The Professionalism and Environmental Factors in the Workplace Questionnaire ® : development and psychometric evaluation

RESEARCH METHODOLOGY

The Professionalism and Environmental Factors in the Workplace

Questionnaire��: development and psychometric evaluation

Andrea Baumann & Camille Kolotylo

Accepted for publication 5 June 2009

Correspondence to A. Baumann:

e-mail: [email protected]

Andrea Baumann PhD RN

Director

Nursing Health Services Research Unit,

McMaster University Site, Hamilton,

Ontario, Canada

Camille Kolotylo PhD RN

Co-Investigator, Senior Research Associate

Nursing Health Services Research Unit,

McMaster University Site, Hamilton,

Ontario, Canada

BAUMANN A. & KOLOTYLO C. (2009)BAUMANN A. & KOLOTYLO C. (2009) The Professionalism and Environmental

Factors in the Workplace Questionnaire�: development and psychometric evalua-

tion. Journal of Advanced Nursing 65(10), 2216–2228.

doi: 10.1111/j.1365-2648.2009.05104.x

AbstractTitle. The Professionalism and Environmental Factors in the Workplace Ques-

tionnaire��: development and psychometric evaluation.

Aim. The aim of this paper is to describe the development and testing of a question-

naire intended to determine key professionalism attributes and key environmental

attributes that influence the professionalism of nurses in their practice environments.

Background. Rapid changes in the healthcare sector and human resource shortages

have had an impact on the stability of global work environments, making maintaining

professionalism a challenge. The literature consists of descriptive research, opinion

and theoretical papers.

Method. The Professionalism and Environmental Factors in the Workplace Ques-

tionnaire� was developed and tested from 2005 to 2007 in three phases: item gener-

ation, pretesting and pilot testing. Convenience sampling was used to obtain

representative samples of the target population in the pretest and pilot test. Mailed

survey methodology was used in the pretest and pilot test. Sample sizes for the pretest

and pilot test were 46 and 848 respectively.

Results. Psychometric testing indicated preliminary instrument validity and reliabil-

ity. Factor analysis resulted in stable factors that mirrored the conceptual basis of the

questionnaire. The results summarize nurses’ ratings of professionalism and envi-

ronmental attributes that play a role in their work lives.

Conclusion. The questionnaire helps nurses reflect on their practice and provides a

starting point for discussion, planning and implementation of methods to support

professionalism in practice and healthy work environments. It is internationally rel-

evant because professionalism is a construct that transcends culture. Confirmatory

factor analysis is needed to validate the results of this study. Testing with populations

in different settings and additional validity and reliability testing will strengthen the

questionnaire.

Keywords: healthcare, instrument development, nursing, Professionalism and

Environmental Factors in the Workplace Questionnaire�, psychometric evaluation,

survey

2216 � 2009 The Authors. Journal compilation � 2009 Blackwell Publishing Ltd

J A N JOURNAL OF ADVANCED NURSING

Introduction

Rapid changes in the healthcare sector and human resource

shortages have had an impact on the stability of global work

environments, making maintaining professionalism a chal-

lenge. Professionalism is a construct that transcends culture.

The constructs of professionalism and environment are

intricately connected. As early as 1915, Flexner (1915)

identified several characteristics of professionalism which

remain relevant today: knowledge, specialization, intellectual

and individual responsibility, and well-developed group

consciousness. Authors provide various perspectives on the

meaning of professionalism, including accountability (Batey

& Lewis 1982), autonomy (Ballou 1998), inquiry and

collaboration [Registered Nurses Association of Ontario

(RNAO) 2007] and ethics and values (Leddy & Pepper

1985). At the same time, positive work environments are

characterized by strong leadership, work demands matching

the skills of the person, balance between effort and reward,

and safety (Karasek 1979, Baumann et al. 2001).

The Professionalism and Environmental Factors in the

Workplace Questionnaire� (PEFWQ�) was developed to

integrate and measure key environmental and professional-

ism attributes influencing nurses’ professionalism within a

conceptual framework developed from the literature (see

Figure 1). It was designed to provide baseline data for

individual reflection and collective dialogue within practice

settings, as well as a baseline for individual and collective

action planning. It can highlight concerns and stimulate

awareness and discussion of professionalism and its relation-

ship to the practice environment.

Background

Relevant published and grey literature, jurisdictional practice

standards, codes of ethics and other pertinent documents

were reviewed. Related databases, keywords and phrases

such as nurses, professionalism, and practice environment,

and other terms derived from the literature were used to

complete a supplemental review in 2008. The literature

consists of descriptive research, opinion and theoretical

papers focussing on aspects of professionalism such as

autonomy (Ballou 1998) and advocacy (Altun & Ersoy

2003). A seminal best practice guideline on professionalism

further informs the reader about key elements (RNAO 2007).

Existing instruments that encompass aspects of profession-

alism and the environment include the application of pro-

fessional nursing practice models (Aiken et al. 1997, Mark

et al. 2003) and the practice environment (Grindel et al.

1996, Aiken & Patrician 2000); the effect of healthcare

Environmental attributes

Environmental cultureand climate

r = 0·77

Professionalism

Professionalism attributes (nurse)

Autonomy r = 0·65

0·52

0·83 0·90

Knowledge r = 0·62

Competence r = 0·65

Professionhood r = 0·66

Accountability r = 0·54

Collaborative practice r = 0·54

Advocacy r = 0·57

Commitment r = 0·61

Control of nursing practicer = 0·65

Quality of nursing worklifer = 0·66

Professional supportr = 0·71

Shared governancer = 0·74

Figure 1 Conceptual framework and Pearson’s Product–Moment correlations between the sum of the environmental factors, the sum of the

professionalism factors, each factor, and the Professionalism and Environmental Factors in the Workplace Questionnaire� (PEFWQ�). Note: all

correlations are statistically significant at P < 0Æ01 (2-tailed).

JAN: RESEARCH METHODOLOGY Professionalism and Environmental Factors in the Workplace Questionnaire�

� 2009 The Authors. Journal compilation � 2009 Blackwell Publishing Ltd 2217

reorganization on nurse (Davidson et al. 1997) and patient

outcomes (Anthony et al. 2004); autonomy (Dwyer et al.

1992, Kramer & Schmalenberg 2003) and teamwork

(Rafferty et al. 2001); magnet hospitals, professional practice

environments and nurse burnout (Aiken & Patrician 2000);

and collaborative practice (Weiss 1983, Weiss & Davis

1985). The environment has an impact on professionalism in

many ways. For example, when nurses feel they have access

to sufficient support, resources and information to complete

their work, they are likely to feel more accountable and

effective (Laschinger & Wong 1999, Baumann et al. 2001).

Nurses are involved in deriving and monitoring solutions for

environmental barriers to responsible care (Cohen et al.

1994). Better patient and staff outcomes result from shared

decision-making, improved working conditions, autonomy

over practice, organizational governance and effective inter-

disciplinary communication (Trofino 1996, Havens & Aiken

1999, Mark et al. 2003).

The PEFWQ� was designed to aid the individual practi-

tioner’s reflection on the concept of professionalism and the

impact of environment. Completion and analysis of the

questionnaire provides a starting point for collective discus-

sion, planning and implementation of methods to support

professionalism in practice and healthy work environments.

Lack of an instrument that relates professionalism to envi-

ronment establishes the need for this study.

The study

Aim

The aim of the study was to develop and test the psycho-

metric properties of the PEFWQ�, which is intended to

determine key professionalism attributes and key environ-

mental attributes that influence the professionalism of nurses

in their practice environments.

Methodology

This self-report instrument was developed and tested between

March 2005 and March 2007 in three phases: item generation,

pretesting and pilot testing. Convenience sampling and the

modified tailored design method� (Dillman 2007) was used for

pretesting and pilot testing. This method is a ‘set of procedures

for conducting successful self-administered surveys that pro-

duce both high quality information and high response rates’

(Dillman 2007, p. 29). Based on social exchange theory, it is a

scientific approach to conducting surveys that focuses on: (a)

the reduction of survey error; (b) methods of contacting and

communicating with participants; and (c) the customization of

survey procedures that build positive social exchange, which

encourages respondents to participate by establishing trust

while increasing the benefits and decreasing the risks of

participation (Dillman et al. 2009). The stages of instrument

development were directed by guidelines for scaling self-report

paper-and-pencil measures of latent social-psychological con-

structs (Netemeyer et al. 2003).

Phase 1: item generation

Instrument development began with a critique of the litera-

ture to: (a) define the construct and content domains; (b)

identify the theoretical basis for questionnaire items; (c)

identify the instrument’s purpose and objectives; (d) locate

items in existing instruments; and (e) formulate objectives

from which items were devised (Netemeyer et al. 2003). The

literature led to the development of a conceptual represen-

tation of professionalism with two main organizing aspects:

professionalism and environment. Each organizing aspect

consists of attributes: eight professionalism attributes, such as

knowledge and competence, and five environmental attri-

butes, such as control of nursing practice and shared

governance (see Table 1). Each attribute has several descrip-

tors from which item objectives and questionnaire items were

developed. For example, internal work motivation and

internal locus of control are among the descriptors for

commitment. The meaning and boundaries of the content

domain and related constructs were established and used to

shape the theoretical framework, constructs and attribute

definitions (Netemeyer et al. 2003).

A domain sampling model, in which the measure is a

sample of items from a larger hypothetical content domain

that it purports to measure, was used to generate items to tap

into the domains of the constructs (Netemeyer et al. 2003,

Waltz et al. 2005). The goal is to systematically sample all

content areas while ensuring items are relevant to and

representative of the targeted constructs. Content experts

and members of the target population helped establish the

content domain and definitions (Netemeyer et al. 2003).

Sources of item generation were the researchers, literature

and content experts from the target population.

Phase 2: pretest

Pretesting, December 2005 to July 2006, included item

analysis and validity and reliability testing.

Validity

Face validity is an informal review by non-experts to deter-

mine if the instrument looks as if it measures what it is

A. Baumann and C. Kolotylo


purported to measure (Litwin 1995). Two volunteers from

our workplace reviewed the instrument, commenting on

language and reading levels, title and instructions.

‘A measure has content validity when its items accurately

represent the thing…being measured’ (Vogt 2005, p. 59). It

includes an organized review of the questionnaire’s content to

ensure that it includes everything it should and does not

include anything it should not (Litwin 1995). In the first three

reviews, 13 experts linked each item with an objective that

guided instrument construction, assessed the relevance of

each item to the content of the objective, and determined if

the intended content domain was adequately and clearly

represented by the items (Netemeyer et al. 2003, Waltz et al.

2005). Six experts also participated in a focus group (Stewart

& Shamdasani 1990). An additional three experts reviewed

the final questionnaire.

To quantify the extent of agreement among the experts’

ratings of items (Lynn 1986, Waltz et al. 2005), a 4-Point

Ordinal Rating Scale (i.e. 1, totally irrelevant; 2, somewhat

relevant; 3, quite relevant; 4, extremely relevant) was used to

assess item relevance, expressed as an alpha coefficient

(Cronbach’s alpha). The Content Validity Index (CVI) is

the proportion of experts who rated each item as content

valid (i.e. a rating of 3 or 4; Lynn 1986) and ranges from 0Æ00

to 1Æ00; a 0Æ00 coefficient signifies lack of expert agreement

and 1Æ00 indicates total expert agreement. Item retention,

based on the CVI, is dependent on the number of experts and

amount of agreement (Lynn 1986). For example, if there

were six content experts, four of whom endorsed an item as

content valid, the CVI would be 0Æ67 and the item would be

retained (a > 0Æ50). To encompass the dimensions of these

broad constructs, five to nine items were retained per

subscale, based on item-relevance agreement (Pett et al.

2003).

Fifteen Registered Nurses (RN) and one licensed practical

nurse (LPN) or registered practical nurse (RPN) participated

in instrument evaluation. An RPN is the equivalent of an

LPN in the province of Ontario, Canada. Participants were

from various healthcare sectors (e.g. primary care, ambula-

tory care and public health) and nursing positions (e.g.

clinical nurse specialist, manager, staff nurse and consultant).

They held diplomas (n = 3), baccalaureate degrees (n = 6)

and graduate degrees (master’s, n = 5; doctorate, n = 1).

Their mean age was 49 years, with a range of 24–59 years.

All but three experts had been practising nursing for more

than 10 years.

Reliability

The introductory letter, questionnaire, participant informa-

tion/consent form and postcard reminders were distributed

according to the modified tailored design method� (Dillman

2007). A small incentive (i.e. a coffee voucher) and a request

for participation in reliability retesting were included with the

initial questionnaire. Willing participants returned signed

consent forms and contact information, and were mailed a

second questionnaire to complete in 2–14 days (Streiner &

Table 1 Pilot test: number of items, test and retest subscale item means and mean inter-item correlations, and test subscale inter-item

correlation range (r, Pearson’s correlations)

Subscales

No.

items

Item

mean�

Inter-item

r range*

Mean

inter-item r

Retest

item mean�

Retest mean

inter-item r

Environmental subscales

1. Control of nursing practice 5 3Æ80 0Æ33–0Æ57 0Æ43 3Æ91 0Æ42

2. Quality of nursing worklife 6 3Æ30 0Æ22–0Æ49 0Æ34 3Æ07 0Æ28

3. Professional support 7 3Æ42 0Æ37–0Æ58 0Æ48 3Æ40 0Æ53

4. Shared governance 7 3Æ21 0Æ45–0Æ70 0Æ55 3Æ30 0Æ53

5. Environment culture and climate 5 4Æ03 0Æ45–0Æ51 0Æ44 4Æ12 0Æ40

Professionalism subscales

1. Autonomy 5 4Æ23 0Æ30–0Æ53 0Æ41 4Æ23 0Æ35

2. Knowledge 5 4Æ23 0Æ42–0Æ65 0Æ57 4Æ23 0Æ52

3. Competence 9 4Æ01 0Æ25–0Æ67 0Æ40 4Æ10 0Æ41

4. Professionhood (Styles 1982) 6 3Æ90 0Æ31–0Æ62 0Æ40 3Æ90 0Æ40

5. Accountability 7 4Æ50 0Æ30–0Æ69 0Æ49 4Æ50 0Æ43

6. Advocacy 6 4Æ31 0Æ37–0Æ81 0Æ57 4Æ40 0Æ54

7. Collaborative practice 6 4Æ03 0Æ34–0Æ71 0Æ55 4Æ00 0Æ48

8. Commitment 8 3Æ70 0Æ46–0Æ82 0Æ57 3Æ80 0Æ53

Total 82 3Æ90 0Æ40–0Æ83 0Æ24 3Æ91 0Æ20

*Statistically significant P > 0Æ01 level (2-tailed).�Range 1–5, higher scores; more agreement.



Norman 2003). Test–retest reliability is an estimate of the

stability of the instrument over time (Litwin 1995). Prior to

reliability and item analysis, reverse order items were re-

scored.

Phase 3: pilot test

Pilot testing consisted of face and construct validity testing

and reliability and item analysis. Item reduction through

factor analysis was the focus of the pilot study. Survey

material, as per the pretest, was mailed to participants

according to the modified tailored design method� (Dillman

2007). Participants returning their consents and contact

information were sent a second questionnaire to complete in

2–14 days (Streiner & Norman 2003). Identification of the

underlying factor structure and comparison with the concep-

tual basis derived from the literature initiated the theory

development process (Tabachnick & Fidell 2007). Respective

authors gave permission to use copyrighted items.

Participants

Convenience sampling was used to obtain samples represen-

tative of the target population: practising RNs and LPN/

RPN, of either gender, fluent in English, in any practice set-

ting (Hulley et al. 2001), excluding non-practising nurses and

those not fluent in English. Two volunteers from our work-

place participated in face validity testing.

Sixteen RN and LPN/RPN participated in content validity

testing from December 2005 to July 2006. The literature

lacks consensus on the required number of experts for

instrument validation (Slocumb & Cole 1991), although five

or more judges are recommended for the detection of

marginal items (Netemeyer et al. 2003). As a result of the

complexity of the constructs, 13 experts (recruited by

ourselves through personal requests and email invitations)

judged questionnaire item-objective agreement.

The aim of the pretest, conducted from April to May 2006,

was the return of a minimum 30 usable questionnaires for

parametric statistics use (Peterson 2008). It consisted of

reliability testing (n = 46, return rate = 76%) and item

analysis. Nurse executives, known to us, agreed to contact

nurse managers in healthcare facilities and request that study

information be posted.

Approximately 3000 potential participants in Ontario and

Nova Scotia were available for participation in the pilot

study, conducted from October 2006 to February 2007,

through contact with the provincial nursing registration

bodies. The recommended number of participants per item

for factor analysis ranges from 5 (Viswanathan 2005) to 10

(Nunnally & Bernstein 1994). Given these recommendations

and the pretest return rate (76%), a sample size of 1302 was

calculated. The final sample size (n = 848, return rate of

35Æ4%) yielded eight participants for each of the 105 items in

the pilot study questionnaire, which is within the recom-

mended range of participants per item for factor analysis.

Final reliability testing, carried out from December 2006 to

February 2007, yielded a sample size of 111 and a return rate

of 68Æ7%. Initially, a questionnaire and information/consent

form was sent requesting participation; a second question-

naire was sent to those returning consent forms and complete

addresses. A large sample size (n > 100) mitigates substan-

tial deviations from normality, allowing for the use of

inferential statistics (Pett 1997, Peterson 2008).

Instrument

An initial pool of 394 items was generated by the

researchers, experts and thematic analysis of the literature.

Item reduction in the pretest (n = 187) and pilot test

(n = 105) left the final instrument with 82 items. The

generation of a large, over-inclusive item pool, as much as

three or four times the final scale, is recommended (DeVellis

2003, Netemeyer et al. 2003). Generated items were

consistent with the theoretical domains of the constructs.

Nine items were adapted, with permission for use, from

existing instruments (Blau 1985, Spreitzer 1995, College of

Nurses of Ontario 2006).

The result was the PEFWQ�, a self-administered 82-item

instrument with three sections: demographic data, key

environmental attributes (five subscales) and key profession-

alism attributes (eight subscales) in the workplace. Theoret-

ical definitions were developed for the attributes. Each item

uses a 5-point Likert Scale (from 1 = strongly agree to

5 = strongly disagree) response format, with a not applicable

option. Following data re-coding, higher scores indicated

greater agreement with the item statements. Scores were

transformed to standardized z and t scores (mean = 50,

SDSD = 10) to ensure positive scores and subscale score compa-

rability (Streiner & Norman 2003, see Table 2). The scores

are used to stimulate discussion about environmental and

professionalism factors that affect nurses’ professionalism in

practice.

This instrument is not intended as a personal profession-

alism measure or a diagnostic benchmarking tool. It is

designed to identify areas of concern to nurses and provide a

starting point for discussion. In instruments with many

dimensions, it is not appropriate to combine all of the items

to form a total instrument score; rather, the dimensions serve

as the basis for subscale construction and scores (Loiselle

et al. 2004, see Table 2).



Ethical considerations

Ethics review board approval was obtained prior to testing.

Following informed consent, respondents agreed to participate

in validity and reliability testing and one focus group. An

information/consent form, explaining the study’s purpose, risks

and benefits, participant responsibilities, withdrawal from the

study and anonymity accompanied the questionnaire for all

study phases. The voluntary nature of the study was emphasized

and confidentiality was assured. Participants had the opportu-

nity to contact the researchers by mail, telephone or email.

Results

Pretest and pilot test data were analysed using SPSSSPSS for Windows

Version 15.0� (SPSS Inc 2006). The data sets were examined

using descriptives, correlations, internal consistency and item

analysis. Pretest data were analysed with nonparametric

statistics (e.g. Spearman’s rank-order correlation) and pilot

test data were analysed with parametric statistics (e.g. Pear-

son’s product-moment correlation), including exploratory

factor analysis. Casewise deletion was applied in cases with

missing data; participants with any missing data were removed

from the analysis (Tabachnick & Fidell 2007, Cole 2008).

Face validity

Pretest and pilot test face validity comments led to instrument

reformatting, instruction clarification and the addition of

subscale headings. The instrument was thought to measure

what it was intended to measure.

Pretest

Content validity

Content validity testing resulted in item reduction, increased

item clarity and decreased redundancy. Problematic items were

deleted and replaced with items from the item pool, rewritten

and/or relocated to other subscales for better conceptual and

statistical fit. Item CVI, item analysis and reliability provided

support for item relocation. Following initial item reduction,

the number of items remained constant throughout validity

testing. The second review resulted in re-writing and re-testing

of five items. The third review resulted in deletion and

replacement of 13 items with tested items from the item pool

(Lynn 1986). Applied feedback included addition of an

example of response format use, a not applicable option, and

number and word anchors to the scale steps. The final evalu-

ation resulted in wording modification of three items.

Reliability

Pretest data exploration revealed non-normality; thus,

Spearman’s Rho stability coefficients were calculated. The

literature lacks agreed on prescriptions for correlations, rec-

ommending judging statistical significance, sample size and

magnitude (Viswanathan 2005). This study’s correlation

criteria are (a) very weak, 0Æ00–0Æ25; (b) weak, 0Æ26–0Æ49; (c)

moderate, 0Æ50–0Æ69; (d) strong, 0Æ70–0Æ89; and (e) very

Table 2 Pilot test: standardized t scores by subscale: means, standard deviations, ranges, and test and retest alpha coefficients and Pearson’s

correlations

Subscales t Scores mean� t Scores SDSD Actual ranges� Test Cronbach’s alpha Retest alpha Test–retest r*

Environmental subscales

1. Control of nursing practice 246Æ86 39Æ23 83–319 0Æ78 0Æ78 0Æ61

2. Quality of nursing worklife 293Æ76 37Æ30 68–381 0Æ75 0Æ73 0Æ70

3. Professional support 345Æ36 55Æ44 86–458 0Æ87 0Æ89 0Æ64

4. Shared governance 344Æ11 59Æ00 56–474 0Æ89 0Æ89 0Æ60

5. Environment culture and climate 247Æ03 39Æ00 84–312 0Æ79 0Æ73 0Æ50

Professionalism subscales

1. Autonomy 248Æ33 38Æ00 27–309 0Æ76 0Æ70 0Æ50

2. Knowledge 248Æ86 42Æ00 103–314 0Æ86 0Æ84 0Æ44

3. Competence 439Æ66 67Æ10 159–568 0Æ85 0Æ84 0Æ63

4. Professionhood 293Æ16 48Æ00 32–383 0Æ79 0Æ77 0Æ50

5. Accountability 348Æ04 55Æ00 81–414 0Æ86 0Æ82 0Æ51

6. Advocacy 296Æ01 53Æ30 49–365 0Æ88 0Æ88 0Æ50

7. Collaborative practice 297Æ72 52Æ30 34–378 0Æ88 0Æ84 0Æ60

8. Commitment 396Æ07 65Æ20 187–495 0Æ92 0Æ90 0Æ63

Total – – – 0Æ96 0Æ94 0Æ70

*P > 0Æ01 level (2-tailed).�Higher scores indicate more agreement with item statements.�t Scores = 10(z) + 50.



strong, 0Æ90–1Æ00 (Munro 2005, p. 249). The pretest stability

coefficients were statistically significant (except the commit-

ment subscale), strong (rs = 0Æ84, P < 0Æ01, 2-tailed) and the

subscale coefficients range from very weak to very strong

(rs = 0Æ22–0Æ90, P < 0Æ01, 2-tailed). Seven subscale coeffi-

cients are <0Æ69, possibly because of the number of items

(Streiner & Norman 2003).

Although no consensus exists regarding a minimum alpha,

0Æ70 is acceptable (Netemeyer et al. 2003, Field 2005). The

instrument’s internal consistency was a = 0Æ96, ranging from

0Æ75 to 0Æ92 for the subscales (see Table 2). Item analysis

resulted in subscale structure modification. Retained items

were clinically relevant and contributed to internal consis-

tency strength.

Pilot test

Sample demographics

The pilot test sample (n = 848) consisted of mostly female

(n = 782, 94Æ9%) RN (n = 675, 80Æ6%), with diplomas

(n = 573, 68Æ9%), between the ages of 40 and 59 years

(n = 484, 60Æ5%), with an average age of 44Æ6 years

(SDSD = 10Æ65, median = 51). Participants reported being in

the nursing profession for more than 21 years (n = 368,

44Æ3%) but at their current positions for <10 years

(n = 569, 68Æ1%). About half were employed in full-time

positions (n = 474, 57Æ4%) and about one-quarter in part-

time positions (n = 239, 28Æ9%). LPN/RPN (n = 154,

18Æ2%) comprised a small percentage of the sample (see

Table 3).

Table 3 Pretest and pilot test: participant

characteristics*Characteristic

Pretest

(n = 46)

Pilot

(n = 848)

Pilot retest

(n = 111)

Age

Range (years) 27–67 21–72 23–71

Mean, median (SDSD) 44Æ9, 43Æ5 (9Æ3) 44Æ6, 44 (10Æ7) 44Æ6, 44Æ5 (11Æ4)

Gender

Male 2 (4Æ3) 42 (5Æ0) 1 (0Æ9)

Female 44 (93Æ5) 782 (92Æ2) 108 (97Æ3)

Missing 1 (2Æ2) 24 (2Æ8) 2 (1Æ8)

Nursing designation

Registered nurse 36 (78Æ3) 675 (79Æ6) 94 (84Æ7)

Licensed practical nurse/registered

practical nurse

7 (15Æ2) 154 (18Æ2) 16 (14Æ4)

Master’s degree in nursing 3 (6Æ5) – –

Registered nurse (extended class) – 8 (0Æ9) 1 (0Æ9)

Missing – 11 (1Æ3) 9 (8Æ1)

Years in nursing

<10 7 (15Æ0) 239 (28Æ7) 33 (30Æ0)

11–20 18 (38Æ3) 225 (27Æ0) 22 (20Æ0)

>21 22 (46Æ8) 368 (44Æ3) 55 (50Æ0)

Employer

Acute care 34 (73Æ9) 565 (66Æ6) 65 (58Æ6)

Community care 3 (6Æ6) 34 (3Æ9) 6 (5Æ4)

Long-term care 6 (13Æ0) 47 (5Æ5) 7 (6Æ3)

Mental health care 1 (2Æ2) 5 (0Æ6) 1 (0Æ9)

Nursing agency – 144 (17Æ0) 25 (22Æ5)

Other – 36 (4Æ3) 7 (6Æ3)

Missing 2 (4Æ4) 17 (2Æ1) –

Positions in nursing

Staff nurse 30 (65Æ2) 563 (68Æ0) 61 (56Æ5)

Visiting nurse – 167 (20Æ2) 29 (26Æ9)

Manager 3 (6Æ5) 29 (3Æ4) 4 (3Æ7)

Clinical educator 4 (8Æ7) 17 (2Æ0) –

Acute care nurse practitioner 1 (2Æ2) 7 (0Æ8) –

Other 7 (15Æ3) 27 (3Æ2) 7 (10Æ2)

Missing 1 (2Æ2) 20 (2Æ3) 3 (2Æ7)

*Values are presented as n (%).



Reliability

Instrument test–retest reliability (n = 111) coefficients (i.e.

Pearson’s product-moment correlation) were statistically

significant and strong (r = 0Æ70, P < 0Æ01, 2-tailed), ranging

from weak to strong for the subscales (r = 0Æ44–0Æ70,

P < 0Æ01, 2-tailed), generally indicating scale stability (see

Table 2). A low stability coefficient (knowledge subscale;

r = 0Æ44) may have been caused by variation in participant

attention or motivation when retesting, which may have led

to variations in scores. Test–retest reliability estimates also

‘captures measurement error due to sampling time or occa-

sions’ (Salkind 2007, p. 994).

Internal consistency was estimated with Cronbach’s alpha,

average inter-item and corrected item-total correlations, and

wording redundancy (Netemeyer et al. 2003). The instru-

ment’s internal consistency reached the level for applied

research (a = 0Æ96; Nunnally & Bernstein 1994). The sub-

scales ranged from 0Æ75 to 0Æ92, indicating that 75–92% of

the scores could be attributed to reliable variance (see

Table 2).

Although consensus does not exist, corrected-item-total

correlations >0Æ30 are acceptable (Viswanathan 2005).

Measures of broad constructs such as professionalism may

have lower corrected item-total correlations, because ‘each

item is tapping into some aspect of a diverse domain’

(Viswanathan 2005, p. 172). The acceptable correlation for

each subscale in multidimensional scales varies; items with

low correlations are candidates for deletion (Netemeyer et al.

2003). Items with high corrected item-total correlations have

more variance relating to what the items have in common and

add more to the test’s reliability than items with low values.

Corrected item-total correlations were acceptable, ranging

from r = 0Æ30 to 0Æ60 for the instrument and 0Æ42–0Æ81 for

the subscales (Nunnally & Bernstein 1994). Items were

not eliminated because of lack of homogeneity with the

construct.

Cronbach’s alpha-if-item-deleted was 0Æ96 and ranged

from 0Æ70 to 0Æ91 for the subscales, indicating that all items

were contributing to the high reliability (Pett et al. 2003).

Cronbach’s alpha-if-item-deleted gives information about

item retention on a factor by indicating change to the alpha

coefficient if the item were deleted (Pett et al. 2003). Two

items were deleted because their removal increased Cron-

bach’s alpha substantially. The assessment of item fit (i.e.

trivial redundancy and conceptual fit) supported item dele-

tion. In trivial redundancy, two items are highly correlated

(i.e. >0Æ90), with minor wording or grammatical differences

(Viswanathan 2005). The conceptual fit is assessed by

evaluating the item with reference to the conceptual model

and item objectives (see Table 4). Pearson’s product–moment

correlations between each subscale and the total scale were

statistically significant and moderate to strong (r = 0Æ54–

0Æ77, P < 0Æ01, 2-tailed; see Figure 1). The sums of the

environmental subscales (0Æ83) and the professionalism

subscales (0Æ90) correlated strongly (P < 0Æ01, 2-tailed) with

the total scale (see Figure 1).

The literature lacks prescriptions for minimum inter-item

correlations. The mean inter-item correlation for a broad

construct may be as low as 0Æ15–0Æ25 (Netemeyer et al. 2003)

and may be a more useful internal consistency index than

coefficient alpha because of the attenuation paradox, espe-

cially in scales with many items (Loevinger 1954, Clark &

Watson 1995). Mean subscale inter-item correlations ranged

from 0Æ34 to 0Æ57, indicating that a single construct underlies

each subscale (Netemeyer et al. 2003, see Table 1).

Individual inter-item correlations should be moderate

(r = 0Æ15–0Æ50; Briggs & Cheek 1988, Netemeyer et al.

2003) and cluster narrowly around the mean inter-item

correlation to ensure dimensionality. A focus on subscale

unidimensionality and internal consistency ensures a more

homogeneous scale (Clark & Watson 1995). Subscale indi-

vidual inter-item correlations were statistically significant

(P < 0Æ01, 2-tailed), ranging from 0Æ22 to 0Æ82, and cluster-

ing around their respective mean inter-item correlations. This

indicates that the subscale items are correlated strongly

enough to be measuring the same concept, but not so strongly

that each item is measuring the same aspect of the concept

(see Table 1).

Construct validity

To reduce items and estimate the scale’s dimensionality, the

data were exploratory factor analysed so that ‘the remaining

items maximize the variance in the scale and maximize…the

scale’s reliability’ (Netemeyer et al. 2003, p. 121). Kaiser–

Meyer–Okin (0Æ93, P < 0Æ001) and individual measures of

sampling adequacy (>0Æ60) indicated that these data could

be factor analysed. Principal axis factoring (PAF) was chosen

for the factor solution because measurement errors are ex-

cluded, the solution had the best scientific utility, consistency,

and meaning (Tabachnick & Fidell 2007), and it made the

most intuitive and conceptual sense (Pett et al. 2003). PAF

with equamax rotation resulted in item factor loadings from

0Æ32 to 0Æ82 (see Table 4).

In the PAF solution, 15 factors were extracted, with

eigenvalues >1Æ0 (1Æ02–24Æ75), accounting for over half

(52Æ27%) of the cumulative variance (Kaiser’s criterion;

Pallant 2007). Adequacy of extraction and factor retention

criteria included the size of the eigenvalues (>1Æ0) and

residuals, scree plot and factor interpretability and usefulness

(Pett et al. 2003, Tabachnick & Fidell 2007). The scree plot



indicated a 13-factor solution. Because these methods tend to

overstate the number of factors to retain, a third technique,

parallel analysis, was completed to further inform this

decision (Pallant 2007). A software program (MONTEONTE CARLOARLO

PCA for Parallel Analysis; Pallant 2007) was run as per the

instructions, requiring the input of three data elements to

calculate random sets of data: the number of variables,

subjects and replications. Results from the parallel analysis

were systematically compared with the pilot test data

eigenvalues. For example, if the pilot test eigenvalue was

greater than the criterion value from the parallel analysis, the

factor was retained; if it was lower, the factor was rejected

(Pallant 2007). The results from the parallel analysis

supported the previous determination to retain 13 factors.

Rotation, following extraction, is used to improve the

interpretability and scientific utility of the factor solution

(Tabachnick & Fidell 2007). A simple structure with at least

two correlated variables loading on each factor indicates

adequacy of rotation (Tabachnick & Fidell 2007). The goal is

to ‘retain enough factors for an adequate fit, but not so many

that parsimony is lost’ (Tabachnick & Fidell 2007, p. 644).

Factor correlations <0Æ32 indicate <10% overlap in

variance, not enough to warrant an oblique solution (Pett

et al. 2003, Tabachnick & Fidell 2007). Orthogonal equa-

max rotation resulted in 12 meaningful and three useful but

small 3-item factors (Nunnally & Bernstein 1994).

An established criterion does not exist for size of factor

loading, but loadings of 0Æ40 in the extraction phase are

sufficiently strong (Pett et al. 2003). After factor rotation,

loadings of ‡0Æ32 indicate a meaningful correlation (Nun-

nally & Bernstein 1994, Tabachnick & Fidell 2007). Lower

factor loadings (0Æ32–0Æ40) for factor interpretation were

warranted because subscale homogeneity was suspected

(Tabachnick & Fidell 2007).

Generally, items with factor loadings from 0Æ40 to 0Æ90

(Netemeyer et al. 2003) were retained, although not all items

with acceptable loadings were kept. Items were eliminated

based on item analysis and lack of conceptual congruity with

the factor. Even though 3–4 items are sufficient to create a

factor (Comrey 1988, Netemeyer et al. 2003), three 3-item

factors were consolidated to form the competence factor

(a = 0Æ85). This was based on conceptual agreement, item

and parallel analysis and because often the ‘last few factors

represent the most interesting and unexpected findings’

Table 4 Rotated factor loadings for the 82-item Professionalism and Environmental Factors in the Workplace Questionnaire�: principal axis

factoring with equamax rotation*

Factor 1

Commitment

Factor 2

Collaborative

practice

Factor 3

Advocacy

Factor 4

Accountability

Factor 5

Knowledge

Factor 6

Competence�

Factor 7

Professional

support

(99) 0Æ81 (93) 0Æ82 (84) 0Æ77 (74) 0Æ73 (55) 0Æ68 (68) 0Æ78 (22) 0Æ64

(101) �0Æ75 (92) 0Æ80 (86) 0Æ73 (81) 0Æ67 (53) 0Æ62 (66) 0Æ66 (23) 0Æ59

(101) �0Æ75 (91) 0Æ79 (83) 0Æ70 (75) 0Æ66 (57) 0Æ62 (72) 0Æ61 (21) 0Æ59

(98) 0Æ74 (90) 0Æ69 (85) 0Æ51 (73) 0Æ64 (56) 0Æ58 (64) 0Æ67 (19) 0Æ50

(103) �0Æ73 (95) 0Æ64 (87) 0Æ49 (88) 0Æ52 (52) 0Æ50 (63) 0Æ51 (27) 0Æ50

(100) 0Æ70 (94) 0Æ45 (82) 0Æ40 (89) 0Æ42 (65) 0Æ45 (24) 0Æ41

(104) 0Æ61 (77) 0Æ39 (60) 0Æ72 (25) 0Æ41

(105) 0Æ57 (58) 0Æ68

(62) 0Æ37

Factor

8 Shared

governance

Factor 9

Control of

nursing practice

Factor 10

Autonomy

Factor 11

Professionhood

Factor 12

Environment

culture and climate

Factor 13

Quality of

nursing worklife

(34) 0Æ56 (3) 0Æ71 (47) 0Æ58 (71) 0Æ60 (36) 0Æ54 (10) �0Æ73�

(33) 0Æ48 (30) 0Æ57 (45) 0Æ57 (78) 0Æ56 (37) 0Æ49 (13) �0Æ67

(17) 0Æ45 (2) 0Æ55 (49) 0Æ55 (79) 0Æ50 (35) 0Æ48 (1) 0Æ33

(32) 0Æ62 (20) 0Æ49 (46) 0Æ54 (67) 0Æ37 (28) 0Æ46 (16) 0Æ38

(29) 0Æ53 (7) �0Æ39 (48) 0Æ42 (51) 0Æ32 (40) 0Æ37 (11) 0Æ48

(18) 0Æ52 (54) 0Æ32 (4) 0Æ48�

(31) 0Æ38 (12) 0Æ37

(38) 0Æ46

*(item number) factor loading.�Factor comprised of three factors.�Item deleted because of trivial redundancy and Cronbach’s alpha-if-item-deleted.



(Tabachnick & Fidell 2007, p. 646) and ‘should be retained’

(Netemeyer et al. 2003, p. 127). Item loadings ranged from

acceptable (n = 12; 0Æ30–0Æ39), fair (n = 16; 0Æ40–0Æ49), good

(n = 22; 0Æ50–0Æ59), very good (n = 17; 0Æ60–0Æ69) to excel-

lent (n = 14; 0Æ70–0Æ79; Nunnally & Bernstein 1994, Pett

et al. 2003). Three items loaded at >0Æ80 (see Table 4).

Factors correlated with the total scale (0Æ54–0Æ77; see

Figure 1) and with each other (0Æ10–0Æ72), indicating consis-

tency within the scale (see Table 5).

Five items were rescaled and reverse scored in the final

instrument. Item analysis, trivial redundancy and appraisal of

conceptual fit resulted in the deletion of two items, leaving an

82-item instrument. Redundancy is an integral aspect of scale

internal consistency; however, if item wording and sentence

structure are too similar between two items, one of the items

must be deleted (DeVellis 2003). When items load strongly

(i.e. >0Æ40) on multiple factors, the items are placed in the

factor with the highest loading that it is ‘most closely related

to conceptually’ (Pett et al. 2003, p. 173). No specific criteria

were used to appraise conceptual fit. However, calculating

Cronbach’s alpha for each factor to determine the factor’s

internal consistency with and without the multiple-loading

items helps determine where best to place an item. It is also

useful to assess the factors’ inter-item reliability with and

without the item and the item’s role in adding to the

interpretability of the factor (Pett et al. 2003). ‘The goal is to

group items together so that the factor on which they are

placed represents a consistent content area’ (Pett et al. 2003,

p. 196).

Scale length was considered; however, a construct with mul-

tiple dimensions typically requires more items to ‘adequately

tap the domain/dimensions of the construct’ (Netemeyer et al.

2003, pp. 145–146). Squared multiple correlations (SMC) of

factor scores, predicted from scores on observed variables,

are an estimate of the factor solution’s internal consistency

(Tabachnick & Fidell 2007). ‘In a good solution…[SMCs]

range between 0 and 1; the larger the…[SMCs], the more

stable the factors’ (Tabachnick & Fidell 2007, p. 649). SMC

for the first 12 factors ranged from 0Æ71 to 0Æ91, suggesting

consistency among the items in the factors (Pett et al. 2003).

Small residuals (<1% >0Æ05) signified that all factors were

extracted from the data.

Items were factored into factors that were consistent with

the original conceptualization; therefore, factor names were

adopted from the conceptual framework attributes. Data

analysis supported the scale’s multidimensionality and the

subscales’ unidimensionality. Preliminary theory develop-

ment was assisted by the exploratory factor analysis

(Tabachnick & Fidell 2007).

Discussion

Study limitations

When interpreting results, the study’s limitations should be

considered. Convenience sampling affects the generalizability

of the results (Hulley et al. 2001, Loiselle et al. 2004). In

addition, measurement error from various sources, such as

response set bias (Pett et al. 2003) and agreement bias

(Viswanathan 2005), must be considered. Temporary per-

sonal factors of the participants (Loiselle et al. 2004), the

process of measurement inducing changes in the concept, or

Table 5 Factor correlations for the Professionalism and Environmental Factors in the Workplace Questionnaire�*

Factors 1 2 3 4 5 6 7 8 9 10 11 12 13

Environmental factors

1. Control of nursing practice –

2. Quality of nursing worklife 0Æ51 –

3. Professional support 0Æ57 0Æ62 –

4. Shared governance 0Æ56 0Æ72 0Æ68 –

5. Environment culture and climate 0Æ63 0Æ57 0Æ66 0Æ64 –

Professionalism factors

1. Autonomy 0Æ42 0Æ23 0Æ31 0Æ30 0Æ48 –

2. Knowledge 0Æ30 0Æ20 0Æ32 0Æ25 0Æ43 0Æ57 –

3. Competence 0Æ25 0Æ20 0Æ30 0Æ24 0Æ36 0Æ54 0Æ59 –

4. Professionhood 0Æ27 0Æ26 0Æ32 0Æ32 0Æ37 0Æ45 0Æ64 0Æ63 –

5. Accountability 0Æ20 0Æ10 0Æ20 0Æ12 0Æ35 0Æ52 0Æ55 0Æ53 0Æ44 –

6. Advocacy 0Æ20 0Æ10 0Æ20 0Æ20 0Æ32 0Æ51 0Æ52 0Æ60 0Æ52 0Æ69 –

7. Collaborative practice 0Æ24 0Æ30 0Æ26 0Æ35 0Æ36 0Æ38 0Æ30 0Æ34 0Æ28 0Æ34 0Æ38 –

8. Commitment 0Æ36 0Æ39 0Æ32 0Æ41 0Æ38 0Æ31 0Æ22 0Æ30 0Æ32 0Æ20 0Æ20 0Æ21 –

Total 0Æ65 0Æ66 0Æ71 0Æ74 0Æ77 0Æ65 0Æ62 0Æ65 0Æ66 0Æ54 0Æ57 0Æ54 0Æ61

*All correlations statistically significant at P < 0Æ01 (2-tailed).



changes in the underlying theoretical concept itself (Carmines

& Zeller 1979), as evidenced by low internal consistency, can

occur. Defining sample selection criteria controlled for

selection bias, although the voluntary nature of participation

can threaten the study’s internal validity (LoBiondo-Wood

et al. 2005). Respondent burden was not identified as an issue

by participants.

The PEFWQ�

The empirical study of professionalism, still in its early stages,

lacks a tested theoretical model; thus, an exploratory survey

was used to construct a ‘picture of a phenomenon’ (LoBi-

ondo-Wood et al. 2005, p. 266). The hypothesis that

attributes from the literature, which became the subscales,

would explain the construct was supported. Exploratory

factor analysis, conducted for instrument and theory devel-

opment (LoBiondo-Wood et al. 2005), grouped individual

items into meaningful subscales reflecting the attributes of the

literature review.

The preliminary results of psychometric testing support

the validity and reliability of the PEFWQ�. The 13-factor

solution resulted in stable factors while retaining parsi-

mony, fit the sample data adequately, and mirrored the

instrument’s conceptual basis. Construct validity is an

ongoing process. The application and testing of the instru-

ment with different populations and settings (Netemeyer

et al. 2003) using confirmatory factor analysis, is planned

for the future.

The PEFWQ� was designed to determine nurses’ key

environmental and professionalism influences affecting their

professionalism in practice. It is meant to help practitioners

reflect on their own practice situation and will be useful to

link the practice environment with professionalism in

practice for managers and educators. Using the instrument

as a guide, discussion of issues affecting work environments

and professionalism will lead to dialogue, planning and

implementation of methods to support professionalism in

practice and healthy work environments. The use of mea-

surement is a unique approach to set the stage for discussions

about professionalism in the work environment. The ques-

tionnaire is also of interest to organizations that have

regulatory and/or professionalism mandates for their guiding

principles.

Conclusion

Exploratory factor analysis supported the conceptual frame-

work upon which this complex multidimensional scale was

based. The conceptual framework proved to be a useful guide

for item generation and subscale development. The PEFWQ�

stimulates awareness and discussion of professionalism and

its relationship to the practice environment. It assists the

individual practitioner to reflect on their practice situation.

Better understanding of how environmental and profession-

alism factors affect practice will help nurses maintain

professionalism in tumultuous healthcare environments.

The results of the questionnaire summarize nurses’ ratings

of professionalism and environmental attributes that play a

role in their daily work lives and facilitate mutual problem-

solving. The instrument offers a mechanism for individual

reflection, collective dialogue and action planning that

What is already known about this topic

• The literature consists of mainly descriptive research,

opinion and theoretical papers regarding aspects of

professionalism.

• No consensus exists in the literature on the meaning of

professionalism.

• Existing instruments encompass aspects of profession-

alism and the environment but do not measure aspects

of individual professionalism attributes, thus establish-

ing the need for this study.

What this paper adds

• The questionnaire facilitates reflection on practice, col-

lective dialogue and action planning that contributes to

quality improvement initiatives to enhance profession-

alism in practice and client care.

• Completion of the questionnaire provides a starting

point for discussion, planning and implementation of

methods to support healthy work environments.

• Better understanding of how environmental and pro-

fessionalism factors affect practice will help nurses

maintain professionalism in tumultuous healthcare

environments.

Implications for practice and/or policy

• The questionnaire assists the individual practitioner to

reflect on their own practice situation.

• The questionnaire will be useful to link the practice

environment with professionalism in practice for

administrators and educators.

• Organizations with regulatory and/or professional

practice mandates will find the questionnaire useful for

guiding principles.



contributes to quality improvement initiatives aimed at

enhancing client care and the practice environment.

Acknowledgements

The authors would like to thank the College of Nurses of

Ontario, specifically Heather Campbell, Director of Practice

and Policy. The authors would also like to thank Janet

Anderson, Manager of Practice; Shandelle Johnson, Practice

Consultant; and Carolina Law, Data Administrator for their

ongoing input and support.

Funding

This research received no specific grant from any funding

agency in the public, commercial, or not-for-profit sectors.

Conflict of interest

No conflict of interest has been declared by the authors.

Author contributions

AB was responsible for the study conception and design. CK

performed the data collection and provided statistical exper-

tise. AB and CK performed the data analysis and responsible

for the drafting of the manuscript. AB made critical revisions

to the paper for important intellectual content; provided

administrative, technical or material support; and supervised

the study.

References

Aiken L.H. & Patrician P. (2000) Measuring organizational traits of

hospitals: the revised nursing work index. Nursing Research 29(3),

146–153.

Aiken L.H., Sochalski J. & Lake E.T. (1997) Studying outcomes of

organizational change in health services. Medical Care 35(Suppl.

11), NS6–NS18.

Altun I. & Ersoy N. (2003) Undertaking the role of patient advocate:

a longitudinal study of nursing students. Nursing Ethics: An

International Journal for Health Care Professionals 10(5), 462–

471.

Anthony M.K., Brennan P.F., O’Brien R. & Suwannaroop N. (2004)

Measurement of nursing practice models using multiattribute

utility theory: relationship to patient and organizational outcomes.

Quality Management in Health Care 13(1), 40.

Ballou K.A. (1998) A concept analysis of autonomy. Journal of

Professional Nursing 14(2), 102–110.

Batey M. & Lewis F. (1982) Clarifying autonomy and accountability

in nursing service: part 1. The Journal of Nursing Administration

12(9), 13–18.

Baumann A., O’Brien-Pallas L., Armstrong-Stassen M., Blythe J.,

Bourbonnais R., Cameron S., Irvine Doran D., Kerr M., McGillis

Hall L., Vezina M., Butt M. & Ryan L. (2001) Commitment and

Care – The Benefits of a Healthy Workplace for Nurses, Their

Patients and the System. Canadian Health Services Research

Foundation and The Change Foundation, Ottawa, Ontario.

Blau G. (1985) The measurement and prediction of career commit-

ment. Journal of Occupational Psychology 58, 277–288.

Briggs S.R. & Cheek J.M. (1988) On the nature of self-monitoring:

problems with assessment, problems with validity. Journal of

Personality and Social Psychology 54(4), 663–678.

Carmines E.G. & Zeller R.A. (1979) Reliability and Validity

Assessment, Quantitative Applications in the Social Sciences Series,

No. 17. Sage, Newbury Park, CA.

Clark L.A. & Watson D. (1995) Construct validity: basic issues in

objective scale development. Psychological Assessment 7(3), 309–

319.

Cohen M.Z., Hausner J. & Johnson M. (1994) Knowledge and

presence: accountability as described by nurses and surgical pa-

tients. Journal of Professional Nursing 10(3), 177–185.

Cole J.C. (2008) How to deal with missing data: conceptual overview

and details for implementing two modern methods. In Best Prac-

tices in Quantitative Methods (Osborne J.W., ed.), Sage, Los

Angeles, CA, pp. 214–238.

College of Nurses of Ontario (2006) Annual Membership Renewal

Form. Author, Toronto, Ontario.

Comrey A.L. (1988) Factor-analytic methods of scale development in

personality and clinical psychology. Journal of Consulting and

Clinical Psychology 56(5), 754–761.

Davidson H., Folcarelli P., Crawford S., Duprat L.J. & Clifford J.

(1997) The effects of health care reforms on job satisfaction and

voluntary turnover among hospital-based nurses. Medical Care

35(6), 634–645.

DeVellis R.F. (2003) Scale Development: Theory and Application,

2nd edn. Sage, Thousand Oaks, CA.

Dillman D.A. (2007) Mail and Internet Surveys: The Tailored Design

Method: 2007 update With New Internet, Visual, and Mixed-

Mode Guide, 2nd edn. Wiley, New York.

Dillman D.A., Smyth J.D. & Christian L.M. (2009) Internet, Mail,

and Mixed-Mode Surveys: The Tailored Design Method, 3rd edn.

Wiley, New York.

Dwyer D.J., Schwartz R.H. & Fox M.L. (1992) Decision-making

autonomy in nursing. Journal of Nursing Administration 22(2),

17–23.

Field A. (2005) Discovering Statistics Using SPSS, 2nd edn. Sage,

London.

Flexner A. (1915) Is social work a profession? Research on Social

Work Practice 1(26), 901.

Grindel C.G., Peterson K., Kinneman M. & Turner T.L. (1996) The

practice environment project: a process for outcome evaluation.

Journal of Nursing Administration 26(5), 43–51.

Havens D.S. & Aiken L.H. (1999) Shaping systems to promote de-

sired outcomes: the magnet hospital model. Journal of Nursing

Administration 29(2), 14–20.

Hulley S., Cummings S., Vrowner W., Grady D., Hearst N. &

Newman T. (2001) Designing Clinical Research: An Epidemio-

logic Approach, 2nd edn. Lippincott, Philadelphia.



Karasek R.A. (1979) Job demands, job decision latitude and mental

strain: implications for job redesign. Administrative Science

Quarterly 24, 285–306.

Kramer M. & Schmalenberg C.E. (2003) Magnet hospital staff

nurses describe clinical autonomy. Nursing Outlook 51(1), 13–19.

Laschinger H.K.S. & Wong C. (1999) Staff nurse empowerment and

collective accountability: effect on perceived productivity and self-

rated work effectiveness. Nursing Economics 17(6), 308–316.

Leddy S. & Pepper J.M. (eds) (1985) Socialization for Professional

Practice. Conceptual Basis of Professional Nursing. Lippincott,

Philadelphia, pp. 33–52.

Litwin M.S. (1995) How to Measure Survey Reliability and Validity,

Survey Kit, Vol. 7. Sage, Thousand Oaks, CA.

LoBiondo-Wood G., Haber J., Cameron C. & Singh M.D. (eds) (2005)

Nursing Research in Canada: Methods, Critical Appraisal, and

Utilization, 1st Canadian edn. Elsevier Mosby, Toronto, Ontario.

Loevinger J. (1954) The attenuation paradox in test theory. Psy-

chological Bulletin 51(5), 493–504.

Loiselle C.G., Profetto-McGrath J., Polit D.F. & Beck C.T. (2004)

Canadian Essentials of Nursing Research. Lippincott, Philadelphia.

Lynn M.R. (1986) Determination and quantification of content

validity. Nursing Research 35(6), 382–386.

Mark B.A., Salyer J. & Wan T.T.H. (2003) Professional nursing

practice: impact on organizational and patient outcomes. Journal

of Nursing Administration 33(4), 224–234.

Munro B.H. (2005) Statistical Methods for Health Care Research,

5th edn. Lippincott, Philadelphia.

Netemeyer R.G., Beardon W.O. & Sharma S. (2003) Scaling Pro-

cedures: Issues and Applications. Sage, Thousand Oaks, CA.

Nunnally J.C. & Bernstein I.H. (1994) Psychometric Theory, 3rd

edn. McGraw-Hill, New York.

Pallant J. (2007) SPSS Survival Manual: A Step-By-Step Guide to

Data Analysis Using SPSS Version 15, 3rd edn. Open University

Press, McGraw-Hill, New York.

Peterson N.J. (2008) Designing a rigorous small study sample. In Best

Practices in Quantitative Methods (Osborne J.W., ed.), Sage, Los

Angeles, pp. 214–238.

Pett M.A. (1997) Nonparametric Statistics for Health Care Research:

Statistics for Small Samples and Unusual Distributions. Sage,

Thousand Oaks, CA.

Pett M.A., Lackey N.R. & Sullivan J.J. (2003) Making Sense of

Factor Analysis: The Use of Factor Analysis for Instrument

Development in Health Care Research. Sage, Thousand Oaks, CA.

Rafferty A.M., Ball J. & Aiken L.H. (2001) Are teamwork and

professional autonomy compatible, and do they result in improved

hospital care? Quality in Health Care 10(Suppl. 2), ii32–ii37.

Registered Nurses Association of Ontario (2007) Healthy Work

Environments Best Practice Guidelines: Professionalism in Nurs-

ing. Author, Toronto, Ontario.

Salkind N.J. (ed.) (2007) Encyclopedia of Measurement and Statis-

tics. Sage, Thousand Oaks, CA.

Slocumb E.M. & Cole F.L. (1991) A practical approach to content

validation. Applied Nursing Research 4(4), 192–200.

Spreitzer G. (1995) Psychological empowerment in the workplace:

dimensions, measurement, and validation. Academy of Manage-

ment Journal 33(5), 1442–1465.

SPSS Inc. (2006) SPSS for windows 15.0� (C006Fmputer software).

Author, Chicago.

Stewart D.W. & Shamdasani P.N. (1990) Focus Groups: Theory and

Practice. Sage, London.

Streiner D.L. & Norman G.R. (2003) Health Measurement Scales: A

Practical Guide to Their Development and Use, 3rd edn. Oxford

University Press, Oxford.

Styles M.M. (1982) On Nursing: Towards a New Endowment.

Mosby, St Louis, MO.

Tabachnick B.G. & Fidell L.S. (2007) Using Multivariate Statistics,

5th edn. Pearson, Boston.

Trofino J. (1996) Developing nursing vision: a professional model for

nursing practice. Nursing Standard 10(15), 31–35.

Viswanathan M. (2005) Measurement Error and Research Design.

Sage, Thousand Oaks, CA.

Vogt W.P. (2005) Dictionary of Statistics and Methodology: A

Nontechnical Guide for the Social Sciences, 3rd edn. Sage, Thou-

sand Oaks, CA.

Waltz C.F., Strickland O.L. & Lenz E.R. (2005) Measurement in

Nursing and Health Research, 3rd edn. Springer, New York.

Weiss S. (1983) Role differentiation between nurse and physician:

implications for nursing. Nursing Research 32(3), 133–139.

Weiss S.J. & Davis H.P. (1985) Validity and reliability of the col-

laborative practice scales. Nursing Research 34(5), 299–305.

The Journal of Advanced Nursing (JAN) is an international, peer-reviewed, scientific journal. JAN contributes to the

advancement of evidence-based nursing, midwifery and health care by disseminating high quality research and

scholarship of contemporary relevance and with potential to advance knowledge for practice, education, management

or policy. JAN publishes research reviews, original research reports and methodological and theoretical papers.

For further information, please visit the journal web-site: http://www.journalofadvancednursing.com



The Professionalism and Environmental Factors in the Workplace Questionnaire ® : development and psychometric evaluation

Documents