86
STRAIGHTFORWARD INCIVILITY SCALE: ADAPTATION AND
VALIDATION OF WORKPLACE INCIVILITY MEASUREMENT FOR
PORTUGUESE SAMPLES
Tito Laneiro1, Luisa Ribeiro2, Martina Nitzsche3, Tânia Ferraro4, Genta Kulari5, Michael Leiter6
PSIQUE • EISSN 21834806 • VOLUME XIX • ISSUE FASCÍCULO 2
1ST JULY JULHO  31ST DECEMBER DEZEMBRO 2023 PP. 86105
DOI: https://doi.org/10.26619/2183-4806.XIX.2.5
Submited on 05/01/2023 Submetido a 05/01/2023
Acceptted on 27/06/2023 Aceite a 27/06/2023
Abstract
The concept of workplace incivility is an underestimated subject in Portugal but a popular
one in the international literature. Workplace incivility does not intent to harm others, but it
harms workplace norms and put peaceful workplace environment into danger. Thus, the main
purpose of this study was to present the adaptation and validation of the Straightforward
Incivility Scale (SIS; Leiter & Day, 2013) for Portuguese healthcare professionals’ samples. SIS has
25 items that cover five different sources of Workplace Incivility (WI): supervisors, colleagues,
subordinates, customers or the participant her/himself. A Portuguese version of the scale was
administered to a total of 737 healthcare professionals from two major public hospital units from
the metropolitan area in Lisbon, Portugal (78% women, 83% nurses, 56% with ages from 25 to
34). To assess the factor structure, we submitted these samples to exploratory and confirmatory
factor analyses. The results provided psychometric support for the new Portuguese measurement
(SIS). Furthermore, it showed good reliability and convergent validity indices with burnout.
Considering the mainstream of studies in the healthcare sector, this study adds to the incivility
literature as a novel area of research. Furthermore, the study provides a validated version of
Straightforward Incivility Scale allowing simultaneous registration of five different workplace
incivility sources, while also providing a measurement with good psychometric properties. It is
our hope that the workplace incivility can be the focus of future studies measuring its outcomes
1 Centro de Investigação em Psicologia, Universidade Autónoma de Lisboa, tlaneiro@autonoma.pt
2 Centro de Investigação em Psicologia, Universidade Autónoma de Lisboa, mribeiro@autonoma.pt
3 Centro de Investigação em Psicologia, Universidade Autónoma de Lisboa, martina.nitzsche@hotmail.com
4 Centro de Investigação em Psicologia, Universidade Autónoma de Lisboa, taniaferraro@upt.pt
5 Centro de Investigação em Psicologia, Universidade Autónoma de Lisboa, gkulari@autonoma.pt
6 Deakin University, michael.leiter@deakin.edu.au
Corresponding author: Luisa Ribeiro, Centro de Investigação em Psicologia, Universidade Autónoma de Lisboa, mribeiro@
autonoma.pt
87
PSIQUE • e-ISSN 2183-4806 • Volume XIX • Issue Fascículo 2 • 1st july julho-31st december dezembro 2023 pp. 86-105
Tito Laneiro, Luisa Ribeiro, Martina Nitzsche, Tânia Ferraro, Genta Kulari, Michael Leiter
among healthcare employees as well as to hospital managers and human resources raising
awareness within the context of healthcare sector.
Keywords: Workplace Incivility, Straightforward Incivility Scale, measurement validation, factor struc-
ture, psychometrics, burnout, health care professionals.
1. Introduction
Workplace incivility (WI) has increased in frequency and severity in the healthcare
environment in the last decade (Cortina et al., 2001; Pearson & Porath, 2009; Tricahyadinata et
al., 2020). There are numerous circumstantial reports of uncivil behaviour in healthcare settings,
although few empirical studies exist in the literature (Laschinger et al., 2009). WI is a subtle form
of workplace violence defined aslow-intensity deviant behaviour with ambiguous intent to
harm the target, in violation of workplace norms for mutual respect” (Andersson & Pearson, 1999,
p. 475). Uncivil behaviours include rude and discourteous comments and actions and generally
displaying a lack of concern for others. Pearson and Porath (2005) found that employees who
experienced uncivil behaviours at work intentionally reduced their work efforts and the quality
of their work. Furthermore, Cortina et al. (2001) linked workplace incivility to job dissatisfaction
and burnout. In particular, in healthcare sector, the prevalence of burnout is the highest when
compared with other professions (Greenglass, Burke, & Fiksenbaum, 2001). Similarly, more
recent research supported the previous findings, linking uncivil behaviours to burnout among
healthcare professionals (Laschinger et al., 2008; Read & Laschinger, 2013; Laschinger, 2012).
However, there is a dearth of research in incivility in healthcare sector in Portugal, which may be
due to the lack of validated versions of well-established measures. Furthermore, the majority of
the international research are focused, with few exceptions, on nurses, providing no perspective
on a broad occupational group of healthcare setting.
The present study, underpins the theory of Andersson and Pearson (1999) regarding the
concept of Workplace Incivility. The authors referred to various counterproductive behaviours in
the workplace such as lack of respect, courtesy or politeness toward co-workers. However, little
is known about the intention of such manifestations, consequently researchers have provided
different interpretations based on their sample and context (Andersson & Pearson, 1999; Cortina
et al., 2001; Pearson et al., 2000). Similarly, Leiter (2013) proposed two dynamics that contribute
to the emotional response of a negative social behaviour. One reason lies behind the intention the
instigator has to exclude the recipient from his or her social group. The other argument highlights
the relation of incivility to risk, as instigators of incivility become targets of hierarchys own
incivility, thus affecting their decision-making, and potentially putting their careers in jeopardy
(Irum et al., 2020; Itzkovich, 2016; Jiménez, 2018). Based on a recent systematic review, incivility
is still a novel area of research, especially in healthcare sector encompassing the perspective of
a broader occupational professions (Vazconcelos, 2020).
The original Straightforward Incivility Scale (Leiter & Day, 2013) was designed to measure
five sources of uncivil behaviours, encompassing supervisors, co-workers, subordinates,
clients, and the respondent itself. The 25-item scale provides evidence on the frequency uncivil
behaviours manifested in the last month. Meanwhile, studies involving healthcare professionals
88
PSIQUE • e-ISSN 2183-4806 • Volume XIX • Issue Fascículo 2 • 1st july julho-31st december dezembro 2023 pp. 86-105
Straightforward Incivility Scale: adaptation and validation of Workplace Incivility measurement for Portuguese samples
have reduced the scale of SIS into 15 items. The work of Portoghese et al. (2015) accounted for
incivility among supervisors, co-workers, and respondents, while Fida et al. (2018) counted only
supervisors, co-workers, and physicians in their study involving Canadian nurses.
However, despite the broad and diverse body of work on incivility, to date it is difficult
for scholars and practitioners alike to integrate and understand the variety of findings on this
negative workplace behaviour (Schilpzand et al., 2016). Moreover, without a clear understanding
of the extant work, practitioners may not be able to incorporate the accumulated knowledge
in their organizational practices (Vazconcelos, 2020). However, based on the original theory,
considering that uncivil behaviour can lead to a spiral of violence, the present study is the first
one to include all five potential sources of incivility as originally anticipated by Leiter and Day
(2013). The scale offers an integrated measurement when applied either in a short term or in a
longitudinal study, which allows the monitoring of variations in uncivil behaviour over time, as
well as considering various sources (including the respondent).
Thus, the goal of this study was to standardize a European Portuguese measure of incivility.
To do so, we aimed to examine the validity and reliability of a Portuguese version of SIS in a
representative healthcare employees population by testing the internal consistency, the
convergent and discriminant validity, and the factorial structure of the scale. To date this is
the first study to integrate the perspective of various occupational groups in healthcare sector
measuring incivility. Considering the mainstream of studies in healthcare sector, this research
adds to the general literature of incivility. Furthermore, there is a dearth of literature studying
incivility in Portugal, thus the present study can provide a new instrument for future research to
measure the outcomes of incivility, in particular among healthcare providers.
The proposed instrument will allow organizations to assess incivility levels suffered by their
staff, informing them about the actual organizational culture. Incivility is a variable related to
poor performance and errors, which in a healthcare setting are particularly dangerous. At the
same time, it is internally controllable and, therefore, can be easily changed than other variables,
which depend on market demands, national policies and budget restrictions. It will distinguish
between different origins of incivility, allowing a targeted intervention. It will further deepen
understanding between incivility (and its different origins) and other variables pertinent to
personal and organizational health and performance.
2. Relevant literature
If anyone has doubts about whether workplace incivility deserves to be concerned, the
research suggests that workplace incivility may act as a precursor to other forms of workplace
violence (Andersson & Pearson, 1999). Incivility has been recognized as being “one of the most
pervasive forms of antisocial behavior in the workplace” (Cortina, 2008, p.56). Furthermore,
the main factor making workplace incivility very crucial in todays work life is the so-called
“incivility spiral” (Andersson & Pearson, 1999). Uncivil acts of one employee may create a
snowball effect and can influence the whole organization because the targets of incivility can
easily transform into instigators.
Among several attempts to characterize workplace incivility, Blau and Andersson (2005)
described it as a type of social interaction, while others even as an interpersonal mistreatment
89
PSIQUE • e-ISSN 2183-4806 • Volume XIX • Issue Fascículo 2 • 1st july julho-31st december dezembro 2023 pp. 86-105
Tito Laneiro, Luisa Ribeiro, Martina Nitzsche, Tânia Ferraro, Genta Kulari, Michael Leiter
(Cortina & Magley, 2003) that jeopardizes relationships not only from a personal but also
organizational and management level as well (Cortina & Magley, 2003; Estes & Wang, 2008;
Lewis & Malecha, 2011; Pearson et al., 2000; Torkelson et al., 2016). McNamara (2012) described
incivility among healthcare professionals in particular as related to status or position of power.
She observed that it is commonplace between co-workers displaying an asymmetry of power in
this context.
This position is further solidified in studies with recently graduated nurses that found them
more vulnerable to workplace incivility as a result of their limited experience and their new
status in the work setting (Laschinger, 2014; Laschinger et al., 2013; Read & Laschinger, 2013;
Smith et al., 2010; Wing et al., 2015). Hence it is of high interest to understand the consequences
of workplace incivility in nurses’ professional and personal life such as increased work stress,
absenteeism, frustration and decreasing satisfaction, which can lead to burnout (Tastan &
Davoudi, 2015; Welbourn et al., 2020; Wing et al. 2015).
In continuity, Laschinger (2014) found that different types of workplace mistreatment,
bullying and incivility in particular, have unfavourable consequences for nurses’ perceptions
of care quality and patient safety risks. Among healthcare professionals, nurses are most prone
to be the victim of uncivil behaviours (Ten Hoeve et al., 2019), especially in cases of physical
and verbal violence by patients or their companions (Speroni et al., 2014). A majority of studies
made use of the first Workplace Incivility Scale developed by Cortina et al. (2001), evidencing
the hostility between supervisor and co-workers (Portoghese et al., 2015). However, the scale
has a uni-dimensional structure that does not differentiate the uncivil behaviours coming from
different elements of the organization.
Leiter (2013) later developed a new measure of workplace incivility, the Straightforward
Incivility Scale (SIS), allows researchers to differentiate the uncivil behaviours, as identified by
the original theory. The scale is a self-reported measure aiming to evidence the nature of incivility
in the workplace, which has been vastly adapted in different languages and cultures (Matthews
et al. 2016; Portoghese et al. 2015; Smidt et al., 2016; Tsuno et al., 2017). The current study provides
support for adaptation and validation of the workplace incivility measure, Straightforward
Incivility Scale (Leiter & Day, 2013) for Portuguese samples as well. The final discussion points out
the main advantages of using the validated instrument and the contributions to the monitoring of
civility-promoting actions. The implications for practice and new directions for future research
are also discussed.
2.1. The outcomes of incivility
Being the target of workplace incivility evokes negative emotions like anger, fear, and
sadness (Porath & Pearson, 2005). Supporting such arguments, there are several studies showing
that uncivil work-place experiences are associated with negative work outcomes, like reduced
job satisfaction, increased job withdrawal, negative mood, and cognitive distraction (Lim et al.,
2008; Miler et al., 2012). Additionally, experiencing incivility is associated with lower energy
levels, higher levels of negative effect, lower task performance, etc. (Giumetti & Harfield, 2013).
Moreover, a number of studies have found that dissatisfaction with the job predicts various job
withdrawal behaviours, including turnover and retirement (Vasconcelos, 2020). Furthermore,
workplace incivility significantly reduces the quality of supervisor and co-worker relationship,
90
PSIQUE • e-ISSN 2183-4806 • Volume XIX • Issue Fascículo 2 • 1st july julho-31st december dezembro 2023 pp. 86-105
Straightforward Incivility Scale: adaptation and validation of Workplace Incivility measurement for Portuguese samples
which negatively impacts job satisfaction, producing long-term effects on mental and physical
health conditions (Carter & Loh, 2017; Lim et al., 2008). As a consequence, organizations suffer
the costs of workplace incivility through job turnover, as well as decreased productivity and
work performance (Pearson & Porath, 2005; Wang & Chen, 2020).
One of the most important things of incivility is the fact that even witnessing incivility has
negative impact on employees. Witnesses of incivility perform less well on complex and creative
tasks. They are also less likely to be helpful, exhibit citizenship behaviours and more likely to
engage in dysfunctional ideation (Pearson & Porath, 2005). Thus, researchers working on work-
pace incivility mostly use Leiter & Day (2013) SIS. This 5-dimensional scale allows researchers
to differentiate the uncivil behaviours coming from different elements of the organization such
as supervisor and co-worker which may have different behavioural forms. Considering the vast
repercussion of incivility in organizations, we believe that would be fundamental for future
studies to measure the manifestation of uncivil behaviours, among healthcare employees in
Portugal. The literature can be of high interest to organizational managers and human resources
to embrace a new strategy that can keep uncivil behaviours at bay.
3. Method
3.1. Translation
The first step, consistent with previous literature (Behling & Law, 2000; Brislin, 1970; Werner
& Campbell, 1970), included translating the Straightforward Incivility Scale by Leiter and Day
(2013). A group of three bilingual research experts in the field were selected to translate the
scale from English to Portuguese. Three versions of translations were obtained and compared
to keep the original meaning of the scale and adapt to the Portuguese context. Differences were
resolved by experts’ discussion, reaching unanimity regarding the best proposal. After reaching
a final version, three other bilingual research experts in the field were invited to provide the
back-translation from Portuguese to English. After concluding this phase, a seventh English
native speaker was introduced to a phase of pre-test. Further, the scale was administered
among 30 nurses in order to evaluate the items’ comprehension for the final version and assess
its psychometric properties. The scale suffered minor changes after obtaining these results,
taking into account suggestions provided by the nurses. The final Portuguese version of the
Straightforward Incivility Scale kept 25 items, same as the original scale (Leiter & Day, 2013). The
instructions, format of items and response options were maintained (Appendix A).
3.2. Study design and data collection
A cross-sectional quantitative study design was conducted for the study. A convenience
sample with a total of 737 healthcare employees participated in the study, where 310 participants
came from Hospital A and 427 from Hospital B. Following ethical approval from the Ethical
Commission for the Psychology Research Center at the Universidade Autónoma de Lisboa, a
request letter to approach health employees was sent to the administration office of two hospitals
91
PSIQUE • e-ISSN 2183-4806 • Volume XIX • Issue Fascículo 2 • 1st july julho-31st december dezembro 2023 pp. 86-105
Tito Laneiro, Luisa Ribeiro, Martina Nitzsche, Tânia Ferraro, Genta Kulari, Michael Leiter
in the great metropolitan area of Lisbon. For anonymity reasons we will refer to these units as
Hospital A and Hospital B. Sampaio (2021) found that registered healthcare employees in Portugal
suffered from higher rates of burnout and turnover rates, when considering that shortage of staff
and overload is more resistant in major hospitals.
After receiving the confirmation from the hospital administrations, all employees received
an introductory e-mail, explaining the study. Subsequently, three authors of the study introduced
themselves to hospital employees, visiting all teams and inviting them to participate. In Hospital
A, data was gathered during December 2015, and in Hospital B from March to April 2016. Although
some time has passed since data collection, and the conditions at Hospitals might have changed,
we retained the data for instrument validation purposes. This first Portuguese validation will
benefit from further studies to verify these results. However, the present study does not use the
sample to measure current incivility levels, nor to explore recent relationships between variables.
Regardless of current levels, the structure of the concept should remain the same across time.
The surveys were self-administered using printed copies of the questionnaire in an enclosed
envelope at their workplace after obtaining formal consent, introduced by the research team.
In Hospital A questionnaires were gathered in a ballot box until recovery by the research team.
In Hospital B questionnaires were offered during team meetings, and immediately gathered by
research team members. Inclusion criteria were adult (+18 years) healthcare employees registered
at the Hospital A and B at the moment of the data collection. Exclusion criteria included intern
health employees, and employees who had ties with either the research project or research team.
3.3. Respondents profile
The study population from the SIS validation consisted from a total of 737 participants from
healthcare employees registered in two major public hospitals in Lisbon, which in both samples,
the majority were women (respectively 76.4% and 78.4%) in the 25-34 age group (52.3%, 59.4%)
with a college degree (66.9%, 85.1%). The majority of healthcare professionals were nurses (59%,
96.5%) with 6-10 years of working experience in the units (34.4%, 32.8%). Table 1 introduces
further sample composition in terms of professional groups.
TABLE1
Sociodemographic Information (N = 737)
Sociodemographic information Study 1
n = 310
Study 2
n = 427
Gender n (%) n (%)
Men 72 (23.6) 90 (21.6)
Women 233 (76.4) 327 (78.4)
Missing values 5 10
Age (years)
≤ 24 16 (8.2) 31 (14.3)
25 – 34 102 (52.3) 129 (59.4)
35 – 44 36 (18.5) 33 (15.2)
92
PSIQUE • e-ISSN 2183-4806 • Volume XIX • Issue Fascículo 2 • 1st july julho-31st december dezembro 2023 pp. 86-105
Straightforward Incivility Scale: adaptation and validation of Workplace Incivility measurement for Portuguese samples
45 – 54 29 (14.9) 20 (9.2)
55 – 64 12 (6.2) 4 (1.8)
Missing values 115 210
Civil status
Single 134 (43.9) 200 (47.8)
Married / non-marital partnership 151 (49.5) 199 (47.6)
Divorced / Separated / Widowed 20 (6.6) 19 (4.6)
Missing values 5 9
Number of children 1.1 (1.2) ; Min: 0 – Max: 7 0.7 (0.9); Min: 0 – Max: 3
Missing values 124
Professional Category Nurses: 177 (59.0) Nurses: 412 (96.5)
Physicians: 17 (5.7) -
Assistents:84 (28.0) -
Technician: 22 (7.3)
Missing values 10 15
Educational Level
PhD - 1 (0.2)
Master degree 23 (8.0) 40 (9.6)
College degree 192 (66.9) 353 (85.1)
Bachelor or equivalent 3 (1.0) 13 (3.1)
General nursing course 3 (1.0) 8 (1.9)
12th Grade / Vocational course 66 (23.0) -
Compulsory education - -
Missing values 23 12
Tenure
< 6 months 10 (3.8) 2 (0.5)
≥ 6 months < 2 years 20 (7.6) 36 (8.9)
2 – 5 years 65 (24.8) 84 (20.8)
6 – 10 years 90 (34.4) 132 (32.8)
11 – 15 years 32 (12.2) 69 (17.1)
16 – 20 years 14 (5.3) 41 (10.2)
21 – 30 years 21 (8.0) 32 (7.9)
> 30 years 10 (3.8) 7 (1.7)
Missing values 48 24
3.4. Instruments
3.4.1. Straightforward Incivility Scale (SIS)
To measure workplace incivility, the Straightforward Incivility Scale (SIS) was administered
(Leiter & Day, 2013). The scale has a total of 25 items on a seven-point Likert scale ranging from
0 (never) to 6 (more than once per day). The scale measures the frequency uncivil behaviour is
manifested in the workplace referring to the last month among the five sources of incivility.
Each source is comprised of the same five items (such as quantifying how often a supervisor
93
PSIQUE • e-ISSN 2183-4806 • Volume XIX • Issue Fascículo 2 • 1st july julho-31st december dezembro 2023 pp. 86-105
Tito Laneiro, Luisa Ribeiro, Martina Nitzsche, Tânia Ferraro, Genta Kulari, Michael Leiter
“Spoke rudely to you”). The overall alpha Cronbach of incivility in sample 1 (EFA) was .96, while
in sample 2 (CFA) .94. Table 3 displays the Cronbach alpha for each of the workplace incivility
dimensions.
3.4.2. Maslach Burnout Inventory – General Survey
Since strong correlation have been found between incivility and burnout, it is relevant to
test for convergent validity, the Maslach Burnout Inventory (MBI-GS; Maslach & Jackson, 1981).
Burnout is defined as “a prolonged response to chronic emotional and interpersonal stressors on
the job” [Maslach et al., (2001), p.397]. The MBI-GS consists of 16 items in a seven-point Likert scale
ranging from 0 (never) to 6 (every day), distributed along three dimensions: emotional exhaustion
(five items), cynicism (five items) and professional efficacy (six items). The concept of burnout
comprises higher scores on exhaustion (e.g., “O meu trabalho deixa-me exausto/a”) and cynicism
(e.g. “Eu duvido do significado do meu trabalho”) and lower scores on efficacy (e.g. “Na minha
opinião eu sou bom naquilo que faço”). The inventory does not provide a score for burnout, but a
medium score of each of the dimensions. In the present study the validated MBI-GS Portuguese
version (available from Mindgarden.com) was used. The overall burnout Cronbach alpha of the
study was .81. Alpha coefficients for the MBI dimensions are emotional exhaustion .90; cynicism
.83; and professional efficacy .76.
3.5. Data analysis
Meyers et al. [(2006), pp.467-468] recommend performing factorial analysis of an inventory
of 25 items with no fewer than 250 participants. Tabachnick and Fidell (2007) highlighted that
as a general rule of thumb, it is comforting to have at least 300 cases for factor analysis” (p.
613). We presented the Exploratory Factor Analysis (EFA) with n = 310 (sample 1) and data were
collected from healthcare professionals from Hospital A. We performed a Confirmatory Factor
Analysis (CFA), with n = 427 health care professionals from several specialized areas in Hospital
B (sample2).
3.5.1. Exploratory Factor Analysis (EFA)
To evaluate the factorial structure of the Portuguese translation of the Straightforward
Incivility Scale, EFA was applied, following the procedure of Hair et al. (2006) and Tabachnick
and Fidell (2007). For this procedure, we worked with 310 healthcare professionals (sample 1).
To determine the suitability of data for factor analysis, the Kaiser–MeyerOlkin (KMO) test was
used. Homoscedasticity of our samples was verified through the Bartlett’s test. The parallel
analysis method was used, following the procedure described by O’Connor (2000). To confirm
the number of components to extract, the scree plot technique (Cattell & Vogelmann, 1977),
based on eigenvalues, known to be one of the most accurate (Zwick & Velicer, 1986) was used.
Communalities were analysed and the factors were obliquely rotated using Promax (Tabachnick
& Fidell, 2007). Promax has been shown to perform either as satisfactory a Varimax, or “much
better” (e.g. at high correlation factors; Finch, 2006). We accepted factor loadings equal to or
greater than .40 as sufficient (Hair et al., 2006). The correspondence between the subscale scores
was assessed applying zero-order correlation (Pearson).
94
PSIQUE • e-ISSN 2183-4806 • Volume XIX • Issue Fascículo 2 • 1st july julho-31st december dezembro 2023 pp. 86-105
Straightforward Incivility Scale: adaptation and validation of Workplace Incivility measurement for Portuguese samples
3.5.2. Confirmatory Factor Analysis (CFA)
CFA of the 25 items was performed with our second sample preponderantly composed of
nurses from a different hospital from the EFA sample. Based on previous research (Leiter et al.,
2015; Leiter and Day, 2013; Mathews and Ritter, 2016), the theoretical framework, the previous
EFA, and comparisons with other validation processes in other languages (Portoghese, et al.,
2015; Smidt et al., 2016; Tsuno et al., 2017), we expected to find five factors. Confirmatory factor
analyses were performed with IBM SPSS AMOS 24.0 (Arbuckle, 2014).
To analyse the goodness-of-fit indices of the model, we began to assess the chi square value
of sample and its significance. However, it is expected that chi-square will almost always be
statistically significant for models with N ≥ 300, because of the sensitivity of the chi-square test
to large sample sizes (Kline, 2011).
Additionally, the model was evaluated using several fit indices: the Comparative Fit Index
(CFI), the Goodness of Fit Index (GFI), the Bender-Bonnet Normed Fit Index (NFI), and the Root
Mean Square Error of Approximation (RMSEA). In general, values close to .95 for the CFI and
NFI indicate an excellent fit (Hu & Bentler, 1999; Kline, 2011; Marôco, 2014; Meyers et al., 2006),
whereas values of .90 or greater reflect a reasonable fit (Hair et al., 2006; Lomax, 2010; Mueller
& Hancock, 2010). However, especially considering the susceptibility of indices to large sample
sizes (Meyers et al., 2006), it is pointed out that for N 250 and a number of observed variables
between 12 < m < 30, CFI values of .90 or above indicate an excellent fit (Hair et al., 2006). Less
affected by sample size, GFI’ results closer to .90 indicates the best fit (Meyers et al., 2006). The
RMSEA for most acceptable models has values below .10 (Kline, 2011). Therefore, with our large
sample size, that is not surprising (Meyers et al. 2006), and we needed to employ alternative fit
measures to evaluate the proposed model. To consider the level of parsimony in the model, we
took into account the PCFI (Parsimony Comparative Fit Index), PGFI (Parsimony Goodness of Fit
Index), PNFI (Parsimony Normed Fit Index) and AIC (Akaike Information Criterion). As is well
known, there are no absolute rules or standards to determine a bad and good fit model (Hair
et al., 2006; McNeish et al., 2017). Adequacy of fit has to be judged considering the statistical
standards, psychometric reflections and internal coherence with the theoretical framework and
practical implications.
4. Results
Our results are presented in three main parts. Firstly, we examine the exploratory factor
analysis. Secondly, we present the confirmatory factor analysis using structural equation
modelling (SEM). Finally, in search of additional effects, we analyse zero-order correlations
among SIS and Burnout to assess the convergent validity.
4.1. Exploratory Factor Analysis
We performed the KMO = .93 and Bartlett = χ2 (300) = 9391.07, p < .001 tests, which indicated
appropriate values. The parallel analysis method (O’Connor, 2000) applied confirmed the five
factors. The scree plot technique based on eigenvalues indicating 5 components. The eigenvalue
95
PSIQUE • e-ISSN 2183-4806 • Volume XIX • Issue Fascículo 2 • 1st july julho-31st december dezembro 2023 pp. 86-105
Tito Laneiro, Luisa Ribeiro, Martina Nitzsche, Tânia Ferraro, Genta Kulari, Michael Leiter
for the first factor not retained was .82 (Henson & Roberts, 2006). All factors together explain 81%
of total variance.
Results confirm the 5-fatorial structure of the instrument, no item saturated outside the
expected dimension. The interpretation assigned to each of the five factors was as follows.
Factor 1 refers to incivility of supervisor behaviour, such as ignoring, excluding, speaking
rudely, behaving rudely (through gestures, facial expressions or others) and/or behaving without
respect for the participant. Factor 2 concerns the incivility behaviour of colleagues and/or other
health care professionals from the same team. Factor 3 concerns the incivility behaviour of
subordinates. Factor 4 is related to incivility behaviour of clients, customers or users, and Factor
5 includes items related to incivility on the part of the respondent. Considering that workplace
incivility can trigger a spiral of violence, the information from different elements of teams
working in the organizations and their perception of uncivil behaviour, including that of clients,
contributes to forming a more comprehensive and integrative perspective of the reality of WI in
the organizations studied.
4.2. Confirmatory Factor Analysis
As already mentioned, we examined items correlating above .40 with the respective factors
in the exploratory factor analysis (after rotation), whose content we considered not redundant
with that of other selected items and coherent with the general meaning of the factor (Hair et al.,
2006). All 5 items for each factor were keeping. The structural model tested included a general,
second-order factor influencing the five factors previously encountered in the exploratory
analysis and explaining their intercorrelations (see Figure 1). In the initial analysis with this
model, Figure 1 shows the factor structure with standardized estimates of SIS for the adjusted
model in our sample 2 (n = 427).
We found that ten items showed raised modification indices and, additionally, still emphasized
the interaction between items 1 and 2 of the subscales relate to supervisors, colleagues,
subordinates, clients and participants themselves. Theoretical plausibility was applied to judge
whether the proposed changes should be adopted. If we resort to a qualitative analysis of these
items: item 1 refers to being ignored and item 2 refers to being excluded.
We can suppose that these two behaviours are closely related, seeming similar in some peoples
understanding, probably being related to the source of these values. We decided to add paths
between residuals (allowing for covariance) and these items. Still observing the modification
indices and taking into account the meaning of some items, we added the other covariances until
reaching the best fit. In all cases, there seemed to be a rationale behind these changes – either
the items reflected behaviours which tend to co-occur (e.g. rude gestures and rude words), or
in the case of different dimensions, the might reflect the same reality from different sources
(inconsiderate behaviours across the organization, from supervisors, colleagues, subordinates,
clients and self). We did not delete any paths. The quality of the final structural model tested
showed good adjustment, as presented in Table 2.