Enhancing the Precision of the Self-Compassion Scale Short Form (SCS-SF) with Rasch Methodology

Abstract

Objectives
Precise measurement of self-compassion is essential for informing well-being–related policies. Traditional assessment methods have led to inconsistencies in the factor structure of self-compassion scales. We used Rasch methodology to enhance measurement precision and assess the psychometric properties of the Self-Compassion Scale Short Form (SCS-SF), including its invariance across Ghana, Germany, India, and New Zealand.

Method
We employed the Partial Credit Rasch model to analyse responses obtained from 1000 individuals randomly selected (i.e. 250 from each country) from a total convenience sample of 1822 recruited from the general populations of Germany, Ghana, India, and New Zealand.

Results
The initial identification of local dependency among certain items led to a significant misfitting of the SCS-SF to the Rasch model (χ² (108) = 260.26, p < 0.001). We addressed this issue by merging locally dependent items, using testlets. The solution with three testlets resulted in optimal fit of the SCS-SF to the Rasch model (χ² (27) = 23.84, p = 0.64), showing evidence of unidimensionality, strong sample targeting (M = 0.20; SD = 0.72), and good reliability (Person Separation Index = 0.71), including invariance across sociodemographic factors. We then developed ordinal-to-interval conversion tables based on the Rasch model’s person estimates. The SCS-SF showed positive correlations with measures of compassion towards others, optimism, and positive affect, alongside negative associations with psychological distress and negative affect.

Conclusions
The current study supports the reliability, as well as the structural, convergent, and external validity of the SCS-SF. By employing the ordinal-to-interval conversion tables published here, the precision of the measure is significantly enhanced, offering a robust tool for investigating self-compassion across different cultures.

Full text

Vol.:(0123456789)
Mindfulness
https://doi.org/10.1007/s12671-024-02462-y
ORIGINAL PAPER
Enhancing thePrecision oftheSelf‑Compassion Scale Short Form
(SCS‑SF) withRasch Methodology
PeterAdu1 · TosinPopoola2 · EmersonBartholomew3 · NavedIqbal4 · AnjaRoemer5 · TomasJurcik6 ·
SunnyCollings1 · CliveAspin1 · OlegN.Medvedev7 · ColinR.Simpson1,8
Accepted: 3 October 2024
© The Author(s) 2024
Abstract
Objectives Precise measurement of self-compassion is essential for informing well-being–related policies. Traditional assess-
ment methods have led to inconsistencies in the factor structure of self-compassion scales. We used Rasch methodology to
enhance measurement precision and assess the psychometric properties of the Self-Compassion Scale Short Form (SCS-SF),
including its invariance across Ghana, Germany, India, and New Zealand.
Method We employed the Partial Credit Rasch model to analyse responses obtained from 1000 individuals randomly selected
(i.e. 250 from each country) from a total convenience sample of 1822 recruited from the general populations of Germany,
Ghana, India, and New Zealand.
Results The initial identification of local dependency among certain items led to a significant misfitting of the SCS-SF to
the Rasch model (χ2 (108) = 260.26, p < 0.001). We addressed this issue by merging locally dependent items, using testlets.
The solution withthree testlets resulted in optimal fit of the SCS-SF to the Rasch model (χ2 (27) = 23.84, p = 0.64), show-
ing evidence of unidimensionality, strong sample targeting (M = 0.20; SD = 0.72), and good reliability (Person Separation
Index = 0.71), including invariance across sociodemographic factors. We then developed ordinal-to-interval conversion
tables based onthe Rasch model’s person estimates. The SCS-SF showed positive correlations with measures of compassion
towards others, optimism, and positive affect, alongside negative associations with psychological distress and negative affect.
Conclusions The current study supports the reliability,as well as the structural, convergent, and external validity of the
SCS-SF. By employing the ordinal-to-interval conversion tables published here, the precision of the measure is significantly
enhanced, offering a robust tool for investigating self-compassionacross different cultures.
Keywords Measurement· Self-Compassion Scale· Rasch Analysis· Mindfulness· Psychometrics
Research on self-compassion has gained substantial interest
in the international literature due to its impact across diverse
dimensions of well-being. For instance, a meta-analysis of
14 studies found a large negative effect size for the relation
between self-compassion and psychopathology (MacBeth &
Gumley, 2012). Hwang etal. (2019) also identified self-com-
passion as the most influential predictor of reduced educa-
tor stress in Australian students. In addition, Lefebvre etal.
(2020) established a connection between individuals’ work-
place resilience and self-compassion. Similarly, a review of
28 studies revealed that self-compassion protects against the
development of poor body image and the emergence of risk
factors for maladaptive behaviours (Braun etal., 2016).
The concept of self-compassion refers to the internal nur-
turing of emotional well-being and mental health. It involves
fully accepting and openly understanding an individual’s life
adversities without self-judgment or excessive self-criticism
(Neff, 2003a). In other words, self-compassion entails the
treatment of oneself with the same kindness, care, love, and
understanding that one will offer to a significant close rela-
tion such as a friend in times of life setbacks. The three
proposed key subconstructs of self-compassion encompass
being kind to oneself (self-kindness), acknowledging that
life challenges are part of common human experience (com-
mon humanity), and practicing awareness of thoughts and
feelings without overly identifying with them (mindfulness;
Neff, 2003a). Notably, the positive impacts of the mindful-
ness subconstruct on overall well-being have received sig-
nificant attention in the literature. For example, mindfulness
Extended author information available on the last page of the article
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Mindfulness
has been associated with reduced levels of depression,
anxiety, and stress in challenging situations, as exemplified
during the COVID-19 pandemic (Hartstone & Medvedev,
2021).
Accurate measurement of this essential positive psycho-
logical resource is vital for advancing our understanding
of its impact on mental health, guiding interventions and
treatments, and informing policies related to well-being. To
date, the assessment of self-compassion in the existing lit-
erature primarily relies on two widely recognised versions of
the same psychometric scale: the 26-item Self-Compassion
Scale (Neff, 2003b, 2016) and a 12-item Self-Compassion
Scale Short Form (SCS-SF; Raes etal., 2011). Using either
of these versions, the conceptually separate yet overarching
aspects of self-compassion are measured through positively
worded items related to self-kindness, common humanity,
and mindfulness, as well as negatively worded items related
to self-judgment, feelings of isolation, and over-identifica-
tion. The negative items reflect behaviours and thought pat-
terns that are less compassionate in nature (Neff, 2003a).
As such, the scales can be employed as a unidimensional
or multi-dimensional measure. Notably, the scale has been
instrumental in the study of self-compassion (Neff & Tóth-
Király, 2022).
Furthermore, evaluation of the psychometric properties
of the scale has predominately been conducted with con-
firmatory factor analyses (CFA; e.g. Rahman etal., 2023), a
method etymologised in Classical Test Theory (CTT). This
method has so far provided evidence supporting the reliabil-
ity and validity of the Self-Compassion Scale, demonstrating
that the scale is a suitable measure for assessing self-com-
passion across various samples (Babenko & Guo, 2019; Neff
etal., 2021). However, an ongoing controversy exists regard-
ing the factor structure of the scale. While some studies have
supported the six-factor structure, others have reported a
two-factor structure for this scale. For instance, López etal.
(2015) found that the Dutch version of the Self-Compassion
Scale supported a two-factor model, self-compassion and
self-criticism, rather than the traditional six-factor structure.
This pattern was supported by Costa etal. (2016), includ-
ing the Bangla and Turkish versions of the scale (Koğar &
Koğar, 2023; Rahman etal., 2023). A two-factor solution,
involving positive and negative self-compassion, was also
confirmed among Spanish nurses (Lluch-Sanz etal., 2022).
Studies have highlighted possible reasons for the contro-
versies surrounding the structure of the scale. For example,
studies have reported that including the uncompassionate
components items (self-judgment, feelings of isolation, over-
identification), which initially served as reversed items for
the main self-compassion measure (self-kindness, common
humanity, mindfulness) in the Self-Compassion Scale, unin-
tentionally introduced new dimensions alongside the com-
passionate components (self-kindness, common humanity,
mindfulness; Wong etal., 2003). Given that uncompassion-
ate components are reverse coded, they intend to comple-
ment compassionate aspects by providing more compre-
hensive coverage of the construct, rather than measuring
two conceptually opposing constructs. However, the uncom-
passionate items were found to be linked to negative health
outcomes such as depression symptoms, while the compas-
sionate items are associated with positive outcomes like hap-
piness. The Self-Compassion Scale now includes multiple
dimensions with different relations to external constructs,
providing an advantage of measuring specific aspects of self-
compassion independently and as a total score by reverse
coding uncompassionate items. This may create confusion
for researchers less familiar with psychometrics and meas-
urement (Muris & Otgaar, 2020), which can be addressed
by applying the unidimensional Rasch measurement model
using testlets. Ultimately, if an acceptable fit to the Rasch
model is achieved using testlets, this findingwould be taken
as evidence of unidimensionality and support for using the
total scale score (Sutton & Medvedev, 2023).
Notwithstanding, methods of analysis based on CTT
are prone to spurious correlations due to method effects,
possibly resulting in incongruent findings regarding the
dimensionality of the scales. These methods are also largely
sample dependent, which leads to the estimation of param-
eters that can over-represent the idiosyncrasies of a specific
sample rather than accurately representing the underlying
structure of the population, resulting in poor model gener-
alisability (Magno, 2009). Again, since instruments are not
perfect, observed scores could be divergent from the true
ability, state, or trait of an individual; thus, estimation of
true score under CTT does not exclude measurement error
(Eluwa etal., 2011; Magno, 2009). These inconsistencies
in the factor structures of the self-compassion scales and
the limitations of CTT raise questions about the unidimen-
sionality and measurement accuracy of the Self-Compassion
Scale, including the extent to which these scales maintain
measurement invariance across diverse countries (Hamble-
ton, 1994).
Rasch analysis, a psychometric analytical technique,
incorporates probabilistic modeling to assess and enhance
the measurement properties of items and scales (Medvedev
& Krägeloh, 2022; Tennant & Küçükdeveci, 2023). It is part
of a family of models under the Modern Test Theory (MTT)
umbrella (Ellis & Mead, 2004). MTT techniques such as
Rasch analysis transcend many of the limitations of the CTT-
derived methods such as CFA. Thus, Rasch models provide
a set of criteria, including the consideration of respondents’
abilities and item difficulties, assessing response options for
polytomous items (i.e. items with more than two response
categories), and converting ordinal-level data into a more
reliable interval-level scale. This ensures a robust assess-
ment of an instrument. Furthermore, Rasch methods align
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Mindfulness
strictly with the fundamental measurement principles laid
out by Thurstone (1931), emphasizing non-discrimination
between instrument users (invariance), unidimensionality,
and equally proportioned scale units (i.e. concatenability).
Despite the advantages of Rasch analysis in psychomet-
ric assessment, only one study (Finaulahi etal., 2021) in
the existing literature has applied this methodology to the
assessment of the self-compassion scales. However, this
study lacked cross-country validation of these scales, as
the study was conducted with an English-speaking popula-
tion only, predominantly composed of individuals of White
British ethnicity, with an overrepresentation of females.
Additionally, the researchers only assessed invariance
based on two demographic factors: age and sex. These
pitfalls lessen the confidence of applying these scales in
other contexts. The study also did not evaluate convergent
or divergent validity. Therefore, there is lack of robust
evidence regarding the psychometric characteristics of
the self-compassion scales. Hence, we sought to provide
further psychometric assessment of the 12-item SCS-SF
employing the Rasch analysis, with a primary focus on
evaluating reliability and various forms of validity, includ-
ing structural, convergent, and divergent validities. In addi-
tion to these, we examined the scale’s invariance across
sociodemographic factors such as country, age, education,
and sex using data from diverse samples in Ghana, Ger-
many, India, and New Zealand.
Based on evidence in the literature, we anticipated a positive
association between the SCS-SF scores and measures of com-
passion towards others, optimism, and positive affect, examining
the convergent validity of the scale (Neff etal., 2007). However,
we assessed divergent validity of the SCS-SF by hypothesising a
weak to zero correlation between theSCS-SF scores and meas-
ures of psychological distress, negative affect, and pessimism
(Medvedev etal., 2021; Shapira & Mongrain, 2010).
Method
Participants
We randomly selected 1000 (e.g. 250 from each country)
participants from a total sample of 1822 recruited from Ger-
many (475), Ghana (523), India (411), and New Zealand
(413) during the months of June and July 2022 for our Rasch
analysis (Fig.1). The age of participants ranged from 18 to
80years in India (Mage = 26.14; SD = 8.57), 18 to 89years
in New Zealand (Mage = 46.35; SD = 18.07), 18 to 63years
in Ghana (Mage = 29.48; SD = 5.69), and 18 to 87years in
Germany (Mage = 44.09; SD = 5.57). The randomly selected
participants differed significantly in terms of education (χ2
(6) = 708.48, p < 0.001), sex (χ2 (3) = 21.72, p < 0.001), and
age (χ2 (6) = 387.74, p < 0.001) across the sample.
Fig. 1 Flowchart for participant
sampling process from the four
countries (n = 250 per country)
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Mindfulness
Power Analysis
Rasch models are less reliant on sample size since they esti-
mate parameters from individual responses rather than data
volume (Hagell & Westergren, 2016; Tennant & Küçük-
deveci, 2023), allowing for precise estimations, even with
smaller sample sizes. This approach reduces sensitivity to
chi-square values, which can inflate statistical significance
without practical impact (Pelton, 2002). Therefore, our
sample selection was to effectively balance the benefits of
larger samples with the challenge of chi-square sensitivity.
To ensure parameter accuracy, a sample size of around 250
to 500 is recommended for Rasch analyses using the Rasch
Unidimensional Measurement Model (RUMM;Hagell &
Westergren, 2016).
Procedure
The data from Ghana and India were collected utilising
SelectSurvey.net software via various online platforms,
including Facebook, WhatsApp, Twitter, Instagram, and
email, using convenience sampling. We relied on our social
network for data from these two countries; as such, partici-
pants were not rewarded. In New Zealand and Germany, data
collection was facilitated by the Qualtrics data collection
company, and participants were remunerated. Online data
collection offers a cost-effective means of reaching diverse
populations in various locations (Lefever etal., 2007). The
questionnaires were presented in English for participants in
Ghana, India, and New Zealand, while a German version
was provided for participants in Germany. Participants ini-
tially provided demographic information and then completed
the main survey, which typically took around 15min. The
data used for the current paper were part of a larger interna-
tional dataset on psychological factors and COVID-19 vac-
cination attitudes. Sections of the current data have been
analysed using different methods and concepts, triangulat-
ing the results. For instance, previous studies have utilised
this data to establish links between psychological factors
and COVID-19 vaccination attitudes (Adu etal., 2024b) and
have adapted and validated the COVID-19 vaccination atti-
tudes scales using CFA (Adu etal., 2023, 2024a).
Measures
Self‑Compassion
The 12-item SCS-SF (Raes etal., 2011) is a self-report
questionnaire designed to assess self-compassion. It com-
prises six subscales (self-kindness, self-judgment, com-
mon humanity, isolation, mindfulness, and over-identifi-
cation), each consisting of two items. Table2 provides
detailed information regarding the sub- and full scales.
This scale is the shortened version of the main and initial
26-item Self-Compassion Scale (Neff, 2003b). The scale
uses a 5-point Likert-scale response format: 1 = Almost
Never to 5 = Almost Always. To calculate the total scores
for the SCS-SF, negative items (self-judgment, isolation,
and over-identification) are reverse-scored. Refer to the
“Results” section for the reliability (Person Separation
Index) of this scale.
Psychological Distress
We measured psychological distress using the Depression
Anxiety Stress Scale (DASS-21; Lovibond & Lovibond,
1995). This 21-item instrument is rated on a 4-point Lik-
ert-scale response option: from 0 = Did not apply to me at
all to 3 = Applied to me very much. Sample items from the
scale encompass: depression (“I couldn’t seem to experi-
ence any positive feeling at all”), anxiety (“I was aware of
dryness in my mouth”), and stress (“I found it hard to wind
down”). This scale demonstrated excellent reliability for the
overall sample (Cronbach’s α = 0.98, McDonald’s ω = 0.98;
M = 30.80, SD = 18.00).
Positive Affect andNegative Affect
We assessed Positive Affect and Negative Affect with the
popular 20-item Positive Affect (PA) and Negative Affect
(NA) Scale (PANAS; Watson etal., 1988). Each adjective
on this scale is rated on 5-point Likert scale ranging from
1 = very slightly to 5 = extremely. Examples of adjectives
measuring PA include “interested”, “strong”, and “proud”,
while NA comprises “anger”, “fear”, and “sadness”. In this
study, the reliability coefficient for the PA subscale for the
whole sample was excellent (α = 0.91; ω = 0.93, M = 31.40,
SD = 7.40). The NA subscale also ranged from very good to
excellent (α = 0.89; ω = 0.92, M = 23.00, SD = 8.70).
Optimism Versus Pessimism
The revised version of the Life Orientation Test (LOT-R;
Scheier etal., 1994) was employed to measure optimism and
pessimism. This 10-item is rated on a 5-point Likert scale
from 0 = strongly disagree to 5 = strongly agree. An example
of a positively worded item on this scale is “In uncertain
times, I usually expect the best”, and a negatively worded
item is “If something can go wrong for me, it will”. The
scale showed relatively low reliability for the total sample
(α =0.57; ω = 0.58, M = 16.10, SD = 3.62). It is not uncom-
mon to find such reliability for scales with few items (Lee
etal., 2016).
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Mindfulness
Compassion Towards Others
We utilised the Santa Clara Brief Compassion Scale
(SCBCS; Hwang etal., 2008) to evaluate Compassion
towards others. This five-item measure is scored on a 7-point
Likert scale ranging from 1 = not at all true of me to 7 = very
true of me. A sample item found on the scale is: “I tend
to feel compassion for people, even though I do not know
them”. The scale exhibited very good reliability for the total
sample (α = 0.86; ω = 0.89, M = 24.10, SD = 7.00).
Data Analyses
Data Preparation andPartial Credit Model
Data imputation was carried out using IBM SPSS (version
28); the Expectation Maximization (EM) algorithm was
employed for this purpose (Dellaert, 2002; Little, 1988).
Descriptive statistics and correlational analysis were per-
formed using SPSS. Total scores for all the multi-item scales
were calculated, and an examination of Q-Q plots, skew-
ness, and kurtosis (i.e. all within − 2 to + 2) demonstrated
normally distributed variables (George & Mallery,2011).
The advanced Rasch analysis utilised RUMM2030 (Andrich
etal., 2009), while applying the unrestricted Partial Credit
model for parameter estimations (Masters, 1982). This spe-
cialised statistical model used in item response theory was
suitable for our dataset, as it incorporates varying levels of
individual items and responses without assuming uniform-
ity of items. It further allows for modification strategies to
improve the overall scale and individual item functioning
(Bartholomew etal., 2023; Tennant & Küçükdeveci, 2023).
Overall Model Fit Estimate
Rasch analysis involves an initial assessment of the overall
model fit using a chi-square test to check how well items
interact with the latent trait. Then, each item’s fit to the
model is evaluated using item fit residuals, and a chi-square
value is calculated for each item. To confirm the Rasch
model’s overall fit, a non-significant interaction between
items and the latent trait (p > 0.05) is required (Tennant &
Küçükdeveci, 2023; Wilkinson etal., 2023). Individual item
fit residuals should fall within the range of − 2.50 to + 2.50,
and the residual correlations between individual items below
0.20 (Bartholomew etal., 2023; Christensen etal., 2017).
Local dependencies (i.e. item redundancy) can introduce
misleading (spurious) correlations affecting the overall
measurement and dimensionality. Fortunately, this concern
can be effectively handled using testlet creation methodol-
ogy (i.e. combining multiple individual items into a single,
more comprehensive assessment;Lundgren Nilsson & Ten-
nant, 2011; Tennant & Küçükdeveci, 2023).
Invariant Measurement
Differential item functioning (DIF) in Rasch analysis
assesses the consistency of a measure across various sam-
ple groups such as country, age, sex, and education (i.e.
primary, secondary, and tertiary), with the aim of avoid-
ing any DIF in individual items (Sutton & Medvedev,
2023; Tennant & Küçükdeveci, 2023). To examine age
group invariance, we applied a standard approach, creat-
ing three balanced age groups based on the 33rd and 66th
percentiles, ensuring roughly three distinct age groups:
18–29years, 30–45years, and 46–89years. DIF was
assessed using between groups ANOVA and visual inspec-
tion of individual item plots (Hagquist & Andrich, 2017;
Pratscher etal., 2022).
Reliability
The Person Separation Index (PSI) is used to evaluate the
scale’s reliability and indicates its effectiveness in distin-
guishing between different levels of an individual’s traits.
PSI values, on a scale from 0 to 1, are interpreted akin to
Cronbach’s alpha. Values exceeding 0.70 signify acceptable
reliability for group measurements, and values at or above
0.80 indicate suitability for individual assessments (Fisher,
1992).
Unidimensionality
The assessment of unidimensionality in Rasch analysis
involves the use of principal components analysis and t-tests
(Hagell, 2015). Unidimensionality is supported when ≤ 5%
of t-tests yield statistically significant results when compar-
ing person estimates between sets of items with high and
low loadings on the first principal component of residuals
(Smith, 2002). Additionally, if the lower boundary of con-
fidence intervals calculated for the number of significant
t-tests falls within the range of 5%, it indicates unidimen-
sionality. When data adhere to Rasch model assumptions, an
ordinal-to-interval transformation table is constructed using
person estimates to enhance the precision of the scale (Med-
vedev etal., 2020). The current study applied the conven-
tional threshold for statistical significance (p-value < 0.05).
Convergent andDivergent Validity
We established convergent and divergent validity by com-
puting Pearson’s correlations between the SCS-SF interval
scores and various measures, including psychological dis-
tress (depression, stress, and anxiety), positive and negative
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Mindfulness
affect, compassion towards others, and life orientation scale
(i.e. optimism versus pessimism).
Results
Initial Analysis
Our initial analysis showed the SCS-SF’s misfit to the
overall Rasch model, as there was evidence of a signifi-
cant interaction observed between the items and the latent
trait of self-compassion (χ2(108) = 260.26, p < 0.001). The
SCS-SF demonstrated a reasonable level of reliability with
a PSI = 0.65, including no evidence of unidimensional-
ity (Table1; A1 Initial). Inspection of individual items
revealed that Items 1, 7, and 11 displayed a significant
misfit to the model, items exceeding − 2.50 to + 2.50
thresholds. The items with their misfitting coefficients
are marked with an asterisk in Table2. Table2 provides
detailed information on individual item fit statistics from
the initial analysis, inclusive of item location, fit residual,
and Chi-square values for item-trait interaction.
Table 1 Rasch model fit
statistics for the initial and
final analyses of the SCS-SF
(n = 1000)
PSI = Person Separation Index without extremes
Analyses Item fit residual Person fit
residual
Goodness of fit PSI Unidimensionality
t-test
Mean SD Mean SD χ2 (df) p% Lower bound
A1 Initial 0.16 1.93 − 0.78 2.18 260.26 (108) < 0.001 0.65 9.2 7.8% (no)
A2 6 Items 0.03 1.94 − 0.66 1.50 141.64 (54) < 0.001 0.51 7.3 5.9% (no)
A3 Final − 0.11 1.00 − 0.57 1.05 23.84 (27) 0.64 0.71 4.7 3.3% (yes)
Table 2 Individual items fit statistics including the initial and final analyses of the SCS-SF (n = 1000)
Items with asterisks should be reverse coded before computing the total ordinal scores
No Initial analysis: 12 items Location Fit residual Chi Square
1 When I fail at something important to me, I become consumed by feelings of inadequacy* 0.338 3.65* 32.03*
2 I try to be understanding and patient towards those aspects of my personality I don’t like − 0.125 0.097 13.62
3 When something painful happens, I try to take a balanced view of the situation − 0.366 0.974 18.95
4 When I’m feeling down, I tend to feel like most other people are probably happier than I am* 0.092 − 0.708 6.48
5 I try to see my failings as part of the human condition − 0.081 2.052 22.77
6 When I’m going through a very hard time, I give myself the caring and tenderness I need − 0.120 − 1.646 21.59
7 When something upsets me, I try to keep my emotions in balance − 0.369 − 1.025 30.16*
8 When I fail at something that’s important to me, I tend to feel alone in my failure* 0.300 − 0.853 13.33
9 When I’m feeling down, I tend to obsess and fixate on everything that’s wrong* 0.132 − 1.189 11.31
10 When I feel inadequate in some way, I try to remind myself that feelings of inadequacy are
shared by most people
0.257 3.47* 52.48
11 I’m disapproving and judgmental about my own flaws and inadequacies* 0.090 − 1.837 24.91*
12 I’m intolerant and impatient towards those aspects of my personality I don’t like* − 0.148 − 0.999 12.64
Analysis 2: 6 super-items (Si)
Si1 Items: 2 + 6 (Self-Kindness subscale) − 0.16 − 1.55 27.27*
Si2 Items: 11 + 12 (Self-Judgment subscale) − 0.02 − 1.31 7.50
Si3 Items: 5 + 10 (common humanity subscale) 0.09 3.76* 51.12*
Si4 Items: 4 + 8 (Isolation subscale) 0.13 − 0.54 10.27
Si5 Items: 3 + 7 (Mindfulness subscale) − 0.23 0.20 25.99
Si6 Items: 1 + 9 (Over-identified subscale) 0.19 − 0.39 19.49
Final analysis: 3 super-items
Si1 Items: Si1 + Si4 0.05 − 1.26 5.11
Si2 Items: Si2 + Si3 − 0.02 0.45 9.47
Si3 Items: Si5 + Si6 − 0.03 0.48 9.25
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Mindfulness
Initial Testlet Creation
To enhance the SCS-SF’s fit to the Rasch model, we exam-
ined the residual correlation matrix, revealing local depend-
encies between items with correlations surpassing the 0.20
threshold. Such local dependencies can affect the overall
fit and dimensionality of a scale. To maintain the scale’s
validity, we addressed this issue in our subsequent analysis
by creating six testlets, aligning with the six subscales of
the SCS-SF (self-kindness, self-judgement, common human-
ity, isolation, mindfulness, over-identification; Table1: A2
6 Items and Table2). This combination of items (i.e. items
that share higher error variability) aimed to reduce meas-
urement error. However, goodness of fit to the Rasch model
was not achieved (χ2 (54) = 141.64, p < 0.001). We achieved
a reasonable level of reliability with a PSI of 0.51. The
assumption of unidimensionality remained unmet, neces-
sitating further analysis.
Final Analysis
The testlets, self-kindness, and common humanity (Table2)
showed a significant misfit to the model. Further assessment
of the residual correlation matrix involving the six testlets
revealed persistent local dependency among some testlets.
We improved the model further following the same above-
mentioned procedure to resolve this issue. This involved
the creation of three final testlets (self-kindness + isolation,
self-judgement + common humanity, mindfulness + over-
identification) from the initial six testlets. This modifica-
tion resulted in achieving overall best fit of the SCS-SF to
the Rasch model (χ2 (27) = 23.84, p = 0.64), indicated by
the lower bound of significant t-tests (3.3%) overlapping the
5% cut-off point (Table1: A3 Final); strong evidence of
unidimensionality was obtained, including the absence of
misfitting items and local dependency. A notable improve-
ment in reliability (PSI = 0.71) was observed at this stage.
Figure2, the item characteristic curve (ICC), illustrates that
alltestlets were working appropriately across different levels
of the latent trait.
DIF, Person‑Item Trait, andOrdinal‑to‑Interval
Conversion
Our DIF analysis for age, sex, education (Fig.S1 in Sup-
plementary Information), and country (Fig.3) indicated no
notable differences across any of the derived final testlets.
The person-item trait distribution of the final testlets showed
no ceiling or floor effects (Fig.4), demonstrating that 100%
of the sample were effectively targeted by items thresh-
olds of the SCS-SF with a person location mean of 0.20
(SD = 0.72). The best fit indices of the SCS-SF led to the
development of the ordinal-to-interval conversion algorithm,
which was based on the Rasch model’s person estimates,
allowing for the transformation of the ordinal scores into
interval-level data. Table3 provides detailed information
about this transformation, including how to use the table and
the scores. A paired samples t-test comparing the means of
the ordinal (M = 37.81; SD = 5.94) and Rasch-transformed
interval (M = 36.67; SD = 5.60) scores using the same scale
range revealed a true statistical difference between the inter-
val and ordinal scores (t(1821) = 42.23, p < 0.000), with a
large effect size of d = 1.00. A significant difference of 0.03
in the standard error was observed, favouring the interval
scores.
This conversion table can only be used for complete
responses to each of 12-item SCS-SF. To use this table, ordi-
nal raw scores (left column) should be obtained by adding
the observed scores for all 12 items. Next, match the ordinal
total score (12–60) to the corresponding interval score in the
right column (scale 12–60). A final converted score between
12 and 60 will be obtained, with higher scores correspond-
ing to higher levels of self-compassion.
Convergent andDivergent Validity
Pearson’s correlation coefficient analysis revealed posi-
tive associations between SCS-SF scores and meas-
ures of positive affect (r = 0.37, p < 0.001), optimism
(r = 0.51, p < 0.001), and compassion towards others
(r = 0.05, p = 0.02). Conversely, negative correlations were
observed between SCS-SF scores and measures of nega-
tive affect (r = − 0.39, p < 0.001), and psychological distress
(r = − 0.43, p < 0.001).
Discussion
We used Rasch methodology to assess the psychometric
properties,measurementinvariance, and enhanced the meas-
urement precision of the SCS-SF using a sample from four
diverse countries. Optimal Rasch model fit was attained for
the SCS-SF after combining items with high shared vari-
ability into three testlets without removing items from the
scale. This was done to preserve the validity of the SCS-
SF, mitigate spurious correlations resulting from method
effects, and reduce measurement error (Medvedev & Krä-
geloh, 2022; Wilkinson etal., 2023). These findings were
consistent with previous Rasch investigations of the SCS-SF
(Finaulahi etal., 2021).
In this study, we combined items with high shared vari-
ances unrelated to the overarching latent trait into testlets
to effectively reduce spurious correlations and related
measurement error. This approach has significant impli-
cations for the ongoing debate about the dimensionality of
the SCS-SF. Specifically, the high variance shared among
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Mindfulness
uncompassionate components (e.g. self-judgment, feelings
of isolation, and over-identification) supported the find-
ings by Wong etal. (2003) that these items represent a
unique underlying latent variable, suggesting they should
be treated as a distinct factor rather than simply reversed
items. However, the latter observation that both uncompas-
sionate and compassionate components share high vari-
ances for testlet creation could also imply the existence
of a common overarching factor that encompasses the six
compassionate and uncompassionate components of the
scale. This finding aligns with Neff’s (2016) perspective,
which supports the idea of a unidimensional measure of
self-compassion. Arguably, from a psychometric perspec-
tive, a construct does not exist if it cannot be measured
using the total score. Therefore, this evidence lends further
credence to using the unidimensional SCS-SF in research,
Fig. 2 SCS-SF item character-
istic curve (ICC) for the final
three testlets
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Mindfulness
Fig. 3 Differential item functioning (DIF) curves for country
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Mindfulness
despite the ongoing debate regarding its dimensionality
(Muris & Otgaar, 2020).
The SCS-SF also demonstrated strong sample targeting,
meaning that item difficulty levels were appropriately dis-
tributed across the range of participants’ abilities in the pre-
sent sample. In essence, the difficulty levels of the items in
the scale accurately match the various levels of proficiency
and knowledge within our sample (Sutton & Medvedev,
2023; Tennant & Küçükdeveci, 2023). The ICCs confirmed
that an item’s probability of endorsement varies across dif-
ferent levels of the latent trait being measured, suggesting
that the item effectively distinguishes between individuals
with varying trait levels (Tennant & Küçükdeveci, 2023;
Wilkinson etal., 2023). In other words, items on the SCS-SF
effectively discriminate between individuals with differing
levels of self-compassion, accurately capturing the range
and nuances between levels of self-compassion within the
sample, an essential aspect of assessment lacking in CTT
methodology. As CTT methods primarily focus on validity
and consistency of scores, MTT methods such as the Rasch
analysis assess a wider and more detailed array of psycho-
metric properties (Eluwa etal., 2011; Magno, 2009).
The established unidimensionality in the SCS-SF implies
that items measure a single overarching latent trait of self-
compassion. Hence, a single score obtained from this ver-
sion of the scale more accurately represents an individual’s
self-compassion level (Finaulahi etal.,2021; Medvedev &
Krägeloh, 2022). The use of the unidimensional SCS-SF is
particularly recommended for assessing self-compassion as
the factor structure of the self-compassion scales is unclear
and often varies between studies in the literature (Muris &
Fig. 4 Person-item thresholds distributions for the SCS-SF
Table 3 Ordinal-to-interval conversion for the 12-item SCS-SF
Ordinal scores Interval Ordinal scores Interval
logits Scale logits Scale
12 − 3.26 12.00 37 0.12 36.04
13 − 2.82 15.15 38 0.27 37.11
14 − 2.54 17.17 39 0.41 38.17
15 − 2.35 18.47 40 0.56 39.21
16 − 2.21 19.46 41 0.70 40.22
17 − 2.10 20.26 42 0.84 41.21
18 − 2.00 20.98 43 0.97 42.14
19 − 1.91 21.61 44 1.09 43.01
20 − 1.83 22.21 45 1.21 43.81
21 − 1.75 22.78 46 1.31 44.55
22 − 1.67 23.35 47 1.41 45.23
23 − 1.59 23.92 48 1.50 45.88
24 − 1.51 24.50 49 1.58 46.49
25 − 1.42 25.11 50 1.67 47.09
26 − 1.33 25.75 51 1.75 47.69
27 − 1.23 26.44 52 1.84 48.29
28 − 1.13 27.20 53 1.93 48.93
29 − 1.01 28.01 54 2.02 49.61
30 − 0.89 28.89 55 2.13 50.39
31 − 0.76 29.82 56 2.26 51.29
32 − 0.62 30.79 57 2.42 52.41
33 − 0.48 31.81 58 2.63 53.93
34 − 0.33 32.85 59 2.96 56.29
35 − 0.19 33.90 60 3.48 60.00
36 − 0.04 34.97
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Mindfulness
Otgaar, 2020). We observed sound reliability for the SCS-SF
that fulfils the conservative criteria for group assessments
(PSI ≥ 0.70) as outlined by Tennant and Conaghan (2007).
In other words, this version of the Self-Compassion Scale is
well-suited for evaluating self-compassion at a group level in
research or clinical settings. However, this reliability was not
sufficiently high for within-group assessment (i.e. repeated
measures or pre- versus post-intervention). Finaulahi etal.
(2021) found the SCS-SF to be a reliable measure of both
groups and individuals. The sightly varying results observed
between these two studies could potentially be attributed to
differences in the samples and languages across the stud-
ies. These variations in findings further complicate the con-
troversies about the reliability and validity of the SCS-SF
across samples (Muris & Otgaar, 2020).
Furthermore, it is essential to emphasise the attainment
of measurement invariance for the SCS-SF in our study
across four countries, and other sociodemographic factors
such as age, sex, and education level. This underscores the
scale’s strength in its ability to be used across a wide spec-
trum of individuals, spanning various countries, age groups,
sexes, and educational backgrounds especially following the
statistically significant difference observed between these
sociodemographic factors across countries. Measurement
invariance increases the applicability, acceptability, and
robustness of the SCS-SF, suggesting that study outcomes
stemming from the use of this scale can be confidently com-
pared (Welzel etal., 2023). Finaulahi etal. (2021) similarly
confirmed the invariance of this scale, yet this pertained
specifically to age and sex. Available studies using a CFA
approach established similar measurement invariance for
this scale, but the original factor structure of the SCS-SF
was not achieved (Meng etal., 2019).
Moreover, we utilised Rasch methodology to transform
the ordinal scores of the SCS-SF into interval-level data,
acknowledging the presence of varying intervals between
response categories (Magno, 2009; Pratscher etal., 2022).
This method provides a real-life precision in measurement
(Magno, 2009; Tennant & Küçükdeveci, 2023) diverg-
ing from the conventional hierarchical assumption among
response categories prevalent in CTT (Courville, 2004).
Notably, the interval scores were found to exhibit reduced
measurement error compared to the ordinal scores, signi-
fying that the interval scores provide a more precise and
less variable estimation of scores compared to the ordinal
scores (Bartholomew etal., 2023). Interval transformation
enhances score precision, ensuring a more accurate repre-
sentation of individual responses in a group in research or
clinical settings (Barber etal., 2022; Medvedev etal., 2018).
Furthermore, interval-level data is appropriate for use with
parametric statistical tests, as they do not violate their under-
lying assumptions. Below is an illustration of how the inter-
val scores demonstrate advantage over the ordinal scores.
Imagine person A’s initial score was 20 and person B’s
initial score was 30 before taking part in a self-compassion
intervention. Following the intervention, person A’s score
rose to 35, and person B’s score increased to 45. Solely rely-
ing on ordinal scores might suggest that both individuals
experienced a similar level of change in their self-compas-
sion levels. However, Rasch interval scores present a dif-
ferent scenario. Person A’s score increased by 11.69 units,
while person B’s score increased by 14.92 units (Table3).
Despite the seemingly comparable changes, person B’s
transformation was more than person A’s, indicating a quite
different outcome that may be clinically significant. This
emphasises the accurate measurements of the Rasch interval
scores in group studies to better discern authentic changes
in attitudes and behaviours.
Additionally, we established the convergent validity of
the SC-SF scores, demonstrating a positive correlation with
related measures such as positive affect, optimism, and com-
passion towards others. Past research consistently indicates
that self-compassion has a significant direct connection with
self-reported measures of positive affect, optimism, and
compassion towards others (Neff etal., 2007). Notwithstand-
ing, there is strong evidence regarding the negative associa-
tion between self-compassion and psychological distress, as
well as negative affect (Medvedev etal., 2021; Shapira &
Mongrain, 2010). While this finding aligns with our study
results and supports the external validity of the SCS-SF, the
evidence of divergent validity was not present. This evidence
suggests that the SCS-SF is an accurate, relevant, and appli-
cable scale for measuring self-compassion (Stöber, 2001).
Limitations andFuture Research
While our study utilised samples from four distinct loca-
tions, signifying the robustness of our findings, it is essential
to note that our instruments were predominately adminis-
tered in the English language across three of these coun-
tries, representing disparate ethnocultural groups. Of note,
the reliability coefficient of the current scale was not high
enough for within-group assessments. Considering the
potential cultural and language influences in responding to
scale items, additional research involving diverse participant
groups and translated versions of the SCS-SF using MTT
is necessary to ascertain cross-cultural consistency, appli-
cability, and the overall robustness of this scale. Another
inconsistency involved the slightly different approaches used
in recruiting the samples (e.g. rewards). The present study
primarily involved a non-clinical sample, emphasising the
need for future research to validate these results in clinical
settings, particularly among groups affected by mood disor-
ders or other psychological health conditions.
In summary, we used Rasch methodology to assess the
psychometric properties of the SCS-SF across four distinct
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Mindfulness
countries. The SCS-SF exhibited a strong fit to the Rasch
model and demonstrated unidimensionality. The SCS-SF
remained consistent across various demographic factors such
as country, age, sex, and educational background. We then
developed an algorithm for converting ordinal to interval-
level data, thereby enhancing the measurement precision
of the scale. The unidimensional SCS-SF was found to be
well-suited for assessing group-level self-compassion, and
displayed strong convergent and divergent validity. While
our large sample size increases confidence in our findings,
we encourage further research to provide similar evidence
using diverse ethnic groups and clinical samples to further
strengthen and broaden the universal tenability of the scale.
Supplementary Information The online version contains supplemen-
tary material available at https:// doi. org/ 10. 1007/ s12671- 024- 02462-y.
Acknowledgements The lead author acknowledges the receipt of the
Wellington Doctoral Scholarship for the conduct of this research and
its authorship.
Author Contribution P.A.: conceptualization, writing — original draft
preparation, data analyses. T.P.: conceptualization, writing — review-
ing and editing, supervision. O.N.M.: conceptualization, data analysis,
writing — reviewing and editing, supervision. N.I.: data collection
(India). E.B.: data analysis, reviewing and editing. T.J.: reviewing and
editing. A.R.: reviewing and editing, instrumentation. C.A.: review-
ing and editing. S.C.: reviewing and editing. C.R.S.: supervision,
reviewing and editing — final manuscript, and approval for submis-
sion. All authors contributed to the study design, drafting the paper,
revising it for important intellectual content, and gave final approval
for publication.
Funding Open Access funding enabled and organized by CAUL and
its Member Institutions.
Data Availability Study participants did not consent to having their data
shared publicly. The deidentified participant dataset from the current
study can be made available to researchers upon a reasonable request
to the corresponding author.
Declarations
Informed Consent Participants provided informed consent by clicking
a button after reading the consent information. They agreed for their
results to be published or used for academic purposes such as reports,
presentations, and public documentation, with data presented in aggre-
gate form (i.e. combined and analysed with others).
Ethics Approval The study received approval from the Human Research
Ethics Committee at Victoria University of Wellington, New Zealand
(#0000029770). The study was also in line with the Declaration of Hel-
sinki, which outlines fundamental ethical principles for health research
involving the use of human participants (World Medical Association,
2001).
Conflict of Interest The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attri-
bution 4.0 International License, which permits use, sharing, adapta-
tion, distribution and reproduction in any medium or format, as long
as you give appropriate credit to the original author(s) and the source,
provide a link to the Creative Commons licence, and indicate if changes
were made. The images or other third party material in this article are
included in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in
the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a
copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
References
Adu, P., Popoola, T., Roemer, A., Collings, S., Aspin, C., Medvedev,
O. N., & Simpson, C. R. (2023). Validation and cultural adapta-
tion of the Motors of COVID-19 Vaccination Acceptance Scale
(MoVac-COVID19S) in German. Psychological Test Adapta-
tion and Development, 4(1), 319–329. https:// doi. org/ 10. 1027/
2698- 1866/ a0000 64
Adu, P., Popoola, T., Collings, S., Aspin, C., Medvedev, O. N.,
& Simpson, C. R. (2024a). Psychometric properties of the
Motors of COVID-19 Vaccination Acceptance Scale (MoVac-
COVID19S) in New Zealand: Insights from confirmatory factor
analysis. Current Psychology, 43, 26628–26638. https:// doi. org/
10. 1007/ s12144- 024- 05877-x
Adu, P., Popoola, T., Iqbal, N., Roemer, A., Collings, S., Aspin, C.,
Medvedev, O. N., & Simpson, C. R. (2024b). Cross-country
assessment of the unique contributions of psychological fac-
tors to vaccination: Perspectives on the COVID-19 pandemic.
Journal of Health Psychology. https:// doi. org/ 10. 1177/ 13591
05324 12665 92
Andrich, D., Sheridan, B., & Luo, G. (2009).RUMM 2030. RUMM
Laboratory.
Babenko, O., & Guo, Q. (2019). Measuring self-compassion in medi-
cal students: Factorial validation of the Self-Compassion Scale-
Short Form (SCS-SF). Academic Psychiatry, 43, 590–594.
https:// doi. org/ 10. 1007/ s40596- 019- 01095-x
Barber, C. C., Middlemiss, W., & Medvedev, O. N. (2022). Applying
Rasch methodology to examine and enhance precision of the
Edinburgh postnatal depression scale. Journal of Affective Dis-
orders, 308, 391–397. https:// doi. org/ 10. 1016/j. jad. 2022. 04. 009
Bartholomew, E., Roemer, A., & Medvedev, O. N. (2023). Validation
and enhancement of the 6-Item Nature-Relatedness Scale using
Rasch analysis. Ecopsychology, 15(1), 81–90. https:// doi. org/
10. 1089/ eco. 2022. 0031
Braun, T. D., Park, C. L., & Gorin, A. (2016). Self-compassion,
body image, and disordered eating: A review of the literature.
Body Image, 17, 117–131. https:// doi. org/ 10. 1016/j. bodyim.
2016. 03. 003
Christensen, K. B., Makransky, G., & Horton, M. (2017). Critical
values for Yen’s Q3: Identification of local dependence in the
Rasch model using residual correlations. Applied Psychological
Measurement, 41(3), 178–194. https:// doi. org/ 10. 1177/ 01466
21616 6775
Costa, J., Marôco, J., Pinto-Gouveia, J., Ferreira, C., & Castilho, P.
(2016). Validation of the psychometric properties of the Self-
Compassion Scale. Testing the factorial validity and factorial
invariance of the measure among borderline personality disor-
der, anxiety disorder, eating disorder and general populations.
Clinical Psychology & Psychotherapy, 23(5), 460–468. https://
doi. org/ 10. 1002/ cpp. 1974
Courville, T. G. (2004). An empirical comparison of item response
theory and classical test theory item/person statistics (Publica-
tion No. 305067822) [Doctoral dissertation, Texas A&M Uni-
versity]. ProQuest One Academic.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Mindfulness
Dellaert, F. (2002). The expectation maximization algorithm. Col-
lege of Computing, Georgia Institute of Technology.
Ellis, B. B., & Mead, A. D. (2004). Item analysis: Theory and prac-
tice using classical and modern test theory. InHandbook of
research methods in industrial and organizational psychol-
ogy(pp. 324–343).https:// doi. org/ 10. 1002/ 97804 70756 669.
ch16
Eluwa, O. I., Eluwa, A. N., & Abang, B. K. (2011). Evaluation of
mathematics achievement test: A comparison between classical
test theory (CTT) and item response theory (IRT). Journal of
Educational and Social Research, 1(4), 99–106.
Finaulahi, K. P., Sumich, A., Heym, N., & Medvedev, O. N. (2021).
Investigating psychometric properties of the Self-Compassion
Scale using Rasch methodology. Mindfulness, 12(3), 730–740.
https:// doi. org/ 10. 1007/ s12671- 020- 01539-8
Fisher, W., Jr. (1992). Reliability, separation, strata statistics. Rasch
Measurement Transactions, 6(3), 238.
George, D., & Mallery, P. (2011). SPSS for Windows step by step: A
simple study guide and reference, 17.0 update (10th ed.). Pearson
Education India.
Hagell, P. (2015). Testing unidimensionality using the PCA/t-test pro-
tocol with the Rasch model: a cautionary note.Rasch Measure-
ment Transactions, 28(4), 1487–1489.
Hagell, P., & Westergren, A. (2016). Sample size and statistical conclusions
from tests of fit to the Rasch model according to the Rasch unidi-
mensional measurement model (Rumm) program in health outcome
measurement. Journal of Applied Measurement, 17(4), 416–431.
Hagquist, C., & Andrich, D. (2017). Recent advances in analysis of
differential item functioning in health research using the Rasch
model. Health and Quality of Life Outcomes, 15, 181. https:// doi.
org/ 10. 1186/ s12955- 017- 0755-0
Hambleton, R. K. (1994). Item response theory: A broad psychomet-
ric framework for measurement advances 1, 2. Psicothema, 6(3),
535–556.
Hartstone, J. M., & Medvedev, O. N. (2021). The role of mindful-
ness and life satisfaction in psychological distress during the
COVID-19 lockdown in New Zealand: A quasi-experimental
study. Mindfulness, 12(11), 2693–2706. https:// doi. org/ 10. 1007/
s12671- 021- 01731-4
Hwang, J. Y., Plante, T., & Lackey, K. (2008). The development of the
Santa Clara Brief Compassion Scale: An abbreviation of Spre-
cher and Fehr’s Compassionate Love Scale. Pastoral Psychology,
56(4), 421–428. https:// doi. org/ 10. 1007/ s11089- 008- 0117-2
Hwang, Y.-S., Medvedev, O. N., Krägeloh, C., Hand, K., Noh, J.-E.,
& Singh, N. N. (2019). The role of dispositional mindfulness and
self-compassion in educator stress. Mindfulness, 10(8), 1692–
1702. https:// doi. org/ 10. 1007/ s12671- 019- 01183-x
Koğar, E. Y., & Koğar, H. (2023). A validation study of the Self-Com-
passion Scale-Short Form (SCS-SF) with ant colony optimization
in a Turkish sample. Psihologija, 56(1), 89–114. https:// doi. org/
10. 2298/ PSI21 11270 15K
Lee, C.-T., Lin, C.-Y., Tsai, M.-C., Strong, C., & Lin, Y.-C. (2016).
Psychometric evaluation and wording effects on the Chinese ver-
sion of the parent-proxy Kid-KINDL. Health and Quality of Life
Outcomes, 14, 123. https:// doi. org/ 10. 1186/ s12955- 016- 0526-3
Lefebvre, J.-I., Montani, F., & Courcy, F. (2020). Self-compassion and
resilience at work: A practice-oriented review. Advances in Devel-
oping Human Resources, 22(4), 437–452. https:// doi. org/ 10. 1177/
15234 22320 94914
Lefever, S., Dal, M., & Matthíasdóttir, Á. (2007). Online data collec-
tion in academic research: Advantages and limitations. British
Journal of Educational Technology, 38(4), 574–582. https:// doi.
org/ 10. 1111/j. 1467- 8535. 2006. 00638.x
Little, R. J. (1988). A test of missing completely at random for mul-
tivariate data with missing values. Journal of the American
Statistical Association, 83(404), 1198–1202. https:// doi. org/ 10.
1080/ 01621 459. 1988. 10478 722
Lluch-Sanz, C., Galiana, L., Vidal-Blanco, G., & Sansó, N. (2022).
Psychometric properties of the Self-Compassion Scale—Short
Form: Study of its role as a protector of Spanish nurses profes-
sional quality of life and well-being during the COVID-19 pan-
demic. Nursing Reports, 12(1), 65–76. https:// doi. org/ 10. 3390/
nursr ep120 10008
López, A., Sanderman, R., Smink, A., Zhang, Y., Van Sonderen, E.,
Ranchor, A., & Schroevers, M. J. (2015). A reconsideration of the
Self-Compassion Scale’s total score: Self-compassion versus self-
criticism. PLoS ONE, 10(7), e0132940. https:// doi. org/ 10. 1371/
journ al. pone. 01329 40
Lovibond, P. F., & Lovibond, S. H. (1995). The structure of negative
emotional states: Comparison of the Depression Anxiety Stress
Scales (DASS) with the Beck Depression and Anxiety Inventories.
Behaviour Research and Therapy, 33(3), 335–343. https:// doi. org/
10. 1016/ 0005- 7967(94) 00075-U
Lundgren Nilsson, A., & Tennant, A. (2011). Past and present issues
in Rasch analysis: The functional independence measure (FIM™)
revisited. Journal of Rehabilitation Medicine, 43(10), 884–891.
https:// doi. org/ 10. 2340/ 16501 977- 0871
MacBeth, A., & Gumley, A. (2012). Exploring compassion: A meta-
analysis of the association between self-compassion and psycho-
pathology. Clinical Psychology Review, 32(6), 545–552. https://
doi. org/ 10. 1016/j. cpr. 2012. 06. 003
Magno, C. (2009). Demonstrating the difference between classical
test theory and item response theory using derived test data. The
International Journal of Educational and Psychological Assess-
ment, 1(1), 1–11.
Masters, G. N. (1982). A Rasch model for partial credit scoring. Psy-
chometrika, 47(2), 149–174. https:// doi. org/ 10. 1007/ BF022 96272
Medvedev, O. N., & Krägeloh, C. U. (2022). Rasch measurement
model. In O. N. Medvedev, C. U. Krägeloh, R. J. Siegert, & N. N.
Singh (Eds.),Handbook of assessment in mindfulness research.
Springer.https:// doi. org/ 10. 1007/ 978-3- 030- 77644-2_ 4-1
Medvedev, O. N., Turner-Stokes, L., Ashford, S., & Siegert, R. J.
(2018). Rasch analysis of the UK Functional Assessment Meas-
ure in patients with complex disability after stroke. Journal of
Rehabilitation Medicine, 50(5), 420–428. https:// doi. org/ 10. 2340/
16501 977- 2324
Medvedev, O. N., Pratscher, S. D., & Bettencourt, A. (2020). Psycho-
metric evaluation of the Interpersonal MindfulnessScale using
Rasch analysis. Mindfulness, 11(8), 2007–2015. https:// doi. org/
10. 1007/ s12671- 020- 01415-5
Medvedev, O. N., Cervin, M., Barcaccia, B., Siegert, R. J., Roemer, A.,
& Krägeloh, C. U. (2021). Network analysis of mindfulness facets,
affect, compassion, and distress. Mindfulness, 12(4), 911–922.
https:// doi. org/ 10. 1007/ s12671- 020- 01555-8
Meng, R., Yu, Y., Chai, S., Luo, X., Gong, B., Liu, B., Hu, Y., Luo, Y.,
& Yu, C. (2019). Examining psychometric properties and meas-
urement invariance of a Chinese version of the Self-Compassion
Scale–Short Form (SCS-SF) in nursing students and medical
workers. Psychology Research and Behavior Management, 12,
793–809. https:// doi. org/ 10. 2147/ PRBM. S2164 11
Muris, P., & Otgaar, H. (2020). The process of science: A critical
evaluation of more than 15 years of research on self-compassion
with the Self-Compassion Scale. Mindfulness, 11(6), 1469–1482.
https:// doi. org/ 10. 1007/ s12671- 020- 01363-0
Neff, K. (2003a). Self-compassion: An alternative conceptualization of
a healthy attitude toward oneself. Self and Identity, 2(2), 85–101.
https:// doi. org/ 10. 1080/ 15298 86030 9032
Neff, K. D. (2003b). The development and validation of a scale to
measure self-compassion. Self and Identity, 2(3), 223–250. https://
doi. org/ 10. 1080/ 15298 86030 9027
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Mindfulness
Neff, K. D. (2016). The Self-Compassion Scale is a valid and theo-
retically coherent measure of self-compassion. Mindfulness, 7(1),
264–274. https:// doi. org/ 10. 1007/ s12671- 015- 0479-3
Neff, K. D., & Tóth-Király, I. (2022). Self-compassion scale (SCS). In O.
N. Medvedev, C. U. Krägeloh, R. J. Siegert, & N. N. Singh (Eds.),
Handbook of assessment in mindfulness research. Springer Interna-
tional Publishing. https:// doi. org/ 10. 1007/ 978-3- 030- 77644-2_ 36-1
Neff, K. D., Rude, S. S., & Kirkpatrick, K. L. (2007). An examination
of self-compassion in relation to positive psychological function-
ing and personality traits. Journal of Research in Personality,
41(4), 908–916. https:// doi. org/ 10. 1016/j. jrp. 2006. 08. 002
Neff, K. D., Tóth-Király, I., Knox, M. C., Kuchar, A., & Davidson, O.
(2021). The development and validation of the state Self-Compas-
sion Scale (long-and short form). Mindfulness, 12(1), 121–140.
https:// doi. org/ 10. 1007/ s12671- 020- 01505-4
Pelton, T. (2002). Where are the limits to the Rasch advantage? Paper
presented at the International Objective Measurement Workshop.
Pratscher, S. D., Oyler, D. L., Bettencourt, B. A., & Medvedev, O. N.
(2022). Interpersonal Mindfulness Scale-Short Form development
using Rasch analyses. Mindfulness, 13(8), 2007–2021. https:// doi.
org/ 10. 1007/ s12671- 022- 01937-0
Raes, F., Pommier, E., Neff, K. D., & Van Gucht, D. (2011). Con-
struction and factorial validation of a short form of the Self-
compassion Scale. Clinical Psychology & Psychotherapy, 18(3),
250–255. https:// doi. org/ 10. 1002/ cpp. 702
Rahman, M. M., Huq, M. M., Shakib, S., & Shahrier, M. A. (2023).
Psychometric evaluation of the Bangla version of the Self-Com-
passion Scale-Short Form (SCS-SF). Mindfulness, 14(11), 2795–
2806. https:// doi. org/ 10. 1007/ s12671- 023- 02237-x
Scheier, M. F., Carver, C. S., & Bridges, M. W. (1994). Distinguishing
optimism from neuroticism (and trait anxiety, self-mastery, and
self-esteem): A re-evaluation of the Life Orientation Test. Journal
of Personality and Social Psychology, 67(6), 1063–1078. https://
doi. org/ 10. 1037/ 0022- 3514. 67.6. 1063
Shapira, L. B., & Mongrain, M. (2010). The benefits of self-compas-
sion and optimism exercises for individuals vulnerable to depres-
sion. The Journal of Positive Psychology, 5(5), 377–389. https://
doi. org/ 10. 1080/ 17439 760. 2010. 516763
Smith Jr, E. V. (2002). Detecting and evaluating the impact of multidi-
mensionality using item fit statistics andprincipal component anal-
ysis of residuals. Journal of Applied Measurement, 3(2), 205–231.
Stöber, J. (2001). The Social Desirability Scale-17 (SDS-17): Con-
vergent validity, discriminant validity, and relationship with age.
European Journal of Psychological Assessment, 17(3), 222–232.
https:// doi. org/ 10. 1027/ 1015- 5759. 17.3. 222
Sutton, A., & Medvedev, O. N. (2023). Development and valida-
tion of the Awareness Outcomes Measure (AOM) using Rasch
approach. Mindfulness, 14(2), 473–481. https:// doi. org/ 10. 1007/
s12671- 022- 02047-7
Tennant, A., & Küçükdeveci, A. A. (2023). Application of the Rasch
measurement model in rehabilitation research and practice: Early
developments, current practice, and future challenges. Frontiers
in Rehabilitation Sciences, 4, 1208670. https:// doi. org/ 10. 3389/
fresc. 2023. 12086 70.
Thurstone, L. L. (1931). The measurement of social attitudes. The
Journal of Abnormal and Social Psychology, 26(3), 249–269.
https:// doi. org/ 10. 1037/ h0070 363
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and
validation of brief measures of positive and negative affect: The
PANAS scales. Journal of Personality and Social Psychology,
54(6), 1063–1070. https:// doi. org/ 10. 1037/ 0022- 3514. 54.6. 1063
Welzel, C., Brunkert, L., Kruse, S., & Inglehart, R. F. (2023). Non-
invariance? An overstated problem with misconceived causes.
Sociological Methods & Research, 52(3), 1368–1400. https:// doi.
org/ 10. 1177/ 00491 24121 995
Wilkinson, S., Ribeiro, L., Krägeloh, C. U., Bergomi, C., Parsons,
M., Siegling, A., Tschacher, W., Kupper, Z., & Medvedev, O.
N. (2023). Validation of the Comprehensive Inventory of Mind-
fulness Experiences (CHIME) in English using Rasch method-
ology. Mindfulness, 14(5), 1204–1218. https:// doi. org/ 10. 1007/
s12671- 023- 02099-3
Wong, N., Rindfleisch, A., & Burroughs, J. E. (2003). Do reverse-
worded items confound measures in cross-cultural consumer
research? The case of the Material Values Scale. Journal of Con-
sumer Research, 30(1), 72–91. https:// doi. org/ 10. 1086/ 374697
World Medical Association. (2001). World Medical Association Decla-
ration of Helsinki. Ethical principles for medical research involv-
ing human subjects. Bulletin of the World Health Organization,
79(4), 373.
Publisher's Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Mindfulness
Authors and Aliations
PeterAdu1 · TosinPopoola2 · EmersonBartholomew3 · NavedIqbal4 · AnjaRoemer5 · TomasJurcik6 ·
SunnyCollings1 · CliveAspin1 · OlegN.Medvedev7 · ColinR.Simpson1,8
* Peter Adu
adupe@staff.vuw.ac.nz
Tosin Popoola
Tosin.Popoola@newcastle.edu.au
Emerson Bartholomew
emerson.bartholomew@auckland.ac.nz
Naved Iqbal
niqbal@jmi.ac.in
Anja Roemer
A.Roemer@massey.ac.nz
Tomas Jurcik
dr.tomas.jurcik@gmail.com
Sunny Collings
sunny.collings@vuw.ac.nz
Clive Aspin
clive.aspin@vuw.ac.nz
Oleg N. Medvedev
oleg.medvedev@waikato.ac.nz
Colin R. Simpson
colin.simpson@vuw.ac.nz
1 School ofHealth, Wellington Faculty ofHealth, Victoria
University ofWellington, Wellington, NewZealand
2 School ofNursing andMidwifery, College ofHealth,
Medicine andWellbeing, The University ofNewcastle,
Callaghan, Australia
3 Department ofPsychological Medicine, Faculty ofMedical
andHealth Sciences, University ofAuckland, Auckland,
NewZealand
4 Jamia Millia Islamia, NewDelhi, India
5 School ofPsychology, Massey University, PalmerstonNorth,
NewZealand
6 University ofBritish Columbia, Vancouver, Canada
7 School ofPsychological andSocial Sciences, University
ofWaikato, Hamilton, NewZealand
8 Usher Institute, The University ofEdinburgh, Edinburgh, UK
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com