Validation and Measurement Invariance of the Personal Financial Wellness Scale: A Multinational Study in 7 Countries

Abstract

In 2020, 17.1% of the population in the European Union was at risk of poverty ( Eurostat, 2021 ). Poverty is often assessed using objective measures such as absolute and relative income levels. However, different individuals may experience different levels of financial stress at the same income level. Therefore, it is crucial to have measures that capture the subjective components of poverty. In this multinational study, we tested the validity and measurement invariance of the Personal Financial Wellness (PFW) scale across six European countries (Germany, Italy, the Netherlands, Slovenia, Spain, and the UK) and the US, and six languages (German, Italian, Dutch, Slovenian, Spanish, and English). Results provided mixed evidence for the fit of the expected one-factor structure. Exploration of a modified one-factor structure indicated an improved fit. The scale showed excellent reliability, and convergent and discriminant validity. This suggests that the PFW scale captures subjective financial stress and is a dependable self-report measure. Measurement invariance testing of the modified one-factor model showed metric invariance across Slovenia, Spain, the UK, and the US. Given that scalar invariance was not achieved and the invariance testing was based on an exploratory model, we do not advise the use of the scale for comparisons between countries.

Full text

Special Issue: Advancing the Reproducibility of Psychological Assessment
Across Borders and Populations
Registered Report
Validation and Measurement
Invariance of the Personal
Financial Wellness Scale
A Multinational Study in 7 Countries
Eike K. Buabang
1,2
, Sarah Ashcroft-Jones
3
, Celia Esteban Serna
4
, Katarina Kastelic
5,6
,
Jakob Kveder
5
, Amanda Lambertus
7
, Tasja S. Müller
8
, and Kai Ruggeri
9
1
Research Group of Quantitative Psychology and Individual Differences, KU Leuven, Belgium
2
Center for Social and Cultural Psychology, KU Leuven, Belgium
3
Department of Experimental Psychology, University of Oxford, UK
4
Division of Psychology & Language Sciences, University College London, UK
5
Department of Psychology, Faculty of Social Sciences, Lund University, Sweden
6
Novo Mesto Health Center, Novo Mesto, Slovenia
7
Institute of Psychiatry Psychology & Neuroscience, King’s College London, UK
8
Faculty of Behavioural and Social Sciences, University of Groningen, The Netherlands
9
Department of Health Policy and Management, Mailman School of Public Health, Columbia University, New York City, NY, USA
Abstract: In 2020, 17.1% of the population in the European Union was at risk of poverty (Eurostat, 2021). Poverty is often assessed using
objective measures such as absolute and relative income levels. However, different individuals may experience different levels of financial
stress at the same income level. Therefore, it is crucial to have measures that capture the subjective components of poverty. In this
multinational study, we tested the validity and measurement invariance of the Personal Financial Wellness (PFW) scale across six European
countries (Germany, Italy, the Netherlands, Slovenia, Spain, and the UK) and the US, and six languages (German, Italian, Dutch, Slovenian,
Spanish, and English). Results provided mixed evidence for the fit of the expected one-factor structure. Exploration of a modified one-factor
structure indicated an improved fit. The scale showed excellent reliability, and convergent and discriminant validity. This suggests that the
PFW scale captures subjective financial stress and is a dependable self-report measure. Measurement invariance testing of the modified one-
factor model showed metric invariance across Slovenia, Spain, the UK, and the US. Given that scalar invariance was not achieved and the
invariance testing was based on an exploratory model, we do not advise the use of the scale for comparisons between countries.
Keywords: Personal Financial Wellness Scale, validation, measurement invariance, financial stress, poverty
In 2020, over 75 million people in the European Union (EU)
were at risk of poverty, meaning they had 60%orlessof
the median disposable income (adjusted for household size)
in their country (Eurostat, 2021). This at-risk-of-poverty
threshold defines poverty based purely on the objective
measure of income. However, this definition fails to
address how individuals perceive their financial situation.
For example, the same income can be perceived differently
by different individuals (Prawitz et al., 2006), creating
unique subjective experiences of financial stress. Sommet
and colleagues (2018) found that the perception of income
is only moderately correlated with actual income level and
that it is the perception itself that relates to increased rates
of unhappiness and poor mental health. Therefore, to
understand financial stress, it is critical to have measures
that accurately capture these subjective experiences.
Subjective financial stress has been measured in a
number of ways. One approach involves defining a subjec-
tive poverty line in economic space (Ravallion, 2012). This
can be achieved by asking people to assess their financial
situation in relation to others by placing themselves on a
European Journal of Psychological Assessment (2022), 38(6), 476–486 Ó2022 Hogrefe Publishing
https://doi.org/10.1027/1015-5759/a000750
https://econtent.hogrefe.com/doi/pdf/10.1027/1015-5759/a000750 - Eike K. Buabang <eike.buabang@tcd.ie> - Monday, December 12, 2022 6:08:33 AM - IP Address:134.226.106.223

rung of an economic ladder, where the bottom two rungs
are indicative of subjective poverty (e.g., Ravallion &
Lokshin, 2002). Alternatively, researchers can ask what
people consider to be the minimum income they require
to “make ends meet”and the subjective poverty line is
drawn at their minimum required income (e.g., Hagenaars
& van Praag, 1985). Another approach simply asks people if
they are able to satisfy their basic needs, thereby measuring
financial strain (e.g., Kahn & Pearlin, 2006). Finally, subjec-
tive experiences can be captured by measuring affective
states, which include feelings of worry, stress, and anxiety
about one’s financial situation (e.g., Shapiro & Burchell,
2012). Individually, these measures give insight into partic-
ular aspects of subjective financial stress, however, to gain a
more complete picture a more holistic measure is required.
A candidate scale to measure subjective financial stress in a
more complete manner is the Personal Financial Wellness
Scale
Ó
(PFW Scale; Prawitz et al., 2006).
The PFW scale (Table 1) consists of eight items measuring
financial stress and well-being (Prawitz et al., 2006). It was
initially developed under the name InCharge Finan-
cial Distress/Financial Well-being (IFDFW) scale and was
validated in the United States in the general population
and in consumer credit counseling clients (Prawitz et al.,
2006). The scale includes items that measure different
aspects of financial stress, such as financial strain and
affective states. Lower scores on the scale indicate lower
levels of financial well-being and higher levels of financial
stress. Different PFW scores have been associated with dif-
ferent financial behaviors. For example, improved budget-
ing and saving when financial well-being is high, or risky
credit card usage and compulsive buying when financial
stress increases (Gutter & Copur, 2011). Furthermore, peo-
ple who contacted a credit counseling agency scored lower
onthePFWscale(Prawitzetal.,2006). Given the breadth
of the PFW scale and its relation to real-life financial behav-
iors, in this study, we aimed to test the validity and measure-
ment invariance of the PFW scale across six European
countries (Germany, Italy, the Netherlands, Slovenia, Spain,
the United Kingdom), and the United States.
The validity and reliability of the scale were first estab-
lishedintheUS.Prawitzandcolleagues(2006)providea
detailed account of how the scale was initially developed.
A principal component analysis (PCA) of the final 8-item
scale indicated that the items measured one factor, explain-
ing 78.9% of the variance. The loadings for each item on the
factor were between .83 to .93.Thescaleshowedhighinter-
nal consistency, indicated by a Cronbach’sαof .96.The
validity and reliability of the scale were replicated in a college
sample from the US (Gutter & Copur, 2011). Using a PCA,
64% of the total variance was explained by one factor, with
factor loadings ranging from .69 to .90 and a Cronbach’sα
coefficient of .91.AfurtherstudybyNielsen(2010)using
confirmatory factor analysis (CFA) indicated that the one-
factor solution explained just 56.6% of the variance. Cron-
bach’sαcoefficient was between .89 and .90 depending
on the handling of missing data. Nielsen (2010)proposed
analternativetwo-factorstructureforthescaledistinguish-
ing between items that measure objective (Items 1,2,3,8)
and subjective (Items 4–7) financial stress. A forced CFA
for the two-factor model demonstrated a reasonable fit, how-
ever, the eigenvalue of the objective factor did not exceed
the cut-off value of 1. There may also be theoretical reasons
against adopting a two-factor model as items from both sub-
scales are still a subjective assessment of financial stress.
Collectively these findings establish the validity and reliabil-
ity of the PFW scale in US and indicate that a one-factor
structure is a reasonable fit across multiple samples.
Outside the US context, Kamaluddin and colleagues
(2018) translated the scale to Malay and examined validity
and reliability in Malaysia. They examined the scale using
an exploratory factor analysis (EFA) with PCA as the extrac-
tion method. They found a one-factor structure, explaining
65.28% of the total variance. The factor loadings were
between .62 and .88.Cronbach’sαcoefficient was .92.In
the current study, we aimed to investigate the validity of
the PFW scale in the European context. Our sample includes
six countries from different areas of Europe (Germany, Italy,
the Netherlands, Slovenia, Spain, and the UK). This study
involved the translation of the scale into five languages
(German, Italian, Dutch, Slovenian, and Spanish). Beyond
the European context, we also attempted to replicate the
one-factor structure in the US. We planned to conduct a
multi-group CFA, testing the structure and measurement
invariance of the scale across these seven countries.
Arguably, the role of financial stress differs between the
countries included in the current study. For example, com-
pared to Europe, income inequality is higher in the US
(Blanchet et al., 2020). Christelis and colleagues (2017)find
that more US households have debt compared to European
households and that the debt is relatively higher. According
to the Stress in America report (American Psychological
Association, 2021), 61% of Americans report that money
is a significant source of stress. Based on these differences,
we might expect higher baseline financial stress levels in
the US compared to the European countries. There are
other differences between the countries, for instance, the
extent to which they are individualistic or collectivistic
cultures (Hofstede et al., 2010)whichmayinfluencesub-
jective financial stress. While we acknowledge these
differences, we did not make specific predictions for the
differences between the countries.
To account for these differences, and to allow for differ-
ences in how the one-factor model of the PFW scale fits
across countries, we planned to conduct separate CFAs for
each country. We planned to then conduct a multi-group
Ó2022 Hogrefe Publishing European Journal of Psychological Assessment (2022), 38(6), 476–486
E. K. Buabang et al., Validation PFW Scale 477
https://econtent.hogrefe.com/doi/pdf/10.1027/1015-5759/a000750 - Eike K. Buabang <eike.buabang@tcd.ie> - Monday, December 12, 2022 6:08:33 AM - IP Address:134.226.106.223

CFA and measurement invariance testing, excluding those
countries with an unacceptable model fit.Further, we exam-
ined convergent validity by assessing the extent to which
the PFW scale is related to the Perceived Stress Scale
(PSS; Cohen et al., 1983). Convergent correlations between
the PSS and objective financial condition have previously
been found in various countries (e.g., Croatia (Fazlić
et al., 2012) and Spain (Vallejo et al., 2018)). Lower scores
on the PFW scale indicate more financial stress. Therefore,
we expected a strong negative correlation (> .50) between
the PFW scale and PSS scores across all countries.
Finally, we assessed discriminant validity by relating the
PFW scale to the Ten Item Personality Inventory (TIPI;
Gosling et al., 2003). Financial stress has been found to
negatively correlate with conscientiousness and positively
with neuroticism (Xu et al., 2015). The relations between
financial stress and extraversion, openness to experience,
and agreeableness are less consistent. In line with these
findings, we expected moderately negative correlations
(.30 to .50) between conscientiousness and PFW scores,
and moderately positive correlations (.30 to .50) between
neuroticism and PFW scores. We expected low correlations
for the remaining three traits. For both convergent and
discriminant validity, correlations were also compared
between the US and each European country.
Methods
Participants
We aimed to recruit comparable participant samples of
at least 200 participants per country to meet the
Table 1.Personal Financial Wellness Scale
Ó
Item Score (1–10)
What do you feel is the level of your financial stress today? 1 = Overwhelming stress
4 = High stress
7 = Low stress
10 = No stress at all
How satisfied are you with your present financial situation? 1 = Completely dissatisfied
4 = Somewhat dissatisfied
7 = Somewhat satisfied
10 = Completely satisfied
How do you feel about your current financial condition? 1 = Feel overwhelmed
4 = Sometimes feel worried
7 = Not worried
10 = Feel comfortable
How often do you worry about being able to meet normal monthly living expenses? 1 = All the time
4 = Sometimes
7 = Rarely
10 = Never
How confident are you that you could find the money to pay for a financial emergency that costs about $1,000? 1 = No confidence
4 = Little confidence
7 = Some confidence
10 = High confidence
How often does this happen to you? You want to go out to eat, go to a movie or do something else
and don’t go because you can’t afford to?
1 = All the time
4 = Sometimes
7 = Rarely
10 = Never
How frequently do you find yourself just getting by financially and living paycheck to paycheck? 1 = All the time
4 = Sometimes
7 = Rarely
10 = Never
How stressed do you feel about your personal finances in general? 1 = Overwhelming stress
4 = High stress
7 = Low stress
10 = No stress at all
Note. Each item is scored from 1 to 10 and the total is divided by 8. Lower scores indicate lower financial well-being. Scale used with permission.
European Journal of Psychological Assessment (2022), 38(6), 476–486 Ó2022 Hogrefe Publishing
478 E. K. Buabang et al., Validation PFW Scale
https://econtent.hogrefe.com/doi/pdf/10.1027/1015-5759/a000750 - Eike K. Buabang <eike.buabang@tcd.ie> - Monday, December 12, 2022 6:08:33 AM - IP Address:134.226.106.223

recommended minimum sample size for multi-group CFA
(Fischer & Karl, 2019). We recruited a total of 1,877 partic-
ipants. Those who did not consent to or complete the study
were excluded from the sample prior to data cleaning leav-
ing 1,553 participants. Participants were excluded based on
the time they took to complete the study. Participants who
took less than 25% of the mean duration per sampling type
were excluded (n=3) leaving a final sample of N=1,550
participants that were included in the analyses across the
seven countries (Germany: (n=205), Italy: (n=213), Nether-
lands: (n=207), Slovenia: (n=212), Spain: (n=203), UK: (n
=288), US: (n=222). Full demographic information is avail-
able in the Supplementary Materials, Table 1and Figure 1:
https://osf.io/gevmr/).
Data missingness was also checked, overall <0.1%was
missing and 100% of the PFW data was present. Some data
was missing for TIPI items, these participants (n=22)were
included in the main analysis but underwent listwise dele-
tion for the validity analyses.
Procedure
The survey was created using Qualtrics software. Partici-
pants were recruited in two ways. First, using the Prolific
platform on which participants were compensated accord-
ing to the minimum fee. Second, using snowball sampling
on social media. The proportions of participants recruited
using each method varied across countries. For example,
all UK and US participants were sampled via snowball sam-
pling, while all participants from the Netherlands were sam-
pled using Prolific. For the other countries, the proportions
sampled on Prolific or via snowball sampling were mixed
(Prolific Proportion: Germany = 0.78, Italy = 0.80, Slovenia
=0.43,Spain=0.88). At the start of the survey, participants
were informed of the confidentiality and anonymity of the
collected data, and the voluntary nature of participation.
Next, they were presented with a series of measures includ-
ing the PFW scale, PSS, and TIPI. The validation of the
PFW scale is part of a larger study (the full list of measures
is available in the Supplementary Materials, https://osf.io/
gevmr/). Finally, participants were debriefed and provided
with the contact details of the research team.
Materials
The Personal Financial Wellness Scale (PFW scale;
Prawitz et al., 2006)
Each item is scored from 1to 10 and the total score is
divided by 8, the total number of items. Lower scores indi-
cate higher financial distress and lower financial well-being.
(See Table 1for item details, scale used with permission
1
).
Perceived Stress Scale (PSS; Cohen, 1988)
The 10-item PSS (Cohen, 1988)wasusedbecausethe
psychometric properties were found to be superior to those
of the 14-item PSS (Lee, 2012). This version includes six
negatively worded items (e.g., “In the last month, how often
have you felt nervous and “stressed”?”) and four positively
worded items (e.g., “In the last month, how often have you
felt that things were going your way?”). Each item uses a 5-
point Likert format ranging from 0=never to 4=very often.
Total scores are obtained by reverse coding the four posi-
tively worded items and summing across all scale items.
Scores will range from 0to 40 points, where higher scores
indicate higher perceived stress.
Ten Item Personality Inventory (TIPI; Gosling et al.,
2003)
The TIPI is a short ten-item measure of the Big-Five per-
sonality dimensions (extraversion, agreeableness, conscien-
tiousness, neuroticism, and openness to experience). Each
dimension is measured with two items on a 7-point scale.
The scale asks to report agreement (disagree strongly to
agree strongly)toitemssuchas“I see myself as extraverted,
enthusiastic”.
0.4 0.0 0.4
O
C
E
A
N
PSS
Pearson Correlation
Scale
Country
Germany
Italy
Netherlands
Slovenia
Spain
UK
US
Figure 1. Correlations between the PFWS and the TIPI traits and PSS
scores country. Correlations between the PFWS and the TIPI and PSS
scales are shown. Correlations were computed within each country.
The five personality traits assessed by the TIPI subscales are denoted
by the letters OCEAN (Openness, Conscientiousness, Extraversion,
Agreeableness, Neuroticism).
1
ÓCopyright by Personal Finance Employee Education Fund (http://www.pfeef.org), E. Thomas Garman, and/or John Hoffmire. (1) E. Thomas
Garman, Fellow and Professor Emeritus, Virginia Tech University; 1761 Pennecamp Drive, The Villages, FL 32162, USA; Tel.: 352-205-4283;
E-mail: ethomasgarman@yahoo.com. (2) John Hoffmire, Emeritus, University of Wisconsin-Madison, 12 The Paddock, Oxford, OX1 5SB, UK; Tel.:
+44 1865 701914; Mobile: +44 78844 72169; E-mail: hoffmire@wisc.edu. Obtain permission for use from any of the above. All rights reserved.
Ó2022 Hogrefe Publishing European Journal of Psychological Assessment (2022), 38(6), 476–486
E. K. Buabang et al., Validation PFW Scale 479
https://econtent.hogrefe.com/doi/pdf/10.1027/1015-5759/a000750 - Eike K. Buabang <eike.buabang@tcd.ie> - Monday, December 12, 2022 6:08:33 AM - IP Address:134.226.106.223

Translation Procedure
The translation was done in three steps. First, native speak-
ers translated the scales from their original English version
into the target languages (German, Italian, Dutch, Slove-
nian, and Spanish). Second, in each country, a set of bilin-
gual speakers who were unaware of the scale and the
purpose of the study conducted a back-translation. Finally,
for each language, a bilingual psychometric expert evalu-
ated the translation of the scale, which was compared to
the original version.
Data Analysis
We used R (Version 4.2.0; R Core Team, 2022) and the R-
packages cocor (Version 1.1.3; Diedenhofen & Musch, 2015),
dplyr (Version 1.0.8; Wickham et al., 2022), Hmisc (Version
4.6.0; Harrell Jr., 2022), kableExtra (Version 1.3.4; Zhu,
2021), knitr (Version 1.39;Xie,2015), lavaan (Version
0.6.11; Rosseel, 2012), lordif (Version 0.3.3;Choietal.,
2016), Olivoto2020 (Olivoto & L’ucio, 2020), papaja (Ver-
sion 0.1.0.9999;Aust&Barth,2020), semTools (Version
0.5.5; Jorgensen et al., 2021), tidyverse (Version 1.3.1;Wick-
ham et al., 2019), userfriendlyscience (Version 0.7.2; Peters,
2017), and visdat (Version 0.5.3;Tierney,2017)forallour
analyses.
Testing measurement invariance of the PFW scale fol-
lows the guidelines by Fischer and Karl (2019)andused
the lavaan package in R. First, we conduct separate CFAs
for each country to examine the one-factor model fit in
each country. The cut-off scores for the interpretation of
each index are presented in Table 2(adapted from Greiff
&Allen,2018).
To assess internal consistency, we report Cronbach’sα
and ω(Crutzen & Peters, 2017). Alpha and omega are
obtained using the userfriendly science package in R and
the code provided by Peters (2014). Confidence intervals
for αand ωare computed through bootstrapping (1,000
samples). Observation of poor fit in any country results in
exclusion from the multi-group CFA and measurement
invariance testing. Excluded countries are subject to further
examination to explore reasons for their poor fit.
We plan to run a multi-group CFA testing the one-factor
model across the remaining countries. The procedure for
the multi-group CFA is as follows: after assessing the model
fit indices, we test configural, metric, scalar, and strict
invariance. We accept a more constrained model if the
ΔCFI between a less constrained and more constrained
model is 0.01.IftheΔCFI is >0.01, we test the partial
invariance of the last model that was not invariant. We
check which parameter would result in the largest ΔCFI if
we allow it to vary across countries. The possible variation
of this parameter is also evaluated on a theoretical level.
A model in which this parameter is free from the constraint
is compared against the last model that was invariant. ΔCFI
is assessed again. If ΔCFI is 0.01,weacceptapartial
invariant model. If it is >0.01, freeing up another parame-
ter will be considered.
Results
Registered Analyses
The one-factor model was first run in each country sepa-
rately (see Table 3). The CFI was excellent (>.95)inGer-
many and the US, it was good (>.92)inItaly,the
Netherlands, and Slovenia, and acceptable (>.89)inSpain
and the UK. The TLI was good (>.92)inGermany,Slove-
nia, and the US, it was acceptable (>.89)inItalyandthe
Netherlands, and questionable (>.86)inSpainandthe
UK. The RMSEA was poor for each country with values
ranging above .10. The SRMR was excellent (<.06)inall
countries. Similarly, the inter-item reliability measures were
excellent (>.90) in all countries. Based on the unacceptable
RMSEA values, no countries met the cut-off for inclusion in
a multi-group CFA and subsequent measurement invari-
ance testing using the registered one-factor model.
To examine the poor fit of the registered one-factor
model, individual item loadings were then assessed within
each country. The item loadings varied across and within
countries with values ranging between 0.67–0.86 in Ger-
many, 0.66–0.82 in Italy, 0.64–0.84 in the Netherlands,
0.71–0.86 in Slovenia, 0.68–0.91 in Spain, 0.69–0.89 in
the UK, and 0.74–0.91 in the US. Across all countries, apart
from the UK (where it was the second lowest), Item 5
(“Confidence regarding financial emergency”)hadthelow-
est loading. Across four countries (Italy, Spain, the UK, and
the US) the four lowest loading items were Items 4–7.Inthe
Netherlands, the three lowest-loading items were Items 5–7,
and in Slovenia, the two lowest-loading items were Items 5–
6. These items (Items 4–7inclusive) appear to load poorly
to the one-factor model across countries. While this does
not mean that these items necessarily load to a second
factor, they have previously been suggested to load to a
Table 2.Model fit cut-off scores
Item CFI TLI RMSEA SRMR αand ω
Excellent > 0.95 > 0.95 < .06 < .06 > 0.90
Good > 0.92 > 0.92 < .07 < .07 > 0.85
Acceptable > 0.89 > 0.89 < .08 < .08 > 0.80
Questionable > 0.86 > 0.86 < .09 < .09 > 0.75
Poor > 0.83 > 0.83 < .10 < .10 > 0.70
Unacceptable < 0.83 < 0.83 > .10 > .10 < 0.70
Note. Model fit cut-off values adapted from Greiff and Allen (2018).
European Journal of Psychological Assessment (2022), 38(6), 476–486 Ó2022 Hogrefe Publishing
480 E. K. Buabang et al., Validation PFW Scale
https://econtent.hogrefe.com/doi/pdf/10.1027/1015-5759/a000750 - Eike K. Buabang <eike.buabang@tcd.ie> - Monday, December 12, 2022 6:08:33 AM - IP Address:134.226.106.223

subscale of the PFW: “Objective Financial Wellbeing”as
part of a two-factor model (Nielsen, 2010). However, the
two-factor model did not meet the eigenvalue cut-off
required to establish this subscale (Nielsen, 2010). This pat-
tern of results is discussed in more detail in the exploratory
analyses section below. All standardized individual item
loading per country countries are reported in Table 4.
Convergent Validity
To assess convergent validity, correlations between scores
on the PFW scale and PSS were calculated (see Figure 1).
Lower scores on the PFW scale indicate more financial
stress. Therefore, we expected a strong negative correlation
(> .50) between the PFW scale and PSS scores. In line
with our prediction, results indicated significantly high
negative correlations in the US (r=0.61,p<.001), in
Spain (r=0.51,p<.001), in Germany (r=0.52,
p<.001), and in Slovenia (r=0.52,p<.001). Results
further indicated significant moderate to high negative cor-
relations in the UK (r=0.47,p<.001), in the Netherlands
(r=0.43,p<.001), and in Italy (r=0.38,p<.001).
Compared to the US, the correlation was significantly
lower in the UK, (z=2.31,p=.021), in the Netherlands,
(z=2.65,p=.008),andinItaly,(z=3.16,p=.002).
The correlations in the other countries were not signifi-
cantly different from the US (all zs<1.44).
Discriminant Validity
Discriminant validity was assessed by examining correla-
tions with the five traits measured with the TIPI (see
Figure 1). We expected moderately negative correlations
(.30 to .50) between conscientiousness and PFW scores,
and moderately positive correlations (.30 to .50) between
neuroticism and PFW scores. We expected low correlations
for the remaining three traits.
Conscientiousness
In line with our prediction, results indicated significant
moderate to high positive correlations in Germany (r=
0.46,p<.001), and in the UK (r=0.35,p<.001). Results
further indicated significant small to moderate correlations
in the Netherlands (r=0.15,p=.036), in the US (r=0.28,
p<.001), and in Slovenia (r=0.16,p=.021). The correla-
tions were low, not significant, and positive in Spain
(r=0.07,p=.311), and negative in Italy (r=0.03,
p=.675).
Compared to the US, the correlation was significantly
lower in Spain (z=2.21,p=.027), and in Italy (z=3.29,
Table 3.Model fit indices for individual country CFAs based on one-factor model of the PFWS
95% CI 95% CI
Country Nw
2
df CFI TLI RMSEA SRMR αLL UL ωLL UL
Germany 205 75 20 0.95 0.93 0.12 0.04 0.92 0.91 0.94 0.93 0.91 0.94
Italy 213 95 20 0.93 0.90 0.13 0.05 0.91 0.89 0.93 0.91 0.89 0.93
Netherlands 207 90 20 0.93 0.90 0.13 0.05 0.91 0.89 0.92 0.91 0.89 0.93
Slovenia 212 84 20 0.95 0.93 0.12 0.04 0.93 0.91 0.94 0.93 0.91 0.94
Spain 203 144 20 0.90 0.87 0.18 0.06 0.93 0.91 0.94 0.93 0.91 0.94
UK 288 184 20 0.91 0.88 0.17 0.05 0.94 0.92 0.95 0.93 0.92 0.95
US 222 100 20 0.95 0.93 0.13 0.04 0.95 0.94 0.96 0.95 0.94 0.96
Note. Alpha and omega reported with bootstrapped 95% confidence intervals (CI). LL indicates the lower limit and UL the upper limit of these intervals.
Results are reported rounded to 2 decimal places, however, unrounded values are used to assess model fit indices.
Table 4.Factor loading per item per country
Item Germany Italy The Netherlands Slovenia Spain UK US
Level of financial stress today 0.69 0.82 0.75 0.80 0.91 0.84 0.89
Satisfaction with present financial situation 0.82 0.82 0.78 0.77 0.81 0.84 0.84
Feelings about current financial situation 0.86 0.80 0.84 0.85 0.88 0.89 0.91
Worry about monthly living expenses 0.81 0.72 0.79 0.82 0.76 0.69 0.80
Confidence regarding financial emergency 0.67 0.66 0.64 0.71 0.68 0.75 0.74
Can’t afford to go out 0.82 0.69 0.72 0.73 0.69 0.80 0.81
Paycheck to paycheck 0.82 0.76 0.71 0.82 0.77 0.81 0.82
Stress about finances in general 0.83 0.82 0.82 0.86 0.88 0.88 0.91
Note. Standardized factor loading per item of the PFW scale per country.
Ó2022 Hogrefe Publishing European Journal of Psychological Assessment (2022), 38(6), 476–486
E. K. Buabang et al., Validation PFW Scale 481
https://econtent.hogrefe.com/doi/pdf/10.1027/1015-5759/a000750 - Eike K. Buabang <eike.buabang@tcd.ie> - Monday, December 12, 2022 6:08:33 AM - IP Address:134.226.106.223

p=.001). Compared to the US, the correlation was signifi-
cantly higher in Germany (z=2.09,p=.036). The corre-
lations in the other countries were not significantly different
from the US (all zs<1.31).
Neuroticism
In line with our prediction, results indicated significant
moderate to high negative correlations in the Netherlands
(r=0.36,p<.001), in the US (r=0.38,p<.001),
and in Germany (r=0.33,p<.001). Results further indi-
cated significant low to moderate negative correlations in
Slovenia (r=0.28,p<.001), in Italy (r=0.21,p=
.002), in the UK (r=0.29,p<.001), and in Spain (r=
0.21,p=.003), No correlations were significantly differ-
ent from the US (all zs<1.94).
Agreeableness, Openness, and Extraversion
Results indicated no significant correlations in any country
for agreeableness or openness. Contrary to our prediction,
results indicated a significant low to a moderate positive
correlation between the PFW and extraversion in Germany
(r=0.20,p=.005). Results indicated no further significant
correlations with extraversion in any other country.
No correlations were significantly different from the US
(for agreeableness all zs<2.08; for openness all zs<
0.79 and for extraversion all zs<1.47).
Exploratory Analyses
Due to the poor fit of the registered one-factor CFA, we
conducted further exploratory analyses. Nielsen (2010)pro-
posed a two-factor model structure (Factor 1:Subjective
Financial Well-being, Items 1,2,3,8;Factor2:Objective
Financial Well-being, Items 4–7inclusive). A two-factor
model would suggest that the two subscales measure differ-
ent underlying constructs. Here, we found a pattern of item
loading that is in line with these subscales, however, all
items still rely on a subjective assessment of financial stress.
Thus, a one-factor structure seems theoretically more
appropriate. Yet, there does appear to be a difference
between the items of the two proposed subscales. The items
of the subjective subscale assess affective financial stress
(e.g., feelings about the financial situation) whereas items
of the objective subscale assess behavioral financial stress
(e.g., living paycheck to paycheck). This suggests a multi-
dimensionality in which all items measure subjective finan-
cial stress but may also contain further internal commonal-
ities not captured by the one-factor structure. This could
lead to measurement-related variance when using a strict
one-factor model. Based on this idea, we explored a modi-
fied one-factor model which allowed for correlated errors of
items assessing affective and behavioral financial stress. All
modifications were made concurrently. Stepwise modifica-
tions based on influential item relations were not explored
because they may be dependent on the data of the current
sample. The results of this modified one-factor CFA are
reported in Table 5. For the modified one-factor model,
the CFI was excellent (>0.95) in all countries. The TLI
was excellent (> 0.95) in all countries except Italy where
it was acceptable (> 0.89). The RSMEA values are
improved across all countries with the modified one-factor
model. The RMSEA was excellent (< 0.06) in Slovenia, the
UK, and the US, it was good (< 0.07)inSpain,questionable
(< 0.09) in Italy, poor (< 0.10) in Germany, and unaccept-
able (> 0.10) in the Netherlands. SRMR was excellent
(< 0.06) in all countries. The item loadings for the modified
one-factor model are reported, in full, in Table 6.
Based on the model fit indices for the modified model,
four countries were included in a multi-group CFA analysis
to test the measurement invariance of the modified one-
factor model across countries. The RMSEA values were
unacceptably high in the Netherlands and were question-
able in Germany and Italy, and thus these countries were
excluded. The fit of the multi-group CFA and measurement
invariance of this model was then assessed, see Table 7.
Based on a cut-off of ΔCFI 0.01 the model achieved met-
ric invariance. This suggests that the factor structure and
loadings, but not the item functions, were similar across
the four countries included in the analysis. The achieved
level of invariance needs to be taken with caution, however,
because they are based on a model that was defined post
hoc. Additional modifications were not explored as the
changes consistent with the theoretical model had already
been made. Further changes to achieve partial scalar invari-
ance would likely depend on the sample data, not the
underlying model.
While overall the modifications improved the fit of the
model, the effect was weaker in the three countries not
included in the multi-group CFA. The “all-in”approach to
model modification may have failed to appreciate the
nuance of individual item functions in individual countries.
Therefore, within the three countries not included in the
Table 5.Model fit indices for individual country CFAs for modified
one-factor model of the PFWS
Country Nw
2
df CFI TLI RMSEA SRMR
Germany 205 22.46 8 0.99 0.96 0.09 0.03
Italy 213 21.76 8 0.99 0.95 0.09 0.02
The Netherlands 207 37.06 8 0.97 0.90 0.13 0.03
Slovenia 212 13.62 8 1.00 0.98 0.06 0.01
Spain 203 14.09 8 1.00 0.98 0.06 0.02
UK 288 11.82 8 1.00 0.99 0.04 0.01
US 222 12.55 8 1.00 0.99 0.05 0.01
Note. Results are reported rounded to 2 decimal places, however,
unrounded values are used to assess model fit indices.
European Journal of Psychological Assessment (2022), 38(6), 476–486 Ó2022 Hogrefe Publishing
482 E. K. Buabang et al., Validation PFW Scale
https://econtent.hogrefe.com/doi/pdf/10.1027/1015-5759/a000750 - Eike K. Buabang <eike.buabang@tcd.ie> - Monday, December 12, 2022 6:08:33 AM - IP Address:134.226.106.223

multi-group CFA, we then explored the individual item
functions to better understand the poorer fit of the modified
model.
In Germany, differential item function analysis (DIF; see
Figure 4in Supplementary Materials, https://osf.io/gevmr/)
indicated that Item 1functioned differently compared with
the other country samples. This may be due to a linguistic
variation in the translation of that item which did not
directly use the word “stress”. The translation may have
created a comparatively less sensitive measurement, as
flagged by the DIF which indicated that the item may not
differentiate between individuals well in the German sam-
ple. Furthermore, Germany and the Netherlands showed
relatively low loading for Items 1and 5(see Table 6). For
Item 5, these differences may relate to a broad cultural dif-
ference in comfort reporting on items that reference explicit
amounts of money which were present in this item only. In
the Netherlands and Italy, examination of the standardized
residuals showed that multiple items held stronger, or
weaker, relations than predicted by the modified one-factor
model (see Figure 5in Supplementary Materials, https://
osf.io/gevmr/).
Yet the pattern of deviations was not mirrored in each
country. This further suggests that the modifications were
not equally influential in all countries. However, the current
modifications were theoretically coherent. Therefore, alter-
native modification pathways within individual countries
were not explored to prevent overfitting specific country
data.
Discussion
Financial stress is a global issue with clear negative conse-
quences for health (Kahn & Pearlin, 2006) and well-being
(Haushofer & Fehr, 2014). Therefore, it is crucial to have
measures that accurately capture individuals’experiences
of financial stress. In the current study, we aimed to test
the validity and measurement invariance of the PFW scale
across six European countries (Germany, Italy, the Nether-
lands, Slovenia, Spain, and UK) and the US. To this end, the
scale was translated into five languages (German, Italian,
Dutch, Slovenian, and Spanish). A one-factor structure
was tested by conducting separate CFAs in each of the
countries. Results were mixed, with some fit indices indicat-
ing an acceptable to the excellent fit, and others indicating
a questionable to poor fit of the one-factor model. Reliabil-
ity measures were excellent and overall, the scale showed
convergent and discriminant validity.
Given the mixed fit indices for the one-factor structure,
we did not conduct the planned multi-group CFA. Instead,
we explored a modified one-factor structure. The modifica-
tions were based on the work by Nielsen (2010)whoorig-
inally suggested a two-factor structure with two latent
factors, namely subjective and objective financial stress.
However, we decided against the two-factor structure for
theoretical reasons. All items of the PFW scale rely on a
subjective assessment of financial stress and thus there is
no substantive support for a two-factor structure. It is also
worth noting that the eigenvalue of the objective factor
Table 6.Factor loading per item per country: Modified one-factor model
Item Germany Italy The Netherlands Slovenia Spain UK US
Level of financial stress today 0.69 0.75 0.67 0.73 0.83 0.76 0.83
Satisfaction with present financial situation 0.76 0.80 0.73 0.73 0.78 0.76 0.82
Feelings about current financial situation 0.80 0.70 0.72 0.80 0.78 0.80 0.85
Worry about monthly living expenses 0.85 0.72 0.84 0.85 0.79 0.72 0.82
Confidence regarding financial emergency 0.66 0.71 0.67 0.70 0.74 0.77 0.77
Can’t afford to go out 0.83 0.71 0.78 0.75 0.72 0.84 0.84
Paycheck to paycheck 0.81 0.78 0.75 0.83 0.82 0.85 0.85
Stress about finances in general 0.80 0.73 0.75 0.83 0.76 0.80 0.85
Note. Standardized factor loading per item of the PFW scale per country using the modified one-factor model.
Table 7.Measurement invariance for modified one-factor model
Testing level Nw
2
df CFI ΔCFI TLI RMSEA SRMR
Configural Invariance 907 52.45 32 1.00 0.99 0.05 0.01
Metric Invariance 907 85.50 53 0.99 0.00 0.99 0.05 0.04
Scalar Invariance 907 181.11 74 0.98 0.01 0.97 0.08 0.05
Note. Configural invariance can be assumed based on acceptable model fit. The change in CFI indicates that metric, but not scalar invariance is achieved as
the cut-off of 0.01 is surpassed at the scalar level. Results are reported rounded to 2 decimal places, however, unrounded values are used to assess model
fit indices.
Ó2022 Hogrefe Publishing European Journal of Psychological Assessment (2022), 38(6), 476–486
E. K. Buabang et al., Validation PFW Scale 483
https://econtent.hogrefe.com/doi/pdf/10.1027/1015-5759/a000750 - Eike K. Buabang <eike.buabang@tcd.ie> - Monday, December 12, 2022 6:08:33 AM - IP Address:134.226.106.223

did not exceed the cut-off value of 1in the study by Nielsen
(2010). At the same time, there is a measurement-related
difference in the items of the two proposed subscales as
items of the subjective subscale assess affective financial
stress (e.g., feelings about the financial situation) whereas
items of the objective subscale assess behavioral financial
stress (e.g., living paycheck to paycheck). Therefore, we
decided to add error correlations for the items assessing
behavioral and affective financial stress respectively.
Results indicated an improved fit for the modified one fac-
tor-structure in all countries. Considering the improved fit
of the modified one-factor model, the PFW scale appears
to be a promising tool to provide insight into the experience
of financial stress within each country.
To assess if the PFW scale can be used to make compar-
isons between countries, we conducted a multi-group CFA
testing the modified one-factor structure across Slovenia,
Spain, the UK, and the US. Germany, Italy, and the Nether-
lands were excluded because the fit indices were still con-
flicted. Testing for measurement invariance, the results
indicated metric invariance was achieved. This suggests
an equal factor structure and equal loadings in the four
countries. However, scalar and strict invariance were not
reached. Considering that we had already explored a mod-
ified model, albeit, with modifications based on theoretical
reasons, we did not explore further changes to achieve
strict invariance based on partial invariance models. Based
onthelackofscalarinvarianceweconcludethatitisnot
advisable to use the scale to make comparisons between
countries.
To understand why these comparisons are not tenable
we investigated the factor loadings and residuals more clo-
sely. We noted that individual items may function differ-
ently in certain countries due to translational or cultural
reasons. For example, Item 1functioned differently in our
German sample which may relate to linguistic differences
in our translation. In turn, Item 5functioned poorly in the
Netherlands and Germany, two countries that are culturally
quite similar in our sample. This difference may reflect a
cultural hesitancy to respond to items relating to specific
amounts of money, as required by that item. The variable
factor loading of Item 5may also relate to the study’sdata
collection period which occurred during the COVID-19
pandemic. Item 5relates to financial emergencies which
were particularly salient during the global pandemic.
Indeed, comparisons between countries in our samples
may be hindered by the fact that different countries
responded to the global pandemic with different financial
support strategies. This may have influenced our country
samples in ways we cannot account for and constrains
the comparisons that can be made. Variation in our sam-
pling proportions across countries, being either Prolific or
Snowball samples, may have also influenced our results
with greater proportions of Prolific data appearing to relate
to a higher likelihood of poor model fit. However, the
underlying reasons for this pattern are not clear.
A noteworthy finding of the current study is the results of
the discriminant validity analysis. Based on previous work
(Xu et al., 2015), we predicted moderate to high correla-
tions between PFW scores and conscientiousness and neu-
roticism, and low correlations for the other traits. Results
were largely in line with our predictions and mostly coun-
try-invariant. However, for conscientiousness, we observed
significant variation in the correlations which were high in
Germany and the UK, low to moderate in the Netherlands,
the US, and Slovenia, and low and nonsignificant in Spain
and Italy. This suggests that in Germany and the UK in par-
ticular conscientiousness is related to financial stress. A
possible explanation is that there are local structural or pol-
icy-related differences between these countries, which are
affected by conscientiousness. One such difference could
be the administrative burden that individuals face when
they are trying to obtain financial benefits (Dubois &
Ludwinek, 2015). When obtaining financial support is bur-
densome, higher levels of conscientiousness may aid indi-
viduals. This could have practical implications as access
to financial benefits should not rely on personal administra-
tive effort. However, further research in which personality
and financial stress are assessed in a cross-cultural context
is needed.
Cross-border comparisons can shed light on cultural vari-
ability within the domain of financial well-being. Previous
work has shown that financial stress can be affected by
cultural differences (Warmath et al., 2021). Indeed, in the
current study, we found that the relationship between con-
scientiousness and financial stress showed significant vari-
ation between countries, with moderate to strong negative
correlations in some countries and no significant correla-
tions in others. However, when considering policy implica-
tions, understanding variability within countries may be
more informative. Reducing financial stress and increasing
well-being through effective policies requires a nuanced
approach and interventions should be tailored to specific
groups (Ruggeri et al., 2020). Therefore, using the PFW
scale to explore differences within countries is a fruitful
area for future research.
In conclusion, the PFW scale shows excellent reliability,
and convergent and discriminant validity. The current
results suggest that the scale can be used to explore specific
differences within countries. However, comparisons
between countries using the PFW scale are not advised.
Understanding the subjective experience of financial stress
has practical applications in terms of understanding and
implementing effective interventions. The PFW scale pro-
vides insight into this domain and its use may benefit both
research and practice on financial stress and well-being.
European Journal of Psychological Assessment (2022), 38(6), 476–486 Ó2022 Hogrefe Publishing
484 E. K. Buabang et al., Validation PFW Scale
https://econtent.hogrefe.com/doi/pdf/10.1027/1015-5759/a000750 - Eike K. Buabang <eike.buabang@tcd.ie> - Monday, December 12, 2022 6:08:33 AM - IP Address:134.226.106.223

References
American Psychological Association. (2021, October). Stress in
America™2021: Stress and Decision-making during the Pandemic.
https://www.apa.org/news/press/releases/stress/2021/decision-
making-october-2021.pdf
Aust, F., & Barth, M. (2020). papaja: Prepare reproducible APA
journal articles with R Markdown. https://github.com/crsh/
papaja
Buabang, E. K., Ashcroft-Jones, S., Esteban Serna, C., Kastelic, K.,
Kveder, J., Lambertus, A., Lambertus, A., Muller, T. S., &
Ruggeri, K. (2022). Validation and measurement invariance of
the Personal Financial Wellness Scale: A multinational study in
7 countries. https://osf.io/gevmr
Blanchet, T., Chancel, L., & Gethin, A. (2020). Why is Europe more
equal than the United States. https://halshs.archives-ouvertes.
fr/halshs-03022133
Choi, S. W., Gibbons, L. E., with contributions from., & Crane, P. K.
(2016). Lordif: Logistic ordinal regression differential item
functioning using IRT. https://CRAN.R-project.org/package=
lordif
Christelis, D., Ehrmann, M., & Georgarakos, D. (2017). Exploring
differences in household debt across the United States and
Euro Area countries (No. 465). Centre for Studies in Economics;
Finance (CSEF), University of Naples. https://ideas.repec.org/p/
sef/csefwp/465.html
Cohen, S. (1988). Perceived stress in a probability sample of the
United States. In S. Spacapan & S. Oskamp (Eds.), The social
psychology of health (pp. 31–67). Sage Publications, Inc.
Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global
measure of perceived stress. Journal of Health and Social
Behavior, 24(4), 385–396. https://doi.org/10.2307/2136404
Crutzen, R., & Peters, G.-J. Y. (2017). Scale quality: Alpha is an
inadequate estimate and factor-analytic evidence is needed
first of all. Health Psychology Review, 11(3), 242–247. https://
doi.org/10.1080/17437199.2015.1124240
Diedenhofen,B.,&Musch,J.(2015).Cocor: A comprehensive solution
for the statistical comparison of correlations. PLoS One, 10(4),
Article e0121945. https://doi.org/10.1371/journal.pone.0121945
Dubois, H., & Ludwinek, A. (2015). Access to social benefits:
Reducing non take up. Publications Office of the European
Union. https://data.europa.eu/doi/10.2806/7077
Eurostat. (2021). Living conditions in Europe –Poverty and social
exclusion. https://ec.europa.eu/eurostat/statistics-explained/
index.php?title=Living_conditions_in_Europe_-_poverty_and_
social_exclusion
Fazlić, H., Brborović, O., VukušićRukavina, T., Fišter, K., Milošević,
M., & Mustajbegović, J. (2012). Characteristics of people with
the perceived stress in Croatia: the CroHort study. Collegium
Antropologicum, 36(1), 165–169.
Fischer, R., & Karl, J. A. (2019). A primer to (cross-cultural) multi-
group invariance testing possibilities in R. Frontiers in Psychol-
ogy, 10, Article 1507. https://doi.org/10.3389/fpsyg.2019.01507
Gosling, S. D., Rentfrow, P. J., & Swann, W. B. (2003). A very brief
measure of the Big-Five personality domains. Journal of
Research in Personality, 37(6), 504–528. https://doi.org/
10.1016/S0092-6566(03)00046-1
Greiff, S., & Allen, M. S. (2018). EJPA introduces registered reports
as new submission format. European Journal of Psychological
Assessment, 34(4), 217–219. https://doi.org/10.1027/1015-
5759/a000492
Gutter, M., & Copur, Z. (2011). Financial behaviors and financial
well-being of college students: Evidence from a national
survey. Journal of Family and Economic Issues, 32(4), 699–
714. https://doi.org/10.1007/s10834-011-9255-2
Hagenaars, A. J. M., & van Praag, B. M. S. (1985). A synthesis of
poverty line definitions. Review of Income and Wealth, 31(2),
139–154. https://doi.org/10.1111/j.1475-4991.1985.tb00504.x
Harrell, F. E. Jr. (2022). Hmisc: Harrell miscellaneous. https://
CRAN.R-project.org/package=Hmisc
Haushofer, J., & Fehr, E. (2014). On the psychology of poverty. Science,
344(6186), 862–867. https://doi.org/10.1126/science. 1232491
Hofstede, G. H., Hofstede, G. J., & Minkov, M. (2010). Cultures and
organizations: Software of the mind: Intercultural cooperation
and its importance for survival (3rd ed.). McGraw-Hill.
Jorgensen, T. D., Pornprasertmanit, S., Schoemann, A. M., &
Rosseel, Y. (2021). semTools: Useful tools for structural equa-
tion modeling. https://CRAN.R-project.org/package=semTools
Kahn, J. R., & Pearlin, L. I. (2006). Financial strain over the life
course and health among older adults. Journal of Health and
Social Behavior, 47(1), 17–31. https://doi.org/10.1177/
002214650604700102
Kamaluddin, M. R., Nasir, R., Sulaiman, W. S. W., Hafidz, S. W. M.,
Abdullah, J. M. A., Khairudin, R., & Zamani, Z. A. (2018). Validity
and reliability of Malay version Financial Well-Being Scale
among Malaysian employees. AKADEMIKA, 88(2), 109–120.
http://ejournal.ukm.my/akademika/issue/view/1064
Lee, E.-H. (2012). Review of the psychometric evidence of the
perceived stress scale. Asian Nursing Research, 6(4), 121–127.
https://doi.org/10.1016/j.anr.2012.08.004
Nielsen, R. B. (2010). Assessing financial wellness via computer-
assisted telephone interviews (SSRN Scholarly Paper No.
2061273). Social Science Research Network. https://papers.
ssrn.com/abstract=2061273
Olivoto, T., & L’ucio, A. D. (2020). Metan: An r package for multi-
environment trial analysis. Methods in Ecology and Evolution,
11(6), 783–789. https://doi.org/10.1111/2041-210X.13384
Peters, G.-J. Y. (2014). The alpha and the omega of scale reliability
and validity: Why and how to abandon Cronbach’s alpha and
the route towards more comprehensive assessment of scale
quality. European Health Psychologist, 16(2), 56–69. https://
doi.org/10.17605/OSF.IO/TNRXV
Peters, G.-J. Y. (2017). Diamond plots: A tutorial to introduce a
visualisation tool that facilitates interpretation and comparison
of multiple sample estimates while respecting their inaccuracy.
PsyArXiv. https://psyarxiv.com/fzh6c
Prawitz, A. D., Garman, E. T., Sorhaindo, B., O’Neill, B., Kim, J., &
Drentea, P. (2006). Incharge Financial Distress/Financial Well-
Being scale: Development, administration, and score interpre-
tation. Financial Counseling and Planning, 17(1), 34–50. https://
doi.org/10.1037/t60365-000
R Core Team. (2022). R: A language and environment for statistical
computing. R Foundation for Statistical Computing. https://
www.R-project.org/
Ravallion, M. (2012). Poor, or just feeling poor? On using subjective
data in measuring poverty. The World Bank. https://doi.org/
10.1596/1813–9450-5968
Ravallion, M., & Lokshin, M. (2002). Self-rated economic welfare in
Russia. European Economic Review, 46(8), 1453–1473. https://
doi.org/10.1016/S0014-2921(01)00151-9
Rosseel, Y. (2012). lavaan: An R package for structural equation
modeling. Journal of Statistical Software, 48(2), 1–36. https://
doi.org/10.18637/jss.v048.i02
Ruggeri, K., Garcia-Garzon, E., Maguire, Á., Matz, S., & Huppert,
F. A. (2020). Well-being is more than happiness and life
satisfaction: A multidimensional analysis of 21 countries.
Health and Quality of Life Outcomes, 18(1), Article 192.
https://doi.org/10.1186/s12955-020-01423-y
Shapiro, G. K., & Burchell, B. J. (2012). Measuring financial
anxiety. Journal of Neuroscience, Psychology, and Economics,
5(2), 92–103. https://doi.org/10.1037/a0027647
Ó2022 Hogrefe Publishing European Journal of Psychological Assessment (2022), 38(6), 476–486
E. K. Buabang et al., Validation PFW Scale 485
https://econtent.hogrefe.com/doi/pdf/10.1027/1015-5759/a000750 - Eike K. Buabang <eike.buabang@tcd.ie> - Monday, December 12, 2022 6:08:33 AM - IP Address:134.226.106.223

Sommet, N., Morselli, D., & Spini, D. (2018). Income inequality
affects the psychological health of only the people facing
scarcity. Psychological Science, 29(12), 1911–1921. https://doi.
org/10.1177/0956797618798620
Tierney, N. (2017). Visdat: Visualising whole data frames. Journal
of Open Source Software, 2(16), Article 355. https://doi.org/
10.21105/joss.00355
Vallejo, M. A., Vallejo-Slocker, L., Fernández-Abascal, E. G., &
Mañanes, G. (2018). Determining factors for stress perception
assessed with the Perceived Stress Scale (PSS-4) in Spanish
and other European samples. Frontiers in Psychology, 9, Article
37. https://doi.org/10.3389/fpsyg.2018.00037
Warmath, D., Chen, P.-J., Grable, J., & Kwak, E. J. (2021). Soft
landings: Extending the cushion hypothesis to financial well-
being in collectivistic cultures. Journal of Consumer Affairs,
55(4), 1563–1590. https://doi.org/10.1111/joca.12408
Wickham, H., Averick, M., Bryan, J., Chang, W., McGowan, L. D.,
François, R., Grolemund, G., Hayes, A., Henry, L., Hester, J.,
Kuhn, M., Pedersen, T. L., Miller, E., Bache, S. M., Müller, K.,
Ooms, J., Robinson, D., Seidel, D. P., Spinu, V., ... Yutani, H.
(2019). Welcome to the tidyverse. Journal of Open Source
Software, 4(43), Article 1686. https://doi.org/10.21105/
joss.01686
Wickham, H., François, R., Henry, L., & Müller, K. (2022). Dplyr: A
grammar of data manipulation. https://CRAN.R-project.org/
package=dplyr
Xie, Y. (2015). Dynamic documents with R and knitr (2nd ed.).
Chapman & Hall/CRC. https://yihui.org/knitr/
Xu, Y., Beller, A. H., Roberts, B. W., & Brown, J. R. (2015).
Personality and young adult financial distress. Journal of
Economic Psychology, 51,90–100. https://doi.org/10.1016/
j.joep.2015.08.010
Zhu, H. (2021). kableExtra: Construct complex table with “kable”and
pipe syntax. https://CRAN.R-project.org/package=kableExtra
History
Received June 24, 2019
Revision received October 23, 2022
Accepted October 24, 2022
Published online December 8, 2022
EJPA Section / Category Miscellaneous / Other
Authorship
Eike K. Buabang, conceptualization, supervision, writing –original
draft, investigation, data curation, formal analysis, visualization,
writing –review & editing; Sarah Ashcroft-Jones, writing –original
draft, investigation, formal analysis, visualization, writing –review
& editing; Celia Esteban Serna, writing –original draft, investiga-
tion, resources; Katarina Kastelic, writing –original draft, inves-
tigation, resources, writing –review & editing; Jakob Kveder,
writing –original draft, investigation, resources; Amanda Lam-
bertus, writing –original draft, investigation; Tasja S. Müller,
writing –original draft, investigation, resources; Kai Ruggeri,
conceptualization, funding acquisition, supervision, writing –
review & editing.
Open Science
We report how we determined our sample size, all data exclu-
sions, all data inclusion/exclusion criteria, whether inclusion/
exclusion criteria were established prior to data analysis, all
measures in the study, and all analyses including all tested
models. If we use inferential tests, we report exact pvalues, effect
sizes, and 95% confidence or credible intervals.
Open Data: I confirm that there is sufficient information for an
independent researcher to reproduce all of the reported results
(Buabang et al., 2022).
Open Materials: I confirm that there is sufficient information for
an independent researcher to reproduce all of the reported
methodology (Buabang et al., 2022).
Preregistration of Studies and Analysis Plans: This study was
preregistered with an analysis plan (Buabang et al., 2022).
The Stage-1 report, materials, data, code, and supplementary
materials are available on the Open Science Framework: https://
osf.io/gevmr/.
Funding
Funding for data collection was provided through Columbia
University Mailman School of Public Health, Department of Health
Policy and Management.
ORCID
Eike K. Buabang
https://orcid.org/0000-0002-3057-0819
Eike K. Buabang
Trinity College Institute of Neuroscience
Trinity College Dublin
Dublin 2
Ireland
eike.buabang@tcd.ie
European Journal of Psychological Assessment (2022), 38(6), 476–486 Ó2022 Hogrefe Publishing
486 E. K. Buabang et al., Validation PFW Scale
https://econtent.hogrefe.com/doi/pdf/10.1027/1015-5759/a000750 - Eike K. Buabang <eike.buabang@tcd.ie> - Monday, December 12, 2022 6:08:33 AM - IP Address:134.226.106.223