Developing a standardised measure of psychological safety.

Abstract

Background: Psychological safety is increasingly recognised as central to mental health and wellbeing. The Polyvagal Theory offers a ‘Science of Safety’ which can help inform clinical practice to promote wellbeing, resilience and post-traumatic growth, whilst mitigating trauma. To date, there is no standardised measure of psychological safety comprising psychological, physiological and social components. The current study aims to develop such a self-report measure.
Methods: The study comprised three stages: First, seven key stakeholders with expertise in trauma, applied psychology and The Polyvagal Theory generated 107 items using a variation of the Delphi method. The second stage aimed to reduce the number of items and assess the underlying factor structure. Participants (n =342) completed the 107 item measure which was disseminated via an anonymised, online Qualtrics survey. Exploratory factor analysis guided by Cronbach’s alpha and omega coefficients resulted in a reduced scale with 29 items of psychological safety. Ethical approval was granted by the University of Strathclyde ethics committee.
Current stage of work: Data collection began for the final phase in February 2021 and should be completed by June 2021. The 29 item ‘safety measure’ has been disseminated via an online Qualtrics survey with five other measures for the purpose of comprehensive psychometric validation via tests of dimensionality, reliability, known groups, criterion, convergent, predictive and discriminant validity.
Expected results: The findings will result in a psychometrically validated measure of psychological safety.
Discussion: The development of a holistic, standardised measure of psychological safety will benefit both academic and clinical work.

Full text

BACKGROUND
Psychological safety is increasingly
recognised as central to mental health &
wellbeing. The polyvagal theory offers a
‘Science of Safety’ which can help inform
clinical practice to promote wellbeing,
resilience & post-traumatic growth, whilst
mitigating trauma. To date, there is no
standardised measure of psychological
safety comprising psychological,
physiological & social components. The
current study aims to develop this.
METHODS & RESULTS
Ethical approval was granted by the
University of Strathclyde ethics committee.
The study comprised 3 stages:
Stage 1: Seven key stakeholders with
expertise in trauma, applied psychology &
the polyvagal theory generated 107 items
using a variation of the Delphi method.
Stage 2: Aimed to reduce the number of
items & assess the underlying factor
structure. Items from stage 1 were
evaluated by participants (n=342) via an
anonymised online survey. Number of
items for retention was guided by
exploratory factor analysis & internal
consistency measures. A preliminary
Neuroception of Psychological Safety Scale
(NPSS) of 29 items emerged, with a best fit
of 3-factor structure described as: 1. Body
sensations; 2. Compassion; 3. Social
engagement.
Stage 3: Aimed to evaluate the reliability &
validity of stage 2. A new pool of
participants (n=318) responded to an
online survey consisting of NPSS & related
existing measures: the Team Psychological
Safety Scale (TPSS; Edmondson, 1999);
Body Perception Questionnaire - Very
Short Form (BPQ; Cabrera et al., 2018);
Abbreviated Post-Traumatic Checklist –
Civilian Version (PCL; Lang & Stein, 2005);
World Assumptions Questionnaire Safety
& Vulnerability Subscale (WAQ-SV;
Kaler et al., 2008); The Ten Item
Personality Inventory (TIPI; Gosling et al.,
2003).
Developing a standardised measure of psychological safety
Known-Groups Validity
Women’s sense of safety is reportedly lower
than men’s (Logan & Walker, 2017),
predicting lower NSS scores. Mann-Whitney
testing showed significant group differences
(p= .002) with male scores (Mdn = 115, IQR
= 17) lower than female scores (Mdn = 120,
IQR = 16.5). At subscale, genders differed on
Social Engagement, (p < .001; female Mdn =
59, IQR = 8; male Mdn = 55.5, IQR = 12.75).
Association with related scales
Spearman’s correlations showed significant
but weak correlation of criterion measure
TPSS with NPSS scores (r = .14, p= .013),
whilst BPQ, PCL & WAQ-SV did not show
relationship.
Spearman’s correlations showed weak but
significant correlation of NSS sum scores &
TIPI subscales extraversion (r = .17, p=
.003), & agreeableness (r = .26, p< .001).
Discussion & Conclusion
The NPSS shows good internal consistency.
The factors Body sensations, Compassion,
& Social engagement are coherent with the
polyvagal theory & analysis of reports from
participants. While safety is an internal
sensation, it is dependent on our social
environment.
The low convergence with existing
measures may highlight the existing gap in
research to measure psychological safety.
Future research will evaluate validity using
assessments more consistent with our three
dimensions of safety & evaluate use in
clinical populations. With these
improvements, the NPSS has the potential
to be a valuable scale with broad
applicability in research & applied
psychological practice.
MORE INFO CONTACT:
Liza.Morton@Glasgow.ac.uk
Nicola.Cogan@strath.ac.uk
Morton, L²., Cogan, N¹., Kolacz, J³., Nikolic, M ¹., Calderwood, C¹., Bacon, T4., Pathe, E 4., Williams, D¹., Porges, S³.
¹ University of Strathclyde, Glasgow, Scotland, ² University of Glasgow, Scotland, ³ Kinsey Institute, Indiana, US, 4 NHS, Scotland
RESULTS
Internal consistency of NPSS
Stage 2 responses centered around value
“Agree” (4): Means (4.15, SD .85) were high
in relation to range (1-5); negatively skewed
(-.92 &), and leptokurtic (.64). Stage 3 sum
score descriptive statistics revealed a similar
pattern, with m 119.48, SD 14.85, (range 29-
145) skew .97, kurtosis .75.
Cronbach’s alpha & Omega estimates
suggest items of the NPSS are reliable at
total scale & subscale levels.
Confirmatory Factor Analysis confirmed
factor structure. The body sensations,
compassion & social engagement factors fit
indices met or closely approached cut-off
levels of good fit. Factor loadings showed
acceptable fit of items.
Content Analysis
Participants were asked to describe a
situation that made them feel safe. 67% of
respondents chose interpersonal situations
involving loved ones, friends, colleagues or
caring professionals. Only 3% chose
explicitly asocial situations.
References
Calderwood, C. (2021) Psychometric Validation of the Neuroception of Safety Scale [MSc Thesis, University of Strathclyde]
Nikolič, M. (2020) Developing the Neuroception of Safety Scale [MSc Thesis, University of Strathclyde]
Porges, S. W. (2011) The Polyvagal Theory: Neurophysiological Foundations of Emotions, Attachment, Communication, and Self-
regulation. New York: WW Norton.