the clarity of each
item of the C-QNWL. The sample size, 150 nurses, provided
N, novice; N1, the first ladder level; N2, the second ladder level; approximately one-tenth of the proposed field-testing study
N3, the third ladder level; N4, the fourth ladder level; SD, standard size, as per the recommendations of Pett et al. (2003). A 94%
deviation. response rate (141/150) was achieved. The respondents rated
item clarity using a 10-point rating scale, ranging from 1 (not
at all clear) to 10 (very clear). Nurses were further invited to
Third, one language expert and the four translators dis- offer written suggestions to improve item clarity for any item
cussed conceptual, semantic, and experiential equivalences rated < 7. An assessment of 2 week test–retest reliability
between the English and the Chinese instruments until they yielded a reliability r value of 0.79 (P < 0.01).
arrived at consensus. Items 5 (leave for family), 8 (unlicensed
support personnel), and 17 (unlicensed support personnel)
Phase III: field-testing process
were deleted, because Taiwan hospital regulations do not
specifically identify family-related leave within leave guide- A sample size of 450–675 was calculated as sufficient to meet
lines, and nurse aides are seldom employed. The process the criteria for a robust factor analysis, requiring at least
resulted in the first version of the C-QNWL, which contained 10–15 participants per item (Pett et al., 2003). To prevent
39 items. problems arising from a low response rate, the population
Fourth, individual 30–60 min content-validation interviews was oversampled to ensure an adequate sample size.
were conducted with six nurses. They had a mean of 7.6 years Three instruments – the developed C-QNWL, the Chinese
of work experience, and their specialties covered internal version of the second edition of the Beck Depression Inven-
medicine care, intensive care, pediatric care, respiratory care, tory (C-BDI-II), and the C-NPES – and two copies of the
and emergency care. These interviews were conducted to informed consent document were put in an envelope and
assess how clear the items in the C-QNWL were, and to delivered to nurses’ workplaces. Nurses were invited to sign
identify additional areas within the domain of nursing work the informed consent and complete the questionnaires;
life for which new items were needed. Adding items demands participants were only identified by a study identification
careful consideration of the trade-off between maintaining number. Weekly follow-up reminders and an institutional
conceptual equivalence between the source and target lan- review board-approved incentive payment of NT$100
guages and enhancing the measure’s cultural validity (Leplège (approximately US$3.33) were provided to encourage a
& Verdier, 1995). Significant revisions were made for items 13 higher response rate.

© 2014 Wiley Publishing Asia Pty Ltd.

Quality of Nursing Work Life Scale 301

Source scale (BNQW, 42 items, English)

1. Forward translation: two independent translators

Integrated translations

2. Back translation: two independent translators

Phase I

3. Consensus discussion: one language expert and the four translators

4. Content validation interview: six nursing professionals

First version of the C-QNWL (39 items)

Second version of the C-QNWL (45 items)

The final version of the C-QNWL (41 items)

Phase II Pilot study (n =150) 10-point item clarity, test-retest reliability

Figure 1. Chinese-version Quality of

Nursing Work Life Scale (C-QNWL) trans- Phase III Internal consistency reliability, concurrent
Field test (n = 1,254) validity, exploratory factor analysis,
lation, cultural adaptation, and field-testing
discriminant validity.
process.

Instruments poration (Chen, 2000). The C-BDI-II is a 21-item, self-report

questionnaire, with each item scored from 0 to 3, and higher
C-QNWL values indicating increasing severity of depression over the
most recent 2 weeks (Beck & Steer, 1996). Its reported inter-
The final version of the C-QNWL is a self-report question-
nal consistency is 0.92, and its construct validity has been
naire consisting of 41 items answered on a six-point Likert-
demonstrated to be satisfactory using the Rasch measure-
type scale (1 = strongly disagree, 6 = strongly agree) adapted
ment model (Pan & Hsu, 2008).
from the 42-item BQNW (Brooks & Anderson, 2005) via the
translation process and 39- and 45-item intermediate ver-
C-NPES
sions outlined above.
The C-NPES (Chiang & Lin, 2008), a measure of the nursing
work practice environment, was chosen to measure the dis-
C-BDI-II criminant validity of the C-QNWL. The C-NPES is a self-
report, four-point Likert-type scale (1 = strongly disagree,
Work/home life imbalance is a risk factor for the onset of
4 = strongly agree) consisting of 30 items over five subscales:
depression (Couser, 2008). Research has indicated that the
management and leadership, nursing professional develop-
QWL of workers is significantly impaired by depressive
ment, nursing quality, staffing and resource adequacy, and
symptoms (Wang, 2009). Hospital nurses are known to work
participation in hospital affairs. Cronbach’s alpha for the
in psychologically- and physically-demanding work environ-
subscales ranges from 0.65 to 0.87, and the validity was
ments, which could lead to depressive symptoms (Gao et al.,
obtained using a content validity index and principal compo-
2012). We believed an instrument targeting such symptoms
nent analysis (PCA) of a five-factor structure, which
would provide a strong measure of concurrent validity. We
explained 47.89% of the variance (Chiang & Lin, 2008).
chose the C-BDI-II (Chen, 2000) for this purpose, which is
used to assess Diagnostic and Statistical Manual of Mental
Data analysis
Disorders, Fourth Edition, depressive symptomatology in
adolescents and adults. The C-BDI-II was previously trans- Data were analyzed using SPSS 17.0 software (SPSS,
lated into Chinese by the Chinese Behavioral Science Cor- Chicago, IL, USA). Missing items were treated by item mean

© 2014 Wiley Publishing Asia Pty Ltd.

302 Y-W. Lee et al.

Table 2. Inter-item correlations and corrected item-total correlations grouped into seven factors

Inter-item correlation† Item-total correlation†

Factor Item (mean) (mean)

1. Supportive milieu with job security and 19, 30, 31, 34, 35, 36, 37, 38, 39, 41 0.25–0.57 (0.39) 0.51–0.67 (0.58)
professional recognition
2. Work arrangement and workload 12, 10, 13, 43‡, 44‡, 46‡ 0.22–0.53 (0.32) 0.42–0.52 (0.48)
3. Work/home life balance 1, 2, 3, 6, 7 0.43–0.77 (0.54) 0.58–0.77 (0.67)
4. Head nurse’s/supervisor’s management style 18, 20, 22, 26, 27, 33, 48‡ 0.40–0.71 (0.54) 0.57–0.80 (0.68)
5. Teamwork and communication 21, 23, 24, 25 0.31–0.70 (0.50) 0.58–0.76 (0.62)
6. Nursing staffing and patient care 9, 14, 15, 16, 47‡ 0.35–0.49 (0.44) 0.49–0.66 (0.58)
7. Milieu of respect and autonomy 11, 28, 29, 32 0.27–0.70 (0.40) 0.40–0.65 (0.52)

Item 45 was deleted due to redundancy; original item numbers have been retained. †Pearson’s correlation; ‡items developed post-pilot study.

substitution; this procedure does not distort the data when rotation at delta = 0 was chosen instead of orthogonal rota-
the number of missing items is less than 5% of the total tion, because we anticipated the concepts within QWL to be
(Streiner & Norman, 2008); there was less than 4% in the highly correlated (O’Brien-Pallas & Baumann, 1992). Factors
present study. were extracted based on eigenvalues being > 1, results of a
scree plot, and their conceptual meanings. Factor loadings of
> 0.3 were used (Kline, 1994; Pett et al., 2003).
Descriptive statistics and item analysis Discriminant validity was examined by calculating
Demographic characteristics were described in terms of per- Pearson’s correlation coefficient of the C-QNWL with the
centage and mean. Item analysis was conducted in terms of C-NPES.
means, variances, inter-item correlations, corrected item-total
correlations, and the change in the Cronbach’s alpha if the
item was deleted. A mean close to the center of the range of RESULTS
possible scores with a relatively high variance was desirable A very good response rate was achieved for the C-QNWL
(DeVellis, 2012), and inter-item correlations of 0.30–0.80 (90.9%), the C-BDI-II (89.4%), and the C-NPES (84.9%),
(Pett et al., 2003) and corrected item-total correlations of with 75.2% of scales employed for concurrent validity and
0.20–0.80 (Streiner & Norman, 2008) were considered satis- discriminant validity analysis after the removal of unusable
factory. Inter-item correlations of less than 0.3 indicated an ones. The results provide strong evidence for the reliability
item that had little in common with other items, whereas and validity of the C-QNWL, thus supporting the study aim.
correlations higher than 0.80 could signify redundant items
(Pett et al., 2003).
Correlations between items and groups of items
Internal consistency reliability Inter-item correlations ranged from 0.22 to 0.77 (P < 0.01),
and item-total correlations grouped over seven factors
The internal consistency reliability of the C-QNWL was
ranged from 0.40 to 0.80 (P < 0.01) (Table 2).
assessed using Cronbach’s alpha; a value greater than 0.7 was
Inter-correlations among the factor-based scales were sig-
taken as satisfactory (Streiner & Norman, 2008; DeVellis,
nificant (P < 0.01), with only one exception; correlations
2012).
ranged from 0.19 to 0.68. These seven factors also signifi-
cantly correlated with the total C-QNWL (P < 0.01)
Concurrent validity (Table 3).

Concurrent validity was examined by calculating Pearson’s

correlation coefficient for the C-QNWL and the C-BDI-II. Internal consistency reliability analysis
Overall scale internal consistency reliability was high
Construct validity (α = 0.93), with good subscale reliability (α = 0.72–0.89)
Construct validity was evaluated using exploratory factor (Table 4).
analysis, together with tests of discriminant validity. The
Kaiser–Meyer–Olkin (KMO) test of sampling adequacy and
Concurrent validity
Bartlett’s test of sphericity for evaluating the correlation
matrix were examined prior to conducting factor analysis. A We observed a significant and weakly-negative Pearson’s
KMO value above 0.9 was considered excellent for factor correlation coefficient between the C-QNWL and the
analysis (Pett et al., 2003). PCA with oblique direct oblimin C-BDI-II (r = −0.22, P < 0.01).

© 2014 Wiley Publishing Asia Pty Ltd.

Quality of Nursing Work Life Scale 303

Table 3. Inter-correlations among the factor-based scales in the Chinese-version Quality of Nursing Work Life Scale

Factor
Factor 1 2 3 4 5 6 7

1. Supportive milieu with job security and professional recognition 1.00

2. Work arrangement and workload 0.25** 1.00
3. Work/home life balance 0.48** 0.28** 1.00
4. Head nurse’s/supervisor’s management style 0.66** 0.19** 0.41** 1.00
5. Teamwork and communication 0.48** −0.02 0.25** 0.54** 1.00
6. Nursing staffing and patient care 0.63** 0.39** 0.55** 0.47** 0.30** 1.00
7. Milieu of respect and autonomy 0.68** 0.19** 0.45** 0.56** 0.43** 0.53** 1.00
Full scale 0.88** 0.47** 0.69** 0.79** 0.55** 0.79** 0.76**

**P < 0.01.

Construct validity patient care, and milieu of respect and autonomy. The
BQNW “work context” dimension was also split into three
Exploratory factor analysis factors: supportive milieu with security and professional
recognition, head nurse’s/supervisor’s management style,
The KMO value was found to be 0.94 and Bartlett’s test
and teamwork and communication. The BQNW “work
of sphericity was significant (χ2 = 23614.59, d.f. = 820,
world” dimension disappeared entirely, with the two items
P < 0.001). The PCA generated nine components, but it was
that had fallen under “work world” loading with supportive
difficult to assign meaning to some of them. Close inspection
milieu with security and professional recognition in the
of the scree plot and eigenvalues of > 1 showed mild discon-
C-QNWL.
tinuity at values five, seven, and eight, and so we targeted our
The findings of our study reflect several ways in which
analysis for five-, seven-, and eight-factor solutions. Meaning
nursing in Taiwan differs from the West. For example, in
was most clearly interpretable for the solution of seven
Taiwan, nurse aides are not employed in most hospitals;
factors, which explained 56.6% of the variance in the QWL of
family-related leave is not available; national nursing short-
Taiwanese nurses (Table 4).
ages make it easy to find and transfer to a job in another
Factor 1 (10 items) was labeled “supportive milieu with
hospital; and Taiwanese nurses are more concerned about the
security and professional recognition”, factor 2 (6 items) was
work conditions, management style, and issues of respect,
labeled “work arrangement and workload”, factor 3 (5 items)
recognition, and support in their work environment.
was labeled “work/home life balance”, factor 4 (7 items) was
The seven factors found in our investigation exhibited
labeled “head nurse’s/supervisor’s management style”, factor
some similarities with the results of past research. Previous
5 (4 items) was labeled “teamwork and communication”,
authors have independently identified our work life/home
factor 6 (5 items) was labeled “nursing staffing and patient
life (O’Brien-Pallas & Baumann, 1992; Brooks & Anderson,
care”, and factor 7 (4 items) was labeled “milieu of respect
2004; Van Laar et al., 2007) and work arrangement and work-
and autonomy”.
load (O’Brien-Pallas & Baumann, 1992; Hsu & Kernohan,
2006; Van Laar et al., 2007) dimensions as essential compo-
Discriminant validity analysis nents of the QWL concept.
The extracted factor, milieu of respect and autonomy, has
A significant positive Pearson’s correlation between the
not been previously described in the QWL-related literature;
C-QNWL and the C-NPES (r = 0.72, P < 0.01) provided
however, autonomy is known to be an important component
evidence of discriminant validity.
of nurses’ job satisfaction and professional status (Varjus
et al., 2011; Papathanassoglou et al., 2012). Conflicts regard-
ing hierarchy and power (Reeves et al., 2008; Holyoake,
DISCUSSION
2011) still exist between Taiwanese doctors and nurses, and
The factor structure determined in this study shared some nurses work in a profession that traditionally carries less
similarities, but many differences, with that of Brooks and respect and autonomy.
Anderson’s study (2005); 85.7% of Brooks and Anderson’s The C-QNWL had acceptable internal consistency, with a
BQNW items were retained in the final version of the Cronbach’s alpha of 0.93 for the total scale, and those for the
C-QNWL. Exploratory factor analysis revealed a seven- seven factors ranging from 0.74 to 0.89; this is even slightly
factor structure for the C-QNWL, in which the scale items higher than the alpha reported for the original BQNW. We
were recategorized, and all of Brooks and Anderson’s factors suspect the improved consistency was directly attributable to
were renamed, with the exception of work life/home life. our tailoring of the C-QNWL to be culturally appropriate
In this study, the items comprising Brooks and Anderson’s through careful translation and creation of new items. The
BQNW “work design” dimension were split into three test–retest reliability analysis of the C-QNWL confirmed
factors: work arrangement and workload, nursing staffing and adequate stability for the overall scale.

© 2014 Wiley Publishing Asia Pty Ltd.

304 Y-W. Lee et al.

Table 4. Exploratory factor analysis results (n = 1254)

Factor loadings
Factor F1 F2 F3 F4 F5 F6 F7

Factor 1: Supportive milieu with job security and professional recognition (10 items)
35. The security department provides a secure environment 0.69 – – – – – –
38. Society has the correct image of nurses 0.67 – – – – – –
41. Job is secure 0.64 – – – – – –
37. Upper-level management has respect for nursing 0.62 – – – – – –
39. Salary is adequate for my job, given the current job market conditions 0.53 – – – – – –
36. Safe from personal harm 0.51 – – – – – –
34. Nursing policies and procedure facilitate my work 0.38 – – – – – –
31. Support to attend in-service and continuing education programs 0.36 – – – – – –
19. Adequate patient care supplies and equipment 0.33 – – – – – –
30. Access to degree-completion programs 0.33 – – – – – –
Factor 2: Work arrangement and workload (6 items)
43. Difficulty to take sick leave† – 0.74 – – – – –
44. Difficulty to arrange holidays† – 0.70 – – – – –
46. To participate in many lectures that occupy my after work time† – 0.70 – – – – –
12. Many non-nursing tasks† – 0.45 – – – −0.43 –
10. Workload is too heavy† – 0.42 – – – −0.40 –
13. Many interruptions in work routine† – 0.44 – – – −0.48 –
Factor 3: Work/home life balance (5 items)
2. Able to arrange for family care – – −0.87 – – – –
1. Able to balance work with family needs – – −0.82 – – – –
6. Able to arrange day care for family – – −0.78 – – – –
7. Able to arrange sick leave when a family member is unwell – – −0.70 – – – –
3. Energy left after work – – −0.69 – – – –
Factor 4: Head nurse’s/supervisor’s management style (7 items)
48. Head nurse’s/supervisor’s mood is adequate – – – 0.85 – – –
26. Receiver feedback on my performance from my head nurse/supervisor – – – 0.81 – – –
18. Good communication with head nurse/supervisor – – – 0.81 – – –
33. Recognized for my accomplishments – – – 0.71 – – –
27. Participation in decision-making process – – – 0.64 – – –
20. Adequate supervision – – – 0.59 – – –
22. Access to career-advancement opportunities – – – 0.45 – – –
Factor 5: Teamwork and communication (4 items)
24. I belong in the work team – – – – 0.81 – –
23. Teamwork in my work setting – – – – 0.72 – –
21. Friendships are important – – – – 0.72 – –
25. Communication with other therapists – – – – 0.63 – −0.31
Factor 6: Nursing staffing and patient care (5 items)
14. Enough time to do my job well – – – – – −0.62 –
15. Enough nursing staff – – – – – −0.57 –
16. Good quality of patient care – – – – – −0.56 –
47. Adequate mealtimes – – – – – −0.53 –
9. Satisfied with my job – – – – – −0.37 –
Factor 7: Milieu of respect and autonomy (4 items)
28. Respect by physicians – – – – – – −0.78
32. Good communication with physicians – – – – – – −0.73
29. High quality of lounge/break area/locker room – – – – – – −0.41
11. Autonomy to make patient care decisions – – – – – – −0.31
Eigenvalue 12.23 3.38 2.04 1.73 1.46 1.24 1.11
Explained variance (%) 29.85 8.24 4.99 4.23 3.55 3.01 2.71
Alpha reliability coefficient
Subscale 0.86 0.74 0.86 0.89 0.80 0.79 0.72
Full scale 0.93

†Reverse-scored item. Component loadings of < 0.30 have been suppressed. Original item numbers have been retained. F1–7, factors 1–7.

© 2014 Wiley Publishing Asia Pty Ltd.

Quality of Nursing Work Life Scale 305

The discriminant validity of the C-QNWL and the C-NPES Beaton DE, Bombardier C, Guillemin F, Ferraz MB. Guidelines for
was satisfactory. The C-QNWL had five additional factors the process of cross-cultural adaptation of self-report measures.
(supportive milieu with job security and professional recog- Spine 2000; 25: 3186–3191.
nition, work arrangement and workload, work/home life Beck AT, Steer RA. Beck depression inventory – II. Behav. Meas.
Lett. 1996; 3: 3–5.
balance, teamwork and communication, and milieu of respect
Bragard I, Dupuis G, Razavi D, Reynaert C, Etienne AM. Quality of
and autonomy) that were not found in the C-NPES. The work life in doctors working with cancer patients. Occup. Med.
nurse–physician relationship is an important issue to con- (Lond) 2012; 62: 34–40.
sider in QWL studies, and it was addressed in the C-QNWL Brooks BA, Anderson MA. Nursing work life in acute care. J. Nurs.
(milieu of respect and autonomy factor) and the PES-NWI Care Qual. 2004; 19: 269–275.
(i.e. the C-NPES source instrument, collegial nurse–physician Brooks BA, Anderson MA. Defining quality of nursing work life.
relations factor), but not in the C-NPES. Nurs. Econ. 2005; 23: 319–326.
Brooks BA, Storfjell J, Omoike O et al. Assessing the quality of
nursing work life. Nurs. Adm. Q. 2007; 31: 152–157.
Limitation of the study Chang M, Yu YM. Nurse Workforce & Nurse Practitioner System:
Vision and Challenges. Miaoli: National Health Research Institute,
We note that the adaptation and developments of this instru- 2010.
ment were conducted in Taiwan; the findings might not be Chen SY. Chinese Version Beck Depression Inventory? Taipei:
generalizable to all areas of the world. Chinese Behavioral Science Cooperation, 2000.
Chiang H, Lin S. Psychometric testing of the Chinese version of
Nursing Practice Environment Scale. J. Clin. Nurs. 2008; 18: 919–
Conclusions 929.
Council of Labor Affairs, Executive Yuan Taiwan. Labor standards.
Findings from this study give a starting point for evidence 2008. [Cited 21 Jan 2014.] Available from URL: http://www
for the acceptable reliability and validity of the 41-item .cla.gov.tw/cgi-bin/siteMaker/SM_theme?page=48eaf454.
C-QNWL, which was translated and culturally adapted to Council of Labor Affairs, Executive Yuan Taiwan. Call for improve-
accurately measure the perspectives of nurses in Taiwan on ment nurses’ working condition. 2011. [Cited 17 Jun 2013.] Avail-
their QWL. One application for which it is very suitable able from URL: http://www.cla.gov.tw/cgi-bin/Message/MM_msg
would be evaluations of the effectiveness of interventions in _control?mode=viewnews&ts=4dc90f21:5771.
improving nurses’ QWL.The C-QNWL was developed based Couser GP. Challenges and opportunities for preventing depression
in the workplace: a review of the evidence supporting workplace
on a sample of Taiwanese hospital nurses, and should be
factors and interventions. J. Occup. Environ. Med. 2008; 50: 411–
tested for suitability in other Asian countries with similar
427.
healthcare systems. Further study is necessary to continue to DeVellis RF. Scale Development: Theory and Applications (3th edn).
validate our results using confirmatory factor analysis. This Thousand Oaks, CA: Sage, 2012.
study provides a useful model of steps to take when adapting Gao YQ, Pan BC, Sun W, Wu H, Wang JN, Wang L. Depressive
an existing research instrument for the target culture. In symptoms among Chinese nurses: prevalence and the associated
addition, the revised instrument will be of interest to nurse factors. J. Adv. Nurs. 2012; 68: 1166–1175.
administrators and researchers in the area of QWL, particu- Holyoake DD. Is the doctor–nurse game still being played? Nurs.
larly in countries with a nursing culture similar to that of Times 2011; 107: 12–14.
Taiwan. Hsu M, Kernohan G. Dimensions of hospital nurses’ quality of
working life. J. Adv. Nurs. 2006; 54: 120–131.
International Council of Nurses. Nurses in the workplace: expecta-
ACKNOWLEDGMENTS tions and needs. 2011. [Cited 21 Jun 2013.] Available from URL:
http://www.icn.ch/news/nurses-in-the-workplace-expectations-and
We gratefully acknowledge the support of the multiple -needs/a-global-survey-of-nurses-may-2009-1335.html.
members and the directors of the nursing department from Kline P. An Easy Guide to Factor Analysis. New York: Routledge,
the seven hospitals: Shu-Chen Chang, Shu-Hui Chiu, Mei-Yu 1994.
Kang, Tsui-Fen Chang, Huei-Shan Chen, Shu-Chuan Lin, and Kutlu Y, Kucuk L, Findik UY. Psychometric properties of the
Turkish version of the Fraboni Scale of Ageism. Nurs. Health Sci.
Mei-Mei Hsieh.
2012; 14: 464–471.
Lake ET. Development of the Practice Environment Scale of the
Nursing Work Index. Res. Nurs. Health 2002; 25: 176–188.
CONTRIBUTIONS
Lee YW, Dai YT, Park CG, McCreary LL. Predicting quality of work
Study Design: YTD, YWL. life on nurses’ intention to leave. J. Nurs. Scholarsh. 2013; 45:
Data Collection and Analysis: YWL, LLM, YTD, GY, BAB. 160–168.
Manuscript Writing: YWL, LLM, YTD. Leplège A, Verdier A. The adaptation of health status measurement:
methodological aspects of the translation procedure. In: Shumaker
SA, Berzon RA (eds). The International Assessment of the Health-
REFERENCES related Quality of Life: Theory, Translation, Measurement and
Analysis. Oxford: Rapid Communications, 1995; 93–101.
Almalki MJ, FitzGerald G, Clark C. The relationship between Martel JP, Dupuis G. Quality of work life: theoretical and methodo-
quality of work life and turnover intention of primary health care logical problems, and presentation of a new model and measuring
nurses in Saudi Arabia. BMC Health Serv. Res. 2012; 12: 314–324. instrument. Soc. Indic. Res. 2006; 77: 333–368.

© 2014 Wiley Publishing Asia Pty Ltd.

306 Y-W. Lee et al.

Mitchell JI. Work life and patient safety culture in Canadian Reeves S, Nelson S, Zwarenstein M. The doctor-nurse game in the
healthcare: connecting the quality dots using national accredita- age of interprofessional care: a view from Canada. Nurs. Inq. 2008;
tion results. Healthc. Q. 2012; 15: 51–58. 15: 1–2.
Mosadeghrad AM, Ferlie E, Rosenberg D. A study of relationship Sarah B, Masoud J, Zeinab Masoumi J, Bahman T. Assessing the
between job stress, quality of working life and turnover intention quality of working life and levels of depression in athletes and
among hospital employees. Health Serv. Manage. Res. 2011; 24: non-athlete staff in National Olympic and Paralympics Academy
170–181. of Iran. Int. J. Acad. Res. Bus. Soc. Sci. 2012; 2: 508–513.
National Audit Office, Taiwan. The report of the clinical Streiner DL, Norman GR. Health Measurement Scales: A Practical
nursing workforce in Taiwan. 2013. [Cited 12 Jun 2013.] Available Guide to Their Development and Use. Oxford: Oxford Medical
from URL: http://www.audit.gov.tw/ezfiles/0/1000/attach/34/pta Publications, 2008.
_862_4339352_87973.pdf. Vagharseyyedin SA, Vanaki Z, Mohammadi E. The nature nursing
Nayeri ND, Salehi T, Noghabi AAA. Quality of work life and pro- quality of work life: an integrative review of literature. West. J.
ductivity among Iranian nurses. Contemp. Nurse 2011; 39: 106–118. Nurs. Res. 2011; 33: 786–804.
O’Brien-Pallas L, Baumann A. Quality of nursing worklife issues – a Van Laar D, Edwards JA, Easton S. The work-related quality of life
unifying framework. Can. J. Nurs. Adm. 1992; 5: 12–16. scale for healthcare workers. J. Adv. Nurs. 2007; 60: 325–333.
Pan AW, Hsu WL. Application of Rasch measurement model in the Varjus SL, Leino-Kilpi H, Suominen T. Professional autonomy of
construct validity of the Beck depression inventory-II. Formos. nurses in hospital settings – a review of the literature. Scand. J.
J. Med. 2008; 12: 284–291. Caring Sci. 2011; 25: 201–207.
Papathanassoglou EDE, Karanicola MNK, Kalafati M, Ventegodt S, Andersen NJ, Kandel I, Enevoldsen L, Merrick J. Sci-
Giannakopoulou M, Lemonidou C, Albarran JW. Professional entific research in the quality of working-life (QWL): generic
autonomy, collaboration with physicians, and moral distress among measuring of the global working life quality with the SEQWL
European intensive care nurses. Am. J. Crit. Care 2012; 21: e41–e52. questionnaire. Int. J. Disabil. Hum. Dev. 2008; 7: 201–217.
Pett MA, Lackey NR, Sullivan JJ. Making Sense of Factor Analysis: Wang XL. P01-300 impact of depressive symptoms on the work-life
The Use of Factor Analysis for Instrument Development in Health quality of financial workers in China. Eur. Psychiatry 2009; 24
Care Research. Thousand Oaks, CA: Sage, 2003. (Suppl. 1): S688.

© 2014 Wiley Publishing Asia Pty Ltd.