Data from an International Multi-Centre Study of Statistics and Mathematics Anxieties and Related Variables in University Students (the SMARVUS Dataset)

Abstract

This large, international dataset contains survey responses from N = 12,570 students from 100 universities in 35 countries, collected in 21 languages. We measured anxieties (statistics, mathematics, test, trait, social interaction, performance, creativity, intolerance of uncertainty, and fear of negative evaluation), self-efficacy, persistence, and the cognitive reflection test, and collected demographics, previous mathematics grades, self-reported and official statistics grades, and statistics module details. Data reuse potential is broad, including testing links between anxieties and statistics/mathematics education factors, and examining instruments’ psychometric properties across different languages and contexts. Data and metadata are stored on the Open Science Framework website (https://osf.io/mhg94/).

Full text

Title
Data from an International Multi-Centre Study of Statistics and Mathematics Anxieties and
Related Variables in University Students (the SMARVUS Dataset)
Authors
Jenny Terry1†, Robert M. Ross2, Tamas Nagy3, Mauricio Salgado4,5, Patricia Garrido-
Vásquez6, Jacob O. Sarfo7, Susan Cooper8, Anke C. Buttner9, Tiago J. S. Lima10, İbrahim
Öztürk11, Nazlı Akay11, Flavia H. Santos12, Christina Artemenko13, Lee T. Copping14,
Mahmoud M. Elsherif15, Ilija Milovanović16, Robert A. Cribbie17,17, Marina G. Drushlyak18,
Katherine Swainston19, Yiyun Shou20, Juan David Leongómez21, Nicola Palena22, Fitri A.
Abidin23,24, Maria F. Reyes-Rodríguez25, Yunfeng He26,13, Juneman Abraham27, Argiro
Vatakis28, Kristin Jankowsky29, Stephanie N. L. Schmidt30, Elise Grimm31, Desirée
González32, Philipp Schmid33,34, Roberto A. Ferreira35, Dmitri Rozgonjuk36,37, Neslihan
Özhan38, Patrick A. O’Connor39, Andras N. Zsido40, Gregor Stiglic41,42, Darren Rhodes43,
Cristina Rodríguez35, Ivan Ropovik44,45, Violeta Enea46, Ratri Nurwanti47, Alejandro J.
Estudillo48,49, Nataly Beribisky17, Karel K. Himawan50,51, Linda M. Geven52,53, Anne H. van
Hoogmoed54, Amélie Bret55, Jodie E. Chapman56, Udi Alter57,17, Tessa R. Flack58, Donncha
Hanna39, Mojtaba Soltanlou59,60, Gabriel Banik61, Matúš Adamkovič62,63, Sanne H. G. van der
Ven54, Jochen A. Mosbacher64, Hilal H. Şen65,66, Joel R. Anderson56, Michael Batashvili67,
Kristel de Groot68,69, Matthew O. Parker70, Mai Helmy71,72, Mariia M. Ostroha73, Katie A.
Gilligan-Lee59, Felix O. Egara74, Martin J. Barwood75, Karuna Thomas76, Grace McMahon77,
Siobhán M. Griffin77, Hans-Christoph Nuerk13, Alyssa Counsell57, Oliver Lindemann68, Dirk
Van Rooy78,79, Theresa E. Wege80, Joanna E. Lewis81, Balazs Aczel3, Conal Monaghan78, Ali
H. Al-Hoorie82, Julia F. Huber13, Saadet Yapan83, Mauricio E. Garrido Vásquez6, Antonino
Callea84, Tolga Ergiyen85, James M. Clay86, Gaetan Mertens87, Feyza Topçu83, Merve G.
Tutlu83, Karin Täht36,88, Kristel Mikkor36, Letizia Caso84, Alexander Karner89, Maxine M. C.

Storm68, Gabriella Daroczy13, Rizqy A. Zein90, Andrea Greco22, Erin M. Buchanan91,
Katharina Schmid92, Thomas E. Hunt93, Jonas De keersmaecker94, Peter E. Branney95,
Jordan Randell96, Oliver J. Clark97, Crystal N. Steltenpohl98,99, Bhasker Malu100, Burcu
Tekeş101, TamilSelvan Ramis102, Stefan Agrigoroaei31, Nicholas A. Badcock103,104, Kareena
McAloney-Kocaman105, Olena V. Semenikhina106, Erich W. Graf38, Charlie Lea107, Fergus M.
Guppy108,109, Amy C. Warhurst110, Shane Lindsay111, Ahmed Al Khateeb82, Frank
Scharnowski89,112, Leontien de Kwaadsteniet54, Kathryn B. Francis113, Mariah Lecompte57,
Lisa A. D. Webster114, Kinga Morsanyi80, Suzanna E. Forwood115, Elizabeth R. Walters95,
Linda K. Tip107, Jordan R. Wagge116, Ho Yan Lai76, Deborah S. Crossland96, Kohinoor M.
Darda117, Zoe M. Flack107, Zoe Leviston78,118, Matthew Brolly119, Samuel P. Hills120, Elizabeth
Collins121, Andrew J. Roberts122, Wing-Yee Cheung96, Sophie Leonard12, Bruno
Verschuere53, Samantha K. Stanley78, Iro Xenidou-Dervou80, Omid Ghasemi104, Timothy
Liew76, Daniel Ansari123, Johnrev Guilaran124, Samuel G. Penny108,125, Julia Bahnmueller80,
Christopher J. Hand126, Unita W. Rahajeng47, Dar Peterburg127, Zsofia K. Takacs128, Michael
J. Platow78, Andy P. Field1
1School of Psychology, University of Sussex, UK, 2Department of Psychology, Macquarie
University, Australia, 3Institute of Psychology, ELTE Eötvös Loránd University, Hungary,
4School of Social Sciences, Universidad Andres Bello, Chile, 5Centre for Research in
Inclusive Education, Universidad Andres Bello, Chile, 6Department of Psychology, University
of Concepción, Chile, 7Department of Health, Physical Education, and Recreation, University
of Cape Coast, Ghana, 8Department of Psychology, Kingston University, UK, 9School of
Psychology, University of Birmingham, UK, 10Department of Social and Work Psychology,
University of Brasília, Brazil, 11Department of Psychology, Middle East Technical University,
Turkey, 12School of Psychology, University College Dublin, Ireland, 13Department of
Psychology, University of Tuebingen, Germany, 14Department of Psychology, Teesside
University, UK, 15School of Psychology, University of Birmingham, 16Department of
Psychology, Faculty of Philosophy, University of Novi Sad, Serbia, 17Department of

Psychology, York University, Canada, 18Mathematics Department, Makarenko Sumy State
Pedagogical University, Ukraine, 19School of Psychology, Newcastle University, UK,
20Research School of Psychology, Australian National University. Australia, 21Faculty of
Psychology, Universidad El Bosque, Colombia, 22Department of Human and Social
Sciences, University of Bergamo, Italy, 23Faculty of Psychology, Universitas Padjadjaran,
Indonesia, 24Center for Innovation and Psychological Research, Faculty of Pychology,
Universitas Padjadjaran, Indonesia, 25Department of Psychology, Universidad de los Andes,
Colombia, 26Liaoning Key Laboratory of Psychological Testing and Behavior Analysis,
Liaoning Univeristy, China, 27Department of Psychology, Faculty of Humanities, Bina
Nusantara University, Indonesia, 28Department of Psychology, Panteion University of Social
and Political Sciences, Greece, 29Psychological Assessment, University of Kassel, Germany,
30Department of Psychology, University of Konstanz, Germany, 31Psychological Sciences
Research Institute, UCLouvain, Belgium, 32Departamento de Didáctica e Investigación
Educativa, Universidad de la Laguna, Spain, 33Media and Communication Science,
University of Erfurt, Germany, 34Implementation Research, Bernhard-Nocht-Insitute for
Tropical Medicine, Germany, 35Facultad de Ciencias de la Educación, Universidad Católica
del Maule, Chile, 36Institute of Mathematics and Statistics, University of Tartu, Estonia,
37Department of Molecular Psychology, Ulm University, Germany, 38School of Psychology,
University of Southampton, UK, 39School of Psychology, Queen’s University Belfast, UK,
40Institute of Psychology, University of Pécs, Hungary, 41Faculty of Health Sciences,
University of Maribor, Slovenia, 42Usher Institute, University of Edinburgh, UK, 43NTU
Psychology, Nottingham Trent University, UK, 44Charles University, Faculty of Education,
Institute for Research and Development of Education, Czech Republic, 45Faculty of
Education, University of Presov, Slovakia, 46Department of Psychology, Alexandru Ioan
Cuza University, Romania, 47Department of Psychology, Brawijaya University, Indonesia,
48Department of Psychology, Bournemouth University, UK, 49School of Psychology,
University of Nottingham, Malaysia, 50Faculty of Psychology, Universitas Pelita Harapan,
Indonesia, 51RELASI Research Lab, Indonesia, 52Institute of Criminal Law and Criminology,

Leiden University, the Netherlands, 53Department of Clinical Psychology, University of
Amsterdam, the Netherlands, 54Radboud University, Behavioural Science Institute, the
Netherlands, 55LPPL, Nantes University, France, 56School of Psychology, Australian Catholic
University, Australia, 57Department of Psychology, Toronto Metropolitan University (formerly
Ryerson University), Canada, 58School of Psychology, University of Lincoln, UK, 59School of
Psychology, University of Surrey, UK, 60Brain and Mind Institute & Department of
Psychology, Western University, Canada, 61University of Presov, Slovakia, 62Institute of
Social Sciences, CSPS, Slovak Academy of Sciences, Slovakia, 63Faculty of Humanities and
Social Sciences, University of Jyväskylä, Finland, 64Institute of Psychology, University of
Graz, Austria, 65Department of Psychology, MEF University, Turkey, 66Faculty of Psychology,
University of Akureyri, Iceland, 67Baruch Ivcher School of Psychology, Reichman University
(formerly IDC Herzliya,, Israel, 68Department of Psychology, Erasmus School of Social and
Behavioural Sciences, Erasmus University Rotterdam, the Netherlands, 69Department of
Applied Economics, Erasmus School of Economics, Erasmus University Rotterdam, the
Netherlands, 70School of Pharmacy and Biomedical Science, University of Portsmouth, UK,
71Department of Psychology, College of Education, Sultan Qaboos University, Oman,
72Department of Psychology, Faculty of Arts, Menoufia University, Egypt, 73Computer
Science Department, Makarenko Sumy State Pedagogical University, Ukraine, 74Department
of Science Education, University of Nigeria, Nsukka, 75School of Health, Sport, and Life
Sciences, Leeds Trinity University, UK, 76Department of Psychology, HELP University,
Malaysia, 77Department of Psychology, University of Limerick, Ireland, 78Research School of
Psychology, Australian National University, Australia, 79Institute for Climate, Energy, and
Disaster Solutions, Australia., 80Centre for Mathematical Cognition, Loughborough
University, UK, 81University of Northern Colorado, USA, 82Independent Researcher, Saudi
Arabia, 83Department of Psychology, Hasan Kalyoncu University, Turkey, 84Department of
Human Sciences, Libera Università Maria SS. Assunta University, Italy, 85Izmir University of
Economics, Turkey, 86Department of Psychology, University of Portsmouth, UK,
87Department of Medical and Clinical Psychology, Tilburg University, the Netherlands,

88Institute of Psychology, University of Tartu, Estonia, 89Department of Cognition, Emotion
and Methods in Psychology, University of Vienna, Austria, 90Department of Psychology,
Universitas Airlangga, Indonesia, 91Harrisburg University of Science and Technology, USA,
92Universitat Ramon Llull, Esade Business School, Spain, 93School of Psychology, University
of Derby, UK, 94Ramon Llull University, Esade Business School, Spain, 95Department of
Psychology, University of Bradford, UK, 96Department of Psychology, University of
Winchester, UK, 97Department of Psychology, Manchester Metropolitan University, UK,
98University of Southern Indiana, USA, 99Dartmouth Center for Program Design and
Evaluation, 100O P Jindal Global University, Sonipat, India, 101Department of Psychology,
Başkent University, Turkey, 102Centre for American Education, Sunway University, Malaysia,
103School of Psychological Science, University of Western Australia, Australia, 104School of
Psychological Sciences, Macquarie University, Australia, 105Department of Psychology,
Glasgow Caledonian University, UK, 106Computer Science Depatment, Makarenko Sumy
State Pedagogical University, Ukraine, 107School of Humanities and Social Science,
University of Brighton, UK, 108School of Applied Sciences, University of Brighton, UK,
109School of Energy, Geoscience, Infrastructure and Society , Heriot-Watt University,
110School of Psychology, University of Winchester, UK, 111Department of Psychology,
University of Hull, UK, 112Psychiatric Hospital, University of Zürich, Switzerland, 113School of
Psychology, Keele University, UK, 114School of Psychology and Therapeutic Studies, Leeds
Trinity University, UK, 115School of Psychology & Sport Science, Anglia Ruskin University,
UK, 116School of Psychology and Cognitive Science, Avila University, USA, 117University of
Pennsylvania, USA, 118School of Arts & Humanities, Edith Cowan University, Perth, WA,
Australia, 119School of Applied Sciences, University of Brighton, 120Faculty of Health and
Social Sciences, Bournemouth University, UK, 121Division of Psychology, Faculty of Natural
Sciences, University of Stirling, UK, 122Department of Philosophy, Macquarie University,
Australia, 123Department of Psychology, Western University, Canada, 124Division of Social
Sciences, College of Arts and Sciences, University of the Philippines Visayas, Philippines,
125Bristol Zoological Society, Bristol, UK, 126School of Education, University of Glasgow, UK,

127Baruch Ivcher School of Psychology, Reichman University (formerly IDC Herzliya), Israel,
128School of Health in Social Science, University of Edinburgh, UK
†Correspondence should be addressed to Jenny Terry; E-mail: jlt26@sussex.ac.uk
Abstract
This large, international dataset contains survey responses from N = 12,570 students
from 100 universities in 35 countries, collected in 21 languages. We measured anxieties
(statistics, mathematics, test, trait, social interaction, performance, creativity, intolerance of
uncertainty, and fear of negative evaluation), self-efficacy, persistence, and the cognitive
reflection test, and collected demographics, previous mathematics grades, self-reported and
official statistics grades, and statistics module details. Data reuse potential is broad, including
testing links between anxieties and statistics/mathematics education factors, and examining
instruments’ psychometric properties across different languages and contexts. Data and
metadata are stored on the Open Science Framework website (https://osf.io/mhg94/).
Keywords
statistics, mathematics, anxiety, education, jangle fallacy

(1) Background
Many university students on non-mathematics-based degrees report feeling anxious
about learning mathematics and statistics (e.g., Field, 2014). Statistics anxiety was initially
assumed to be the same as mathematics anxiety, but many now consider it distinct (Chew &
Dillon, 2014). Statistics anxiety has been defined as "a negative state of emotional arousal
experienced by individuals as a result of encountering statistics in any form and at any level
[...] and is related to but distinct from mathematics anxiety" (Chew & Dillon, 2014, p. 199).
Mathematics anxiety is similarly defined as “a feeling of tension and anxiety that interferes
with the manipulation of numbers and the solving of mathematical problems in [...] ordinary
life and academic situations” (Richardson & Suinn, 1972). Neither definition is clear about
how these two constructs differ, and students may perceive them to be the same because
both mathematics and statistics involve the manipulation and interpretation of numerical
information. This conflation could be a shared root of students’ anxiety, rather than their
anxiety being specific to mathematics or statistics.
These definitions have informed the scales that measure these constructs (Baloğlu &
Zelhart, 2007; Cruise et al., 1985). For these scales to be valid, we need clarity about
whether they measure facets of anxiety specific to statistics/mathematics or reflect a
common numeric anxiety. In short, we must rule out a jangle fallacy, where two scales are
incorrectly assumed to measure different constructs (Kelley, 1927). Jangle fallacies can lead
to independently evolving theoretical literatures for each construct that should instead be
mutually informative (Block, 1995).
Few studies have tested the distinctiveness of statistics and mathematics anxiety
scales. Most concluded statistics anxiety is related to mathematics anxiety, but some
variance remains unaccounted for, suggesting a unique component (r = 0.41 to r = 0.67;
Baloğlu, 2002; Birenbaum & Eylath, 1994; Paechter et al., 2017; Zeidner, 1991). What this
unique component is remains unclear. It is possible the unexplained variance does not

reflect differences in statistics and mathematics anxieties, but differences in the scales’
dimensions. For example, because the Statistics Anxiety Rating Scale (STARS; Cruise et
al., 1985) includes a “Fear of asking for help” subscale and the Revised Maths Anxiety
Rating Scale (R-MARS; Baloğlu & Zelhart, 2007) does not, the unique variance may have
been driven by the fear of asking for help only captured by the STARS.
It is important to use a range of methods to study the constructs’ independence, such
as various confirmatory factor analysis techniques (Lawson & Robins, 2021), extrinsic
convergent validity analysis (Gonzalez et al., 2020), and multi-trait-multi-method designs
(Campbell & Fiske, 1959). However, previous studies that compared measures of
mathematics and statistics anxiety (e.g., Baloğlu, 2002; Paechter et al., 2017) have based
their conclusions on correlations, which are only one of the 10 criteria that can determine the
extent that two scales overlap (Lawson & Robins, 2021).
To address these concerns, Terry et al. (2022) explored these constructs’
distinctiveness in two samples of UK-based undergraduate psychology students (N = 465
and N = 245). They measured statistics anxiety with the STARS (Cruise et al., 1985) and
mathematics anxiety with the R-MARS (Baloğlu & Zelhart, 2007), and developed versions of
each scale modified to reflect the other construct (i.e., a mathematics version of the STARS
and a statistics version of the R-MARS). By doing so, Terry et al. (2022) created equivalent,
comparable subscales (e.g., there was now a mathematics version of the “Fear of asking for
help” subscale). Their results suggested a jangle fallacy. Specifically, the factor analyses
and latent profile analyses of the four measures, as well as their experimental studies, found
converging evidence that the scales were measuring the same construct.
However, construct validation work should be conducted for all populations that use
a given measure (Flake, 2021) and with statistics being a required module
1
for
1
We use the term ‘module’ to describe the smaller units that make up a degree course (often lasting
one semester) to distinguish them from full degree programmes, which we refer to as ‘courses’.

undergraduate students of most social and physical sciences in universities throughout the
world (Schwab-McCoy, 2019), the extent to which these findings are generalisable should
be examined.
Therefore, the first aim of the present study was to assess generalisability by
repeating Terry et al.’s (2022) study in a large, international sample of university students
from different academic disciplines for whom statistics was part of their degrees.
Our second aim was to explore whether specific facets of the STARS and R-MARS
are driven by a superordinate parent construct (Lawson & Robins, 2021). For example,
scores on the scales’ test anxiety items might be driven by general test anxiety, and not
specific to mathematics or statistics tests. Therefore, we added further measures of fear of
negative evaluation (Carleton et al., 2011), intolerance of uncertainty (Carleton et al., 2007),
social interaction and performance anxiety (Baker et al., 2002; Liebowitz, 1987), creativity
anxiety (Daker et al., 2020), test anxiety (Benson & El‐Zahhar, 1994), and trait anxiety (Ree
et al., 2008) to assess whether they underpin STARS and R-MARS items.
Our third aim was to examine the constructs’ extrinsic convergent validity (ECV; the
extent two measures correlate with other constructs in the same ways; Gonzalez et al.,
2020). The more similar the correlations are, the more probable it is that the measures are
tapping the same construct (Gonzalez et al., 2020). To examine ECV, we included five
additional variables shown to correlate with statistics and/or mathematics anxieties: Self-
efficacy (e.g., Z = |0.52|; Trassi et al., 2022), persistence (e.g., r = -.746; González et al.,
2016), analytic thinking (using a revised version of the Cognitive Reflection Test; CRT;
Shenhav et al., 2012)
2
, pre-university mathematics qualifications (e.g., r = -.27; Beurze et al.,
2013), and university statistics module grades (although, this relationship varies from r = -.56
to r = .10; Terry & Field, 2022).
2
We also added a single item measure of participants’ belief in God(s) to test a CRT-related research
question, outside of our core aims.

Besides our core aims, we maximised the reuse potential of this dataset - particularly
its capacity to address important questions in the statistics education literature (see Section
4 - Reuse Potential) - by collecting data from the student participants’ statistics instructors
about their module format, content, and assessment.
(2) Methods
2.1 Study design
The data were collected via a cross-sectional, online, self-report questionnaire-
based, multi-centre study. The final dataset was generated from the following three sources
(see Section 2.5 for full details of each variable):
1) The student survey, containing survey responses from university students on
measures of statistics and mathematics anxieties (including the modified versions from Terry
et al., 2022), test anxiety, trait anxiety, fear of negative evaluation, social interaction anxiety
and performance anxiety, intolerance of uncertainty, creativity anxiety, self-efficacy,
persistence, analytical thinking, and belief in God/s. Students also provided demographic
information (age, gender/sex, ethnicity, and any specific learning difficulties), information
about their pre-university mathematics qualifications (highest level, grades, and how long
ago they were taken), self-reported grades for completed statistics modules, and their
degree course details (major, year of study, and whether they are studying any non-statistics
mathematics-based modules). Student survey data also includes selected information auto-
recorded by Qualtrics (Qualtrics, Provo, UT; the start and end dates, duration, and
completion percentage for each response) and key identifiers added by the lead author
(participant ID, survey ID, country, and language).
2) An instructor survey, containing information about the statistics modules students
were taking at the time of completing the survey. The instructor survey recorded dates of the
student participants’ statistics module, mode of teaching delivery (e.g., lectures/workshops,

online/face-to-face), module content, and types and dates of assessments. Instructors also
indicated how assessments were graded, necessary to standardise grades across
institutions.
3) Students’ grade data from university records (where permitted to obtain and
share).
2.2 Time of data collection
Data were collected between January 2021 and September 2021
3
. Due to the
differences in term/semester dates cross-nationally, different research teams had different
start and end dates. The date participants began and finished the survey is included in the
dataset.
2.3 Location of data collection
Table 1.
A table detailing the universities data were collected from, the country they were in
4
,
associated survey language, and the number of responses at the country and university
level (after exclusions).
Country (ISO Code; N)
Language
University
N
Australia (AU; N = 315)
English
Macquarie University
237
English
Australian National University
53
English
University of Western Australia
25
Austria (AT; N = 230)
German
University of Vienna
120
3
The period of data collection coincided with the Covid-19 pandemic, which affected teaching delivery
(e.g., the move to online learning, some details of which were captured by the instructor survey) and
general anxiety levels may have been higher than usual.
4
Note that the UK includes the devolved nations of England, Scotland, and Northern Ireland (we did
not collect data in Wales). Note, however, that the devolved nations have different education systems
both pre-university (e.g., different mathematics exams) and during university (e.g., different degree
durations).

German
University of Graz
108
German
Medical University of Graz
1
German
Technical University of Graz
1
Belgium (BE; N = 184)
French
Catholic University of Louvain (UCLouvain)
184
Brazil (BR; N = 68)
Portuguese
University of Brasília
58
Portuguese
UNESP - São Paulo State University
10
Canada (CA; N = 986)
English
Toronto Metropolitan University (formerly Ryerson University)
520
English
York University
228
English
Memorial University of Newfoundland
127
English
Western University
111
Chile (CL; N = 191)
Spanish
Andrés Bello National University
98
Spanish
University of Concepción
93
China (CN; N = 323)
Chinese
Tianjin Normal University
196
Chinese
Qufu Normal University
127
Colombia (CO; N = 114)
Spanish
El Bosque University
113
Spanish
Other (unspecified)
1
Egypt (EG; N = 1390)
Arabic
Menoufia University
1,390
Estonia (EE; N = 98)
Estonian
University of Tartu
91
Estonian
Tallinn University
7
France (FR; N = 248)
French
University of Nantes
248
Germany (DE; N = 506)
German
University of Erfurt
231
German
University of Konstanz
114
German
University of Tübingen
110
German
University of Kassel
50
German
International University of Applied Sciences
1
Ghana (GH; N = 41)

English
University of Education, Winneba
19
English
University of Cape Coast
9
English
University of Ghana
7
English
All Nations University
2
English
Kwame Nkrumah University of Science and Technology
1
English
Other (unspecified)
1
English
University of Health and Allied Sciences
1
English
University of Professional Studies, Accra
1
Greece (GR; N = 99)
Greek
Panteion University
94
Greek
Aristotle University of Thessaloniki
2
Greek
National and Kapodistrian University of Athens
2
Greek
University of Crete
1
Hungary (HU; N = 206)
Hungarian
ELTE Eötvös Loránd University
184
Hungarian
University of Pécs
22
India (IN; N = 41)
English
CHRIST (deemed to be) University
41
Indonesia (ID; N = 697)
Bahasa Indonesia
Bina Nusantara University
223
Bahasa Indonesia
Brawijaya University
171
Bahasa Indonesia
Airlangga University
131
Bahasa Indonesia
Pelita Harapan University
96
Bahasa Indonesia
Padjadjaran University
62
Bahasa Indonesia
Atma Jaya Catholic University of Indonesia
14
Ireland (IE; N = 82)
English
University of Limerick
60
English
University College Dublin
22
Israel (IL; N = 285)
Hebrew
Reichman University (née Interdisciplinary Center Herzliya)
285
Italy (IT; N = 248)
Italian
University of Bergamo
176
Italian
LUMSA University
72
Malaysia (MY; N = 369)
English
HELP University
369
Netherlands (NL; N = 508)

Dutch
Radboud University
165
Dutch
Tilburg University
133
English
University of Amsterdam
114
Dutch
Erasmus University Rotterdam
96
Nigeria (NG; N = 255)
English
University of Nigeria
255
Philippines (PH; N = 47)
English
University of the Philippines Visayas
47
Poland (PO; N = 69)
Polish
University of Silesia
58
Polish
WSB University, Poznan
11
Romania (RO; N = 317)
Romanian
Alexandru Ioan Cuza University
317
Saudi Arabia (SA; N = 100)
Arabic
King Faisal University
100
Serbia (RS; N = 117)
Serbian
University of Novi Sad
117
Slovakia (SL; N = 88)
Slovakian
University of Prešov
88
Slovenia (SL; N = 94)
Slovenian
University of Maribor
94
Spain (ES; N = 346)
Spanish
University of La Laguna
218
English
ESADE Business School, Universitat Ramon Llull
128
Turkey (TR; N = 834)
Turkish
Hasan Kalyoncu University
339
Turkish
MEF University
160
Turkish
Baskent University
158
Turkish
Izmir University of Economics
100
Turkish
Middle East Technical University
77
UK (GB; N = 2962)
English
University of Sussex
413
English
University of Birmingham
363
English
Bournemouth University
214
English
Nottingham Trent University
202
English
University of Southampton
163

English
Kingston University
157
English
Queen's University Belfast
137
English
Loughborough University
134
English
University of Stirling
125
English
University of Lincoln
124
English
University of Hull
123
English
University of Portsmouth
116
English
University of Winchester
107
English
University of Brighton
99
English
University of Surrey
99
English
Teesside University
90
English
University of Derby
90
English
Glasgow Caledonian University
60
English
University of Bradford
56
English
Anglia Ruskin University
36
English
Manchester Metropolitan University
32
English
Leeds Trinity University
22
Ukraine (UA; N = 25)
Ukrainian
Sumy Makarenko State Pedagogical University
25
USA (US; N = 87)
English
University of Southern Indiana
51
English
University of Northern Colorado
33
English
Avila University
3
Total
12,570
Figure 1.
The top panel is a map showing the countries from which data was collected and their
respective sample sizes. The bottom panel is a treemap of sample sizes for each country,
organised by continents (see Table 1 for ISO country codes).

2.4 Sampling, sample and data collection
Collaborating research teams were recruited via Twitter and word-of-mouth, with
efforts made to invite researchers from geographically and culturally diverse countries with
varying education systems to produce more generalisable results. In the end, data were
collected from 100 universities in 35 countries and in 21 languages.
Student Survey
Student participants were invited to take part by collaborating researchers (or those
with access to the sample on researchers’ behalf, such as statistics module instructors) via
email, virtual learning environments, university-specific student social media platforms, and
university participant pools. Some students were also invited to complete the survey as part
of an in-class exercise. The study was hosted via Qualtrics online survey software (Qualtrics,
Provo, UT) and students completed it using a suitable electronic device (e.g., laptop, mobile
phone, or tablet) with internet access.
A total of N = 18,841 student survey responses were recorded. For the present
version of the data, we have excluded duplicates (identified using a combination of
participant-generated ID code and demographic responses; n = 72) and any cases where
the participant did not respond to any items on the measurement scales; n = 6,199). We
have not excluded any other data. Note that for our primary research study, we planned to
recruit undergraduate students that had taken or were taking statistics as part of their
research methods training on any degree course that is not typically associated with
mathematics. For example, we would exclude courses such as physics, engineering, and
data science, whilst courses such as social sciences, business, and geography were eligible.
Despite this stipulation, some responses were received from postgraduate students (n = 3),
and from those on mathematics and statistics degrees (n = 2), mathematics-adjacent
degrees (e.g., physics, engineering, computer sciences; n = 151), and degrees that are
unlikely to have included a statistics module (e.g., arts & humanities subjects; n = 232). We

have included these responses in the present dataset to afford other researchers the
opportunity to set their own exclusion criteria. Similarly, the pre-registration for the primary
empirical study (https://osf.io/be4yh/) states that we would only retain responses that passed
the attention checks, but we have not removed them in the present data (n = 8,597 passed
all seven).
After exclusions, the final sample presented here contains n = 12,570 responses
(68.2% of initial responses). Table 1 contains a breakdown of the number of responses from
each university.
Participants’ ages ranged from 18 to 67 years (M = 21.01, SD = 4.12); with 3,119
participants choosing not to respond to this question and 14 values (>=99 years) recoded as
implausible. The majority of participants identified as
5
a woman/female (n = 8,298, 66.0%),
with a further 2,002 identifying as a man/male (15.9%), 74 as non-binary (0.6%), 13
preferred to describe their gender in another way (0.1%)
6
, and 2,183 (17.4%) did not
respond to this question. Most participants (n = 9,026, 71.8%) reported they did not have a
diagnosis of any of the following Specific Learning Differences (SpLDs): ADHD/ADD,
Dyslexia, Dyscalculia, Dyspraxia, or Dysgraphia/Dysorthography. However, 738 (5.91%)
participants reported having one or more SpLD, whilst a further 111 (0.8%) responded
“other” (including self-diagnosis), 3 were unsure (0.02%), and 2,692 (21.4%) did not
respond.
Most participants indicated they were in the first year of their degree course (n =
4,505, 35.8%), with a further 3,126 in second year (24.9%), 1,859 in third year (14.8%), 689
in fourth year (5.5%), and 40 in fifth year (0.33%). An additional 61 participants (0.5%)
5
Although we recognise that “female” and “male” refer to biological sex and that “woman” and “man”
refer to gender identity, the adapted response options to the question asking participants’ gender
identity were inconsistent with some listing “woman” and “man”, others listing “female” and “male”,
and others listing “woman/female” and “man/male”, so we have merged the responses.
6
An anonymised list of the ways participants described their gender is available in the OSF
supplementary materials.

indicated their degree year as ‘other’, three participants (0.02%) were postgraduates, and
2,287 (18.2%) did not respond. Psychology was the most common degree major amongst
participants (n = 8,759, 69.7%), followed by Business and Finance (n = 768, 6.1%),
Education (n = 397, 3.2%), Health and Medical Sciences (n = 273, 2.2%), and Computer
Sciences (n = 128, 1.0%)
7
. A further 1,526 (12.14%) of students did not indicate their degree
major.
Each university provided their own participation incentives based on local norms and
availability. Half of participants were offered ungraded course credits (50.0%) and around a
third were offered no incentive (33.0%), with the remaining being offered either a prize draw
(up to a maximum of £50 or local equivalent per 100 participants; 10.0%); payment
(maximum £5 or local equivalent; 3.1%); a choice of a prize draw or course credits (3.3%), or
both payment and course credits (0.5%). Incentive information is unavailable for 0.14% of
participants.
Instructor Survey
The student participants’ statistics module instructors were invited to take part by
email (either by the lead researcher, where the collaborating researchers were also module
instructors, or by the collaborating researchers where there were not). In some cases,
someone other than the primary instructor completed the survey (e.g., graduate teaching
assistants). The instructor survey was also hosted via Qualtrics online survey software
(Qualtrics, Provo, UT).
A total of N = 176 instructor survey responses were recorded. We have excluded
responses given in error (e.g., for a postgraduate course or for more than one module per
response; n = 21) and any responses where no data was entered (n = 36).
7
See the supplementary materials for the frequency of degree major categories below 1%.

After exclusions, the final sample contained n = 119 responses (67.6% of initial
responses), representing n = 96 modules in n = 57 universities in n = 27 countries,
corresponding to n = 4,867 student survey responses.
Grade Data
Where permitted by the student participants and by their universities, we also
collected grades (and grading scales) for the statistics module students were taking at the
time of completing the survey from university records
8
. A total of N = 20 universities provided
this data, corresponding to n = 1,804 student participants in n = 41 modules in n = 9
countries.
2.5 Materials/ Survey instruments
Student Survey Adaptations
The student survey was prepared in stages. First, a generic master version of the
survey was created in English by the lead researcher (available here: https://osf.io/enc29).
This version was then adapted from English into the local language by collaborating
research teams as required, resulting in a generic master version for each language. A
translation guide was provided (available here: https://osf.io/v3qxf), which advised
translators to adopt a team-based approach (Behr & Shishido, 2016). This approach
involved a minimum of two people translating the scales individually and resolving any
differences as a team. It was chosen over the more ubiquitous back-translation technique,
because it is more effective in producing equivalent scales across languages (Behr, 2017).
The generic master version for each language was then copied for each research
team for modification to the local context, following guidelines provided to encourage
consistency (available here: https://osf.io/t2pc5). Modifications were kept minimal and
8
Some universities also provided students’ grades from their previous modules. This data will be
provided as a supplementary file on the OSF at a later date.

primarily pertained to course/module details (e.g., the names of the statistics modules), the
mathematics education questions (e.g., to reflect the structure of pre-university education
locally), and the demographic questions (e.g., adapting the ethnicity options to reflect local
populations). Researchers could also adapt it to award participant incentives (e.g., linking to
local course credit systems). The measurement scales were not altered, with minor
exceptions (detailed in the Measures section below). Very rarely, and where it did not impact
on our core research aims, questions were removed altogether to meet the requirements of
the local ethics boards and/or to be appropriate in the local context (e.g., some ethics boards
requested we did not ask about ethnicity).
All materials, including copies of all adapted/modified surveys are available on the
project’s OSF page (https://osf.io/3bmqz/).
Measures
9
Student Survey
Statistics Anxiety. Statistics anxiety was measured by the Statistics Anxiety Rating
Scale (STARS; Cruise et al., 1985). The three anxiety subscales (Hanna et al., 2008;
Papousek et al., 2012) of the STARS were used (23 items in total); test and class anxiety (8
items), interpretation anxiety (11 items), and fear of asking for help (4 items). Each item
describes a situation involving statistics such as “Doing an examination in a statistics
course” (test and class anxiety), “Interpreting the meaning of a table in a journal article”
(interpretation anxiety), or “Going to ask my statistics teacher for individual help with material
I am having difficulty understanding” (fear of asking for help). Participants were asked to
indicate how much anxiety they feel in those situations on a Likert scale ranging from 1 = “no
anxiety” to 5 = “a great deal of anxiety”.
9
We do not provide reliability coefficients for the measures because such coefficients should be
calculated for the specific subsample chosen for any secondary research studies.

Several items use outdated language and were modified to reflect modern
equivalents (e.g., “Asking one of my teachers for help in understanding a printout” was
changed to “Asking one of my teachers for help in understanding statistical output”). These
modifications are the same as those made in Terry et al. (2022).
An attention check was also included in this scale, which asked participants to
“Please select '1 - no anxiety' for this question”.
Mathematics Anxiety. Mathematics anxiety was measured with the Revised
Mathematics Anxiety Rating Scale (R-MARS; Baloğlu & Zelhart, 2007). There are three
subscales in the R-MARS which measure mathematics test anxiety (15 items), numerical
task anxiety (5 items), and mathematics course anxiety (5 items). Each item describes a
situation involving mathematics such as “Taking an exam in a math course” (mathematics
test anxiety), “Being given a set of division problems to solve” (numerical task anxiety), or
“Listening to another student explain a math formula” (mathematics course anxiety).
Participants are asked to indicate how much anxiety they feel in those situations on a Likert-
type scale ranging from 1 = “no anxiety” to 5 = “a great deal of anxiety”.
Where the local context required it, items were modified to reflect local equivalents of
US terms (e.g., in the UK, “Taking the math section of a college entrance exam” was
changed to “Taking the maths section of a university entrance exam”).
Modified STARS and R-MARS. The modified versions of the STARS and R-MARS
used in Terry et al. (2022) were also included. In these versions, the STARS items were
revised to reflect mathematics-related situations (e.g., “Doing the coursework for a statistics
course” was changed to “Doing the coursework for a mathematics course”) and the R-MARS
statements were revised to reflect statistics-related situations (e.g., “Walking into a
mathematics class” was changed to “Walking into a statistics class”). The response scales
were kept the same as the originals.

Three items in the STARS were not easily distinguishable as being about either
mathematics or statistics so equivalent items were not created (“Arranging to have a body of
data put into the computer”, “Reading an advertisement for a car which includes figures on
miles per gallon, depreciation, etc.”, and “Trying to understand the odds in a lottery”).
Additionally, one item on the R-MARS was deemed untranslatable to a statistics context so,
again, an equivalent was not created (“Reading a cash register receipt after your purchase”).
The exploratory factor analysis in Terry et al. (2022) indicated that the R-MARS
numerical task anxiety subscale was the only subscale where the revised items did not load
onto the same factor as the corresponding original items. We believe the inconsistency in
factor loadings in the original study could be because the modifications were not equivalent.
For example, "Being given a set of numerical problems involving addition to solve on paper"
was modified for the statistical context to "Calculating the sum of squared deviances by
adding the squared deviances together” and, although the two both involved addition, the
latter would be less familiar to participants and thus could be perceived as more a complex
mathematical task. To rule out the possibility that the original and modified items loaded onto
separate factors due to differences in perceived complexity, we re-modified the numerical
task anxiety items and added these to the present version (as well as the original
modifications, for comparison). Four items were modified from mathematics items to
statistics items whilst keeping the language more consistent (e.g., “Being given a set of
numerical problems involving addition to solve on paper” was modified to “Being given a set
of statistical problems involving addition to solve on paper”) and four items were changed
from our original modifications back to mathematics but matching the more complex
language used (e.g., “Calculating the sum of squared deviances by adding the squared
deviances together” has been modified to “Finding the codomain of the function h(x, y) = x +
y when x = {3,4,5,6} and y = {5,7,9,13}”).
An attention check was also included in this modified STARS, which asked
participants to “Please select '5 - a great deal of anxiety' for this question”.

Trait Anxiety. Trait anxiety was measured using the trait subscale of the State Trait
Inventory for Cognitive and Somatic Anxiety (STICSA; Ree et al., 2008). The STICSA has
been developed and evidenced to differentiate anxiety from depression more effectively than
other popular anxiety measures (e.g., the State-Trait Anxiety Inventory (STAI); Spielberger,
1983; Tindall et al., 2021). The trait subscale is further broken down into cognitive (10 items)
and somatic symptoms (11 items). Cognitive symptoms are measured with statements such
as “I cannot concentrate without irrelevant thoughts intruding” and somatic symptoms are
measured with statements such as “My heart beats fast”. Participants are asked to indicate
the extent to which each item is true of them on a Likert scale ranging from 1 = “not at all” to
4 = “very much so”.
An attention check was also included in this scale, which asked participants to
“Please select '1 - not at all' for this question”.
Test Anxiety. Test anxiety was measured with the Revised Test Anxiety Scale (R
TAS; (Benson & El‐Zahhar, 1994). The scale contains four subscales: 6 worry items (e.g.,
“During tests I find myself thinking about the consequences of failing”), 5 tension items (e.g.,
“During tests I feel very tense”), 4 test-irrelevant thinking items (e.g., “During tests I find I am
distracted by thoughts of upcoming events”), and 5 bodily symptoms items (e.g., “I get a
headache during an important test”). Participants were asked to respond to each item in
terms of how they feel when taking tests in general on a scale of 1 = “almost never” to 4 =
“almost always”.
Fear of Negative Evaluation. Following recommendations by Carleton et al. (2011),
fear of negative evaluation was measured using the Brief Fear of Negative Evaluation Scale
– Straightforward (BNFE-S; Leary, 1983; Rodebaugh et al., 2004). The scale contains 8
items, including statements such as, “I am afraid that people will find fault with me” and “I
often worry that I will say or do the wrong things”. The BNFE-S omits the reverse-scored
items in the original BNFE scale (items 2, 4, 7, and 10) which were found to be measuring a
different construct (Carleton et al., 2011). Participants were asked to indicate how

characteristic each item is of them on a Likert scale ranging from 1 = “not at all characteristic
of me” to 5 = “extremely characteristic of me”.
An attention check was also included in this scale, which asked participants to
“Please select '3 - moderately characteristic of me' for this question”.
Social Interaction Anxiety and Performance Anxiety. Social interaction anxiety
and performance anxiety were measured using the experienced fear/anxiety dimension of
the Liebowitz Social Anxiety Scale - Self Report (LSAS-SR; Baker et al., 2002; Liebowitz,
1987). The scale is broken down into social interaction anxiety (12 items, e.g., “Talking with
people you don’t know very well”) and performance anxiety (12 items, e.g., “Participating in
small groups”). Participants were asked to indicate how anxious they would feel in each
situation on a Likert scale ranging from 0 = “not at all” to 3 = “very much so”.
Some LSAS-SR items were adapted to respect local laws/norms in Saudi Arabia.
Specifically, “Drinking with others” was reworded to “Drinking coffee with others”, “Urinating
in a public bathroom” was changed to “Using a public bathroom”, and “Trying to make
someone's acquaintance for the purpose of a romantic/sexual relationship” was changed to
“Making someone's acquaintance for the purpose of making a marriage proposal”.
Intolerance of Uncertainty. Intolerance of uncertainty was measured using the
Intolerance of Uncertainty Scale - Short Form (IUS-SF; Carleton et al., 2007). The scale
contains 12 items, including statements such as, “It frustrates me not having all the
information I need” and “The smallest doubt can stop me from acting”. Participants were
asked to indicate how characteristic each item is of them on a Likert scale ranging from 1 =
“not at all characteristic of me” to 5 = “entirely characteristic of me”.
Creativity/Non-Creativity Anxiety. Creativity/Non-Creativity Anxiety was measured
using the Creativity Anxiety Scale (Daker et al., 2020). The scale contains 16 items: 8
creativity anxiety items (e.g., “Having to solve a problem for which the solution is open-
ended”) paired with 8 non-creativity items (e.g., “Working in a situation where there is an

established correct and incorrect way of doing things”). Participants were asked to indicate
how much each situation would make them feel anxious on a Likert scale ranging from 0 =
“not at all” to 4 = “very much”.
An attention check was also included in this scale, which asked participants to
“Please select '2 - a little' for this question”.
Analytic Thinking. Analytic thinking was measured using a revised version of the
Cognitive Reflection Test (CRT; Frederick, 2005), developed by Shenhav et al. (2012). We
selected a revised version because participants were less likely to be familiar with it than the
original. Like the original, the revised CRT contains three word problems, each of which
requires a numerical response. Questions are open-ended, but respondents typically give
either the correct response (indicating greatest analytic thinking), a single incorrect and
intuitively compelling response, or varying incorrect and unintuitive responses. The data set
contains the raw numerical responses given by participants so that researchers can code
them according to their chosen criteria.
Participants were also asked, “You have just answered three reasoning problems.
How many of them do you think you answered correctly?” and - to help ensure the integrity
of the revised CRT - “You have just answered three reasoning problems. Did you look any of
the answers up online?”, to which they could respond “Yes” or “No”.
Belief in God. Participants’ belief in God(s) was recorded using a single item.
Participants were asked, “How strongly do you believe in God (or gods) from 0-100? If you
are certain that God (or gods) does not exist, then enter “0” and if you are certain that God
(or gods) does exist then enter “100”.” Possible responses ranged between 0 and 100.
Self-Efficacy. Self-efficacy was measured with the 8-item New General Self Efficacy
Scale (NGSE; Chen et al., 2001) which contains items such as “When facing difficult tasks, I
am certain that I will accomplish them”. Participants were asked to indicate the extent to
which they agree with each statement on a Likert scale of 1 = “strongly disagree” to 5 =

“strongly agree”. An attention check was also included in this scale which asked participants
to “Please select '4 - agree' for this question”.
Persistence. Persistence was measured with the persistence subscale of the
Attitude Towards Mathematics Survey (ATMS; Miller et al., 1996), which contains 8 items
such as “If I have trouble understanding a problem, I go over it again until I understand it”.
Although the ATMS as a whole focusses on mathematics, the persistence subscale items
refer to academic persistence more generally. Some items were modified to make them
more appropriate for the higher education context. Specifically, in item 3 the words “in the
book” were removed, in item 6 the words “hope that the teacher explains it” were changed to
“hope that it is explained”, and the word “homework” was removed from items 2, 7, and 8.
Item 4, “If I have trouble solving a homework problem in the book, I copy down the answer in
the back of the book if it is available”, was removed because the required modifications
would have changed the meaning too far from the original. All items except 1 and 7 are
reverse scored. Participants were asked to indicate the extent to which they agree with each
statement on a Likert scale of 1 = “strongly disagree” to 5 = “strongly agree”.
An attention check was also included in this scale, which asked participants to
“Please select '4 - agree' for this question”.
Mathematics Education. Participants were asked for their highest level of pre-
university mathematics education (GCSE or A Level or international equivalents), the grade
they received at each level, and how long ago (in months) they took each qualification.
These questions were modified for the local context of each partner university and,
consequently, some include additional questions (see codebook for full details). Note that
grades are in their raw form and will need to be standardised before they can be compared.
Statistics Grades (Self-Reported). We asked participants whether they had
previously taken any university-level statistics modules and, for those that had, to self-report
their grades for these modules. Note that grades are in their raw form and will need to be
standardised before they can be compared.

Degree Course Details. Participants were asked to indicate their (intended) major
subject of study (i.e., the subject of the degree that they are pursuing), their current year of
study, and whether they were studying any other (i.e., non-statistics) mathematics-based
modules on their degree. Where the local researchers already know these details (e.g., they
were only sharing the survey with their own students) these questions were omitted to
reduce the length of the survey and the information was instead added into the data during
data processing.
Demographics. Participants were also invited to provide their age (in years), gender
identity, ethnicity, and whether they have been diagnosed with a specific learning difference
(SpLD), such as dyslexia or dyscalculia.
Attention Check. In addition to the attention checks embedded within the
measurement scales, participants were presented with the following at the end of the
survey: “Please indicate whether you feel you have answered the previous questions
carefully and truthfully. Answering 'yes' will ensure that your data is included in our analyses.
Answering 'no' will mean that your data is excluded from our analyses but will have no
impact upon you (i.e., you will still earn your incentive for taking part)”. Participants could
respond “Yes, I have answered all questions carefully and truthfully” or “No, I have not
answered the questions carefully and truthfully”.
Survey Metadata. The dataset also contains selected metadata that was
automatically collected by Qualtrics (Qualtrics, Provo, UT), which researchers may find
useful. Specifically, we include the percentage of the survey completed, the time it took to
complete the survey, and the date participants began and finished the survey.
Identifiers. We have also added relevant identifiers. Specifically, the country in
which the survey was taken, the language in which the survey was taken, the survey ID
(because some surveys were made available to students in more than one university), and a
randomly generated participant ID, which replaced the participant-generated ID code for
anonymisation purposes. The type of incentive offered to students has also been recorded.

Instructor Survey
Statistics Module Details. The instructor survey asked for the following information
about each module: Name and/or code, start and end dates, the statistical software taught,
the approximate content of the modules (via a checklist of different statistics topics), the
primary academic discipline of the instructors, the mode(s) of teaching and number of hours
per format (e.g., 1-hour online lecture, 2-hour in-person workshop), the types, format, and
date of assessments, how assessments were graded, opportunities for formative feedback,
average grade from previous cohorts, and any other information that would be useful to
contextualise the assessment information.
Grade Data
Statistics Grades (Official). At universities where it was approved by the local
ethics committees, we asked student participants to provide their names and/or student ID
codes, so that the grades for the statistics modules they were taking at the time they
completed the survey could be obtained from their university records. Note that grades are in
their raw form, but we also provide grading scale/system information to enable
standardisation.
Procedure
Student Survey
Upon receiving the invitation to take part, students were directed to the online survey
where they read the information sheet and provided consent before continuing. Participants
were then asked to complete an eligibility check (if they had not been pre-screened), and to
provide their name and/or student ID code (to obtain grade data from student records, where
relevant), a unique participant ID code (to withdraw their data, if desired), and their primary
degree subject and statistics module names (if researchers were unsure of these details in
advance). All participants then completed the first block of measurement scales containing

all four measures of statistics anxiety and mathematics anxiety, randomised at the measure
and item level. This block was presented first because it contained the measures most
critical to the study and if participants did not proceed to the next block, their data would still
be useful. The second block of measurement scales - also randomised at the question and
item level - contained measures of trait anxiety, test anxiety, fear of negative evaluation,
social interaction/performance anxiety, creativity anxiety, intolerance of uncertainty, self-
efficacy, persistence, and the revised cognitive reflection test (CRT). The question asking
about participants’ belief in God (or gods) was randomly presented before or after the
revised CRT. The two follow-up questions about the revised CRT were then asked.
Participants were next asked about their pre-university mathematics education, their
statistics grades from previous modules at university (if applicable), the year of their degree
course, and demographics. Finally, participants answered an attention check question, were
debriefed, and if required, redirected to collect their survey incentives. The median
completion time for the survey was 30 minutes).
Instructor Survey
Upon receiving the invitation to take part, statistics module instructors were directed
to the online survey where they read the information sheet and provided consent before
continuing. Participants were first reminded that they should complete the instructor survey
once for every statistics module that the student participants were taking at the time of
completing the student survey and provided with a unique code they could use if they later
wished to remove their data. The survey then requested (in order) the university name, the
statistics module name and code, and the start and end dates of the module. Participants
could then select the software(s) taught on the module, whether the module was frequentist,
Bayesian, both, or other, and select the topics taught from a checklist (e.g., ANOVA, Bayes
factors, Data visualisation). We then asked whether the module was taught by the
mathematics/statistics department or from the students’ main discipline (e.g., psychology
Lecturers that teach statistics). The survey then requested the percentage of in-person

teaching and whether there was less than usual due to COVID-19. We then asked for details
about the mode of teaching (e.g., lectures, practicals), including how many hours per week
were spent on each, whether they were online or in-person, and synchronous or
asynchronous. The next section was about module assessments. We asked for the type of
assessment (e.g., exams, coursework), the percentage of the final grade each type
contributed to, the length of any timed assessments, whether assessments were online/in-
person (where appropriate), the date of exams/deadline for coursework, and the scale used
for grading (e.g., numeric continuous, letter grades). Where respondents reported using
regular testing, we also asked for the frequency and format (e.g., quizzes, tasks) of testing,
whether they were timed, and whether all grades counted towards the final, overall grade.
Next, instructors could indicate the types of any formative assessment (e.g., verbal /written,
peer/instructor), what the average final overall grade for the module usually, and, finally,
instructors were invited to record any additional information about their assessments that
could be useful for contextualising their data.
Grade Data
Where permitted by collaborating universities’ ethics committees and legal teams,
grade data was obtained by the collaborating researchers and shared with the lead
researcher using password-protected files.
2.6 Quality Control
Attention Checks
At the end of the student survey, participants were asked whether they had answered
all questions truthfully and carefully, to which 10,281 (81.8%) responded “yes”, 172 (1.4%)
responded “no”, and 2117 (16.8%) responses are missing (where participants did not reach
that stage of the survey).

Additionally, six attention checks were embedded within the measurement scales
which asked participants to select a specific response (e.g., “please select ‘1 – strongly
disagree’”). There were two in the first block which contained the statistics and mathematics
anxiety measures and four in the second block which contained all other scales. In the first
block, 2,052 (16.3%) participants responded incorrectly to the first check and 2,194 (17.5 %)
responded incorrectly to the second. In the second block, the number of students responding
incorrectly to each check were 2,779 (22.1%) to the third, 2,812 (22.4%) to the fourth, 2,872
(22.9%) to the fifth, and 2,752 (21.9%) to the sixth.
CRT Check
To help ensure the integrity of the revised CRT, participants were asked “You have
just answered three reasoning problems. Did you look any of the answers up online?”, to
which they could respond “Yes” or “No”. There were 382 (3.0%) participants that answered
“Yes’ to this question.
2.7 Data anonymisation and ethical issues
Ethics
This study was approved (ER/JLT26/7) by the Sciences & Technology Cross-
Schools Research Ethics Committee (C-REC) at the University of Sussex in adherence to
the British Psychological Society’s Code of Human Research Ethics (2018). Partner
universities were covered by the overarching University of Sussex ethics approval, but were
asked to check with their own ethics boards whether further approval was required at the
local level and, if necessary, to obtain it before beginning data collection. Ethics approval
documentation is available here: https://osf.io/2aumd/.
For those universities that shared students’ grade data with us, a data protection
agreement was in place to allow the legal transfer of the non-anonymised data (i.e., student

names and/or ID codes) required to obtain, share, and link grades to participants’ survey
data.
Anonymisation
Raw data has and will only ever be available to the research leads at the University
of Sussex. To anonymise the data for sharing, the student names and ID codes have been
replaced with a randomly generated unique ID code, and the demographic variables and
some course/module details from the student surveys have been edited as required to
ensure that participants are not identifiable via a combination of these data. Specifically,
students’ age, degree major, and any specified non-statistics mathematics modules have
been categorised, gender identities and SpLDs have been partially re-categorised, and
ethnicity data have been removed completely. Full details on how the data has been
processed for anonymisation is available in the codebook and data processing notes
(available here: https://osf.io/374vn/).
2.8 Existing use of data
There are presently no published articles or other outputs originating from this data.
However, following the embargo period, researchers will be able to register their planned
secondary analyses on an open document, which we encourage use of to prevent
duplication of efforts.
(3) Dataset description and access
3.1 Repository location
DOI 10.17605/OSF.IO/MHG94

3.2 Object/file name
The dataset will be available in its complete form and also in its component parts, as
follows:
● SMARVUS_complete.csv – all data
● SMARVUS_demo_meta.csv – demographics, Qualtrics meta-data, and key
identifiers
● SMARVUS_measures.csv – measurement scales
● SMARVUS_maths_edu.csv – prior mathematics education data
● SMARVUS_stats_edu.csv – statistics education data (from official records
and self-reported)
3.3 Data type
Partially processed primary data.
10
3.4 Format names and versions
All versions of the dataset are available as .csv files, which can be opened using
most spreadsheet and statistics software.
3.5 Language
All data are stored in English (UK), except proper nouns (e.g., names of pre-
university mathematics qualifications). Free-text responses were mostly short and
straightforward to translate (e.g., degree major or gender identity) so were translated back
into English by the lead author using Google Translate. Where a translation was ambiguous,
it was clarified with native speakers.
10
Whether and how each variable has been processed is detailed in the ‘Data Processing Notes’
column of the code book and is summarised here:
https://osf.io/374vn/?view_only=9d70d0facfeb476987f32d4ab156ecdc.

3.6 License
CC BY-NC-ND
3.7 Limits to sharing
Data will be under embargo until 1st October 2024 to allow the authors sufficient time
to publish from it first. During this time, data will be made available upon request, provided
the intended research does not overlap with projects being undertaken by the present
authors.
3.8 Publication date
N/A
3.9 FAIR data/Codebook
Data are stored in .csv format on the OSF, along with a detailed codebook and all
materials, using a CC BY-NC-ND licence.
(4) Reuse potential
The SMARVUS dataset has the potential to address many important questions,
particularly regarding statistics and mathematics education, anxiety, psychometrics, and
survey methodology. It uniquely facilitates cross-lingual and cross-cultural comparisons and
the larger-than-usual sample size is more likely to produce reliable and robust estimates.
Below, we highlight just some of the ways this could benefit specific fields.
Statistics Education
These data enable the exploration of relationships between many constructs. For
example, a much-debated question is whether statistics anxiety effects achievement (e.g.,
statistics module grades). A recent meta-analysis of this relationship produced a non-

significant effect size of just Z = |0.07| (Trassi et al., 2022). However, the authors noted
considerable variability in their systematic review, explaining it may be attributed to
moderators, such as pre-university mathematics grades and self-efficacy. Another review
identified mode of assessment as a potential moderator (Terry & Field, 2022). These
moderators could be tested with the SMARVUS data.
Variability could also be due to the multi-dimensionality of the STARS (Cruise et al.,
1985). There are three subscales which measure statistics attitudes, not statistics anxiety
(Hanna et al., 2008; Papousek et al., 2012), thus should not be conflated. The data required
for Trassi et al. (2022) to separate these subscales was unavailable, forcing them to use
composite scores. A large-scale analysis of the relationship between statistics anxiety and
achievement using the anxiety subscales alone is possible with the SMARVUS data.
Trassi et al. (2022) further note that studies in their meta-analysis mainly tested
psychology students within Europe and North America and many had low sample sizes,
which the SMARVUS data addresses. Such limitations are pervasive in psychology
(Ioannidis, 2005; Rad et al., 2018), so these data could benefit many other research
questions in the same ways. Furthermore, the sample is sufficiently large to enable multi-
level modelling to estimate variation across different languages, geographic regions, or
educational systems.
Construct Validity
To understand how generalisable research is, the scales we use to measure
constructs must be validated in different populations (Flake, 2021). This includes ensuring
adaptations (e.g., translations) are valid and reliable, and that different groups respond to
measures in the same ways, such that the factor structure, loadings, and item intercepts are
equivalent (i.e., are measurement invariant).
Our student survey included eight scales adapted to 21 languages. We also modified
some scales to be appropriate for the local context. In most cases, this was minimal (e.g.,

changing “college” to “university”). However, we made more substantial modifications to our
measure of social interaction and performance anxiety - the LSAS-SR (Baker et al., 2002) -
for use in Saudi Arabia (e.g., modifying inappropriate references to alcohol and dating).
Validating adapted scales would ensure these versions are appropriate for use in different
countries and cultural contexts, opening up fresh opportunities for cross-cultural research.
Our data could also be used for measurement invariance testing. There is a dearth of
invariance testing for most psychological scales (D’Urso et al., 2022), so many gaps to be
filled. For example, we know that mathematics anxiety scores vary between cultures (Hunt et
al., 2021), which could be indicative of cultural non-invariance. If that is the case, the
generalisability of predominantly Western research findings should not be assumed. Such
findings might be misleading for other cultures, with consequences for education. This
dataset could address this problem via the cross-cultural investigation of mathematics
anxiety scale properties.
Cognitive Reflection Test (CRT)
The data includes responses to a revised version of the Cognitive Reflection Test
(CRT; Shenhav et al., 2012), a hugely popular measure of reflective thinking tendencies
(cited over 5000 times, according to Google Scholar). Projects are underway to test the
psychometric properties of Shenhav et al.’s (2012) version and, assuming the scale shares
key properties of the original (e.g., excellent validity, reasonable reliability, and incorrect
responses converging on the same typical response), the SMARVUS opens up opportunities
for research into cross-cultural and gender comparisons of cognitive reflection and its
relationship to various types of anxiety.
Survey Methodology
For survey-based research to be robust, it is essential that care and attention is
employed by respondents. One study found 10-12% of responses to long surveys by
undergraduates completing it for course credit are given without such care (Meade & Craig,

2012). Some researchers have proposed using attention checks to help identify and
eliminate such responses (Huang et al., 2012). The present study included attention checks
within the survey measures, asking participants to choose a particular response option, and
an ‘amnesty’ at the end, asking if they had answered carefully and truthfully throughout.
SMARVUS data could be used to compare the effectiveness of these checks as well as
other measures of careless responding, such as response time and ‘long-string analysis’
(providing the same response to all items on a scale; Curran, 2016).
Pedagogy
Finally, we suggest the SMARVUS dataset has unique pedagogical reuse potential.
First, students might find a dataset related to mathematics and statistics anxieties to be
relatable, something qualitative evidence suggests can aid learning (e.g., Blackburn, 2015)
and reduce anxiety (e.g., Trakulphadetkrai, 2017) in statistics education. Second, using
these data in a statistics class would give instructors an opportunity to make students
conscious of any anxieties, show them they are far from alone, and encourage students to
notice and, subsequently, challenge the influence anxiety may be having on their attitudes
and behaviours regarding learning statistics. Third, there are general benefits of using
authentic, secondary data in statistics education that could further enhance the specific
benefits. For example, students can learn data processing strategies that are usually
unavailable with pre-prepared datasets – such as dealing with missing data – alongside
statistical procedures and tests. Additionally, students who use this data for research
projects (e.g., undergraduate dissertations) could do so without the worry of ethics approval
or recruiting a large enough sample, and could present their work at conferences and in
publications, as previously done by Long & Chalk (2020) with Grahe et al.'s (2018) Emerging
Adulthood Measured at Multiple Institutions 2 (EAMMi2) data.

Contribution Statement
Conceptualization: Jenny Terry and Andy P. Field conceptualised the overall project, and
Robert M. Ross conceptualised the inclusion of the Cognitive Reflection Test (CRT).
Data curation: Jenny Terry managed and processed the data with coding support from
Tamas Nagy.
Funding acquisition: Patricia Garrido-Vásquez and Mauricio Salgado sourced funding for
incentive payments at their institutions.
Investigation: The following researchers contributed to data collection: Jenny Terry, Robert
M. Ross, Tamas Nagy, Mauricio Salgado, Patricia Garrido-Vásquez, Jacob O. Sarfo, Susan
Cooper, Anke C. Buttner, Tiago J. S. Lima, İbrahim Öztürk, Nazlı Akay, Flavia H. Santos,
Christina Artemenko, Lee T. Copping, Mahmoud M. Elsherif, Ilija Milovanović, Robert A.
Cribbie, Marina G. Drushlyak, Katherine Swainston, Yiyun Shou, Juan David Leongómez,
Nicola Palena, Fitri A. Abidin, Maria F. Reyes-Rodríguez, Yunfeng He, Juneman Abraham,
Argiro Vatakis, Kristin Jankowsky, Stephanie N. L. Schmidt, Elise Grimm, Desirée González,
Philipp Schmid, Roberto A. Ferreira, Rozgonjuk Dmitri, Neslihan Özhan, Patrick A.
O'Connor, Andras N. Zsido, Gregor Stiglic, Darren Rhodes, Cristina Rodríguez, Ivan
Ropovik, Violeta Enea, Ratri Nurwanti, Alejandro J. Estudillo, Nataly Beribisky, Karel K.
Himawan, Linda M. Geven, Anne H. van Hoogmoed, Amélie Bret, Jodie E. Chapman, Udi
Alter, Tessa R. Flack, Donncha Hanna, Mojtaba Soltanlou, Gabriel Banik, Matúš Adamkovič,
Sanne H. G. van der Ven, Jochen A. Mosbacher, Hilal H. Şen, Joel R. Anderson, Michael
Batashvili, Kristel de Groot, Matthew O. Parker, Mai Helmy, Mariia M. Ostroha, Katie A.
Gilligan-Lee, Felix O. Egara, Martin J. Barwood, Karuna Thomas, Grace McMahon, Siobhán
M. Griffin, Hans-Christoph Nuerk, Alyssa Counsell, Oliver Lindemann, Dirk Van Rooy,
Theresa E. Wege, Joanna E. Lewis, Balazs Aczel, Conal Monaghan, Ali H. Al-Hoorie, Julia
F. Huber, Saadet Yapan, Mauricio E. Garrido Vásquez, Antonino Callea, Tolga Ergiyen,

James M. Clay, Gaetan Mertens, Feyza Topçu, Merve G. Tutlu, Täht Karin, Mikkor Kristel,
Letizia Caso, Alexander Karner, Maxine M. C. Storm, Gabriella Daroczy, Rizqy A. Zein,
Andrea Greco, Erin M. Buchanan, Katharina Schmid, Thomas E. Hunt, Jonas De
keersmaecker, Peter E. Branney, Jordan Randell, Oliver J. Clark, Crystal N. Steltenpohl,
Bhasker Malu, Burcu Tekeş, TamilSelvan Ramis, Stefan Agrigoroaei, Nicholas A. Badcock,
Kareena McAloney-Kocaman, Olena V. Semenikhina, Erich W. Graf, Charlie Lea, Fergus M.
Guppy, Amy C. Warhurst, Shane Lindsay, Ahmed Al Khateeb, Frank Scharnowski, Leontien
de Kwaadsteniet, Kathryn B. Francis, Mariah Lecompte, Lisa A. D. Webster, Kinga
Morsanyi, Suzanna E. Forwood, Elizabeth R. Walters, Linda K. Tip, Jordan R. Wagge, Ho
Yan Lai, Deborah S. Crossland, Kohinoor M. Darda, Zoe M. Flack, Zoe Leviston, Matthew
Brolly, Samuel P. Hills, Elizabeth Collins, Andrew J. Roberts, Wing-Yee Cheung, Sophie
Leonard, Bruno Verschuere, Samantha K. Stanley, Iro Xenidou-Dervou, Omid Ghasemi,
Timothy Liew, Daniel Ansari, Johnrev Guilaran, Samuel G. Penny, Julia Bahnmueller, and
Christopher J. Hand.
Methodology: Jenny Terry, Robert M. Ross, and Andy P. Field.
Project administration: Jenny Terry, Robert M. Ross, Tamas Nagy, Mauricio Salgado,
Patricia Garrido-Vásquez, Jacob O. Sarfo, Susan Cooper, Anke C. Buttner, Tiago J. S. Lima,
İbrahim Öztürk, Nazlı Akay, Flavia H. Santos, Christina Artemenko, Lee T. Copping,
Mahmoud M. Elsherif, Ilija Milovanović, Robert A. Cribbie, Marina G. Drushlyak, Katherine
Swainston, Yiyun Shou, Juan David Leongómez, Nicola Palena, Fitri A. Abidin, Maria F.
Reyes-Rodríguez, Yunfeng He, Juneman Abraham, Argiro Vatakis, Kristin Jankowsky,
Stephanie N. L. Schmidt, Elise Grimm, Desirée González, Philipp Schmid, Roberto A.
Ferreira, Rozgonjuk Dmitri, Neslihan Özhan, Patrick A. O'Connor, Andras N. Zsido, Gregor
Stiglic, Darren Rhodes, Cristina Rodríguez, Ivan Ropovik, Violeta Enea, Ratri Nurwanti,
Alejandro J. Estudillo, Nataly Beribisky, Karel K. Himawan, Linda M. Geven, Anne H. van
Hoogmoed, Amélie Bret, Jodie E. Chapman, Udi Alter, Tessa R. Flack, Donncha Hanna,

Mojtaba Soltanlou, Gabriel Banik, Matúš Adamkovič, Sanne H. G. van der Ven, Jochen A.
Mosbacher, Hilal H. Şen, Joel R. Anderson, Michael Batashvili, Kristel de Groot, Katie A.
Gilligan-Lee, Felix O. Egara, Martin J. Barwood, Karuna Thomas, Grace McMahon, Siobhán
M. Griffin, Hans-Christoph Nuerk, Alyssa Counsell, Oliver Lindemann, Dirk Van Rooy,
Theresa E. Wege, Joanna E. Lewis, Thomas E. Hunt, Peter E. Branney, TamilSelvan
Ramis, Nicholas A. Badcock, Kareena McAloney-Kocaman, Olena V. Semenikhina, Frank
Scharnowski, Mariah Lecompte, Kinga Morsanyi, Suzanna E. Forwood, Ho Yan Lai,
Deborah S. Crossland, Zoe M. Flack, Samuel P. Hills, Timothy Liew, Daniel Ansari, and Julia
Bahnmueller.
Resources: Jenny Terry, Robert M. Ross, Tamas Nagy, Mauricio Salgado, Patricia Garrido-
Vásquez, Jacob O. Sarfo, Susan Cooper, Anke C. Buttner, Tiago J. S. Lima, İbrahim Öztürk,
Nazlı Akay, Flavia H. Santos, Christina Artemenko, Lee T. Copping, Mahmoud M. Elsherif,
Ilija Milovanović, Robert A. Cribbie, Marina G. Drushlyak, Katherine Swainston, Yiyun Shou,
Juan David Leongómez, Nicola Palena, Fitri A. Abidin, Maria F. Reyes-Rodríguez, Yunfeng
He, Juneman Abraham, Argiro Vatakis, Kristin Jankowsky, Stephanie N. L. Schmidt, Elise
Grimm, Desirée González, Philipp Schmid, Roberto A. Ferreira, Rozgonjuk Dmitri, Neslihan
Özhan, Patrick A. O'Connor, Andras N. Zsido, Gregor Stiglic, Darren Rhodes, Cristina
Rodríguez, Ivan Ropovik, Violeta Enea, Ratri Nurwanti, Alejandro J. Estudillo, Nataly
Beribisky, Karel K. Himawan, Linda M. Geven, Anne H. van Hoogmoed, Amélie Bret, Jodie
E. Chapman, Udi Alter, Tessa R. Flack, Donncha Hanna, Mojtaba Soltanlou, Gabriel Banik,
Matúš Adamkovič, Sanne H. G. van der Ven, Jochen A. Mosbacher, Hilal H. Şen, Joel R.
Anderson, Michael Batashvili, Kristel de Groot, Matthew O. Parker, Mai Helmy, Mariia M.
Ostroha, Balazs Aczel, Conal Monaghan, Ali H. Al-Hoorie, Julia F. Huber, Saadet Yapan,
Mauricio E. Garrido Vásquez, Antonino Callea, Tolga Ergiyen, James M. Clay, Gaetan
Mertens, Feyza Topçu, Merve G. Tutlu, Täht Karin, Mikkor Kristel, Letizia Caso, Alexander
Karner, Maxine M. C. Storm, Gabriella Daroczy, Rizqy A. Zein, Andrea Greco, Unita W.
Rahajeng, Dar Peterburg, and Zsofia K. Takacs.

Supervision: Jenny Terry, Robert M. Ross, Tamas Nagy, Mauricio Salgado, Patricia
Garrido-Vásquez, Jacob O. Sarfo, Susan Cooper, Anke C. Buttner, Tiago J. S. Lima, İbrahim
Öztürk, Nazlı Akay, Flavia H. Santos, Christina Artemenko, Lee T. Copping, Mahmoud M.
Elsherif, Ilija Milovanović, Robert A. Cribbie, Marina G. Drushlyak, Katherine Swainston,
Yiyun Shou, Juan David Leongómez, Nicola Palena, Fitri A. Abidin, Maria F. Reyes-
Rodríguez, Yunfeng He, Juneman Abraham, Argiro Vatakis, Matthew O. Parker, Mai Helmy,
Mariia M. Ostroha, Katie A. Gilligan-Lee, Felix O. Egara, Martin J. Barwood, Karuna
Thomas, Grace McMahon, Siobhán M. Griffin, Hans-Christoph Nuerk, Alyssa Counsell,
Oliver Lindemann, Dirk Van Rooy, Theresa E. Wege, Joanna E. Lewis, Stefan Agrigoroaei,
Erich W. Graf, Lisa A. D. Webster, and Andy P. Field.
Visualization: Jenny Terry, Tamas Nagy, Erin M. Buchanan, and Andy P. Field.
Writing - original draft: Jenny Terry.
Writing - review & editing: Jenny Terry, Robert M. Ross, Tamas Nagy, Mauricio Salgado,
Patricia Garrido-Vásquez, Jacob O. Sarfo, Susan Cooper, Anke C. Buttner, Tiago J. S. Lima,
İbrahim Öztürk, Nazlı Akay, Flavia H. Santos, Christina Artemenko, Lee T. Copping,
Mahmoud M. Elsherif, Ilija Milovanović, Robert A. Cribbie, Marina G. Drushlyak, Katherine
Swainston, Yiyun Shou, Juan David Leongómez, Nicola Palena, Fitri A. Abidin, Maria F.
Reyes-Rodríguez, Yunfeng He, Juneman Abraham, Argiro Vatakis, Kristin Jankowsky,
Stephanie N. L. Schmidt, Elise Grimm, Desirée González, Philipp Schmid, Roberto A.
Ferreira, Rozgonjuk Dmitri, Neslihan Özhan, Patrick A. O'Connor, Andras N. Zsido, Gregor
Stiglic, Darren Rhodes, Cristina Rodríguez, Ivan Ropovik, Violeta Enea, Ratri Nurwanti,
Alejandro J. Estudillo, Nataly Beribisky, Karel K. Himawan, Linda M. Geven, Anne H. van
Hoogmoed, Amélie Bret, Jodie E. Chapman, Udi Alter, Tessa R. Flack, Donncha Hanna,
Mojtaba Soltanlou, Gabriel Banik, Matúš Adamkovič, Sanne H. G. van der Ven, Jochen A.
Mosbacher, Hilal H. Şen, Joel R. Anderson, Michael Batashvili, Kristel de Groot, Matthew O.

Parker, Mai Helmy, Mariia M. Ostroha, Katie A. Gilligan-Lee, Felix O. Egara, Martin J.
Barwood, Karuna Thomas, Grace McMahon, Siobhán M. Griffin, Hans-Christoph Nuerk,
Alyssa Counsell, Oliver Lindemann, Dirk Van Rooy, Theresa E. Wege, Joanna E. Lewis,
Balazs Aczel, Conal Monaghan, Ali H. Al-Hoorie, Julia F. Huber, Saadet Yapan, Mauricio E.
Garrido Vásquez, Antonino Callea, Tolga Ergiyen, James M. Clay, Gaetan Mertens, Feyza
Topçu, Merve G. Tutlu, Täht Karin, Mikkor Kristel, Letizia Caso, Alexander Karner, Maxine
M. C. Storm, Gabriella Daroczy, Rizqy A. Zein, Andrea Greco, Erin M. Buchanan, Katharina
Schmid, Thomas E. Hunt, Jonas De keersmaecker, Peter E. Branney, Jordan Randell,
Oliver J. Clark, Crystal N. Steltenpohl, Bhasker Malu, Burcu Tekeş, TamilSelvan Ramis,
Stefan Agrigoroaei, Nicholas A. Badcock, Kareena McAloney-Kocaman, Olena V.
Semenikhina, Erich W. Graf, Charlie Lea, Fergus M. Guppy, Amy C. Warhurst, Shane
Lindsay, Ahmed Al Khateeb, Frank Scharnowski, Leontien de Kwaadsteniet, Kathryn B.
Francis, Mariah Lecompte, Lisa A. D. Webster, Kinga Morsanyi, Suzanna E. Forwood,
Elizabeth R. Walters, Linda K. Tip, Jordan R. Wagge, Ho Yan Lai, Deborah S. Crossland,
Kohinoor M. Darda, Zoe M. Flack, Zoe Leviston, Matthew Brolly, Samuel P. Hills, Elizabeth
Collins, Andrew J. Roberts, Wing-Yee Cheung, Sophie Leonard, Bruno Verschuere,
Samantha K. Stanley, Iro Xenidou-Dervou, Omid Ghasemi, Timothy Liew, Daniel Ansari,
Johnrev Guilaran, Samuel G. Penny, Julia Bahnmueller, Christopher J. Hand, Unita W.
Rahajeng, Dar Peterburg, Zsofia K. Takacs, Michael J. Platow, and Andy P. Field.
Acknowledgements
The authors would like to thank everyone that contributed to administration and data
collection for this project, including: Alissa Beath (Macquarie University); Andrés
Castellanos-Chacón (Universidad El Bosque); Dr. Christiany Suwartono (Universitas Katolik
Indonesia Atma Jaya); Edoardus Gilang Wardana, Muhammad Raihan, and Isyraq
Qurratul'Aini (Brawijaya University); Martina Daiser, Alexander Soell, and Tang Kuan
(University of Tübingen); Megan Davies (University of Surrey); Roberto Capelli, Lucrezia
Cavagnis, Jacopo Stringo, and Mariantonella Todaro (University of Bergamo); Bethany

Lewington, Robert Spell, and Iga Ewa Zlotucha (University of Sussex). We also thank Peter
Lugtig (Utrect University) for his advice on translation.
Conflict of Interest
The authors declare no conflict of interest associated with the publication of this
manuscript.
Funding Statement
Kristel de Groot was supported by Dutch Research Council (NWO); James M. Clay
was supported by ESRC (ES/P000673/1); Elizabeth Collins was supported by ESRC; Philipp
Schmid was supported by European Commission Horizon 2020 Grant (964728 JITSUVAX);
Robert M. Ross was supported by Macquarie University Research Fellowship; Andras N.
Zsido was supported by New National Excellence Program of the Ministry for Innovation and
Technology from the source of the National Research, Development and Innovation Fund
(OTKA PD-137588; ÚNKP-21-4); Jenny Terry was supported by School of Psychology PhD
Studentship, University of Sussex; Gabriel Banik was supported by Slovak Research and
Development Agency (APVV-17-0418); Matúš Adamkovič was supported by Slovak
Research and Development Agency (APVV-20-0319); Ivan Ropovik was supported by
Systemic Risk Institute (PRIMUS/20/HUM/009; LX22NPO5101); Mahmoud M. Elsherif was
supported by The Baily Thomas Charitable Fund; Mauricio Salgado was supported by The
National Agency for Research and Development (ANID), Ministry of Science, Technology,
Knowledge and Innovation, through the Centre for Research in Inclusive Education (PIA-
ANID CIE160009; 2017); Sophie Leonard was supported by University College Dublin, Ad
Astra Scholarship; Flavia H. Santos was supported by University College Dublin, Ad Astra
Start Up Scholarship.

References
Baker, S. L., Heinrichs, N., Kim, H.-J., & Hofmann, S. G. (2002). The Liebowitz social anxiety
scale as a self-report instrument: A preliminary psychometric analysis. Behaviour
Research and Therapy, 40(6), 701–715. https://doi.org/10.1016/S0005-
7967(01)00060-2
Baloğlu, M. (2002). Psychometric Properties of the Statistics Anxiety Rating Scale.
Psychological Reports, 90(1), 315–325. https://doi.org/10.2466/pr0.2002.90.1.315
Baloğlu, M., & Zelhart, P. F. (2007). Psychometric Properties of the Revised Mathematics
Anxiety Rating Scale. The Psychological Record, 57(4), 593–611.
https://doi.org/10.1007/BF03395597
Behr, D. (2017). Assessing the use of back translation: The shortcomings of back translation
as a quality testing method. International Journal of Social Research Methodology,
20(6), 573–584. https://doi.org/10.1080/13645579.2016.1252188
Behr, D., & Shishido, K. (2016). The Translation of Measurement Instruments for Cross-
Cultural Surveys. In C. Wolf, D. Joye, T. Smith, & Y. Fu, The SAGE Handbook of
Survey Methodology (pp. 269–287). SAGE Publications Ltd.
https://doi.org/10.4135/9781473957893.n19
Benson, J., & El‐Zahhar, N. (1994). Further refinement and validation of the revised test
anxiety scale. Structural Equation Modeling: A Multidisciplinary Journal, 1(3), 203–
221. https://doi.org/10.1080/10705519409539975
Beurze, S. M., Donders, A. R. T., Zielhuis, G. A., de Vegt, F., & Verbeek, A. L. M. (2013).
Statistics Anxiety: A Barrier for Education in Research Methodology for Medical
Students? Medical Science Educator, 23(3), 377–384.
https://doi.org/10.1007/BF03341649
Birenbaum, M., & Eylath, S. (1994). Who is afraid of statistics? Correlates of statistics
anxiety among students of educational sciences. Educational Research, 36(1), 93–
98. https://doi.org/10.1080/0013188940360110

Blackburn, G. (2015). Effectiveness of eLearning in statistics: Pictures and stories. E-
Learning and Digital Media, 12(5–6), 459–480.
https://doi.org/10.1177/2042753016653704
Block, J. (1995). A contrarian view of the five-factor approach to personality description.
Psychological Bulletin, 117(2), 187–215. http://dx.doi.org/10.1037/0033-
2909.117.2.187
Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the
multitrait-multimethod matrix. Psychological Bulletin, 56(2), 81–105.
https://psycnet.apa.org/doi/10.1037/h0046016
Carleton, R. N., Collimore, K. C., McCabe, R. E., & Antony, M. M. (2011). Addressing
revisions to the Brief Fear of Negative Evaluation scale: Measuring fear of negative
evaluation across anxiety and mood disorders. Journal of Anxiety Disorders, 25(6),
822–828. https://doi.org/10.1016/j.janxdis.2011.04.002
Carleton, R. N., Norton, P., & Asmundson, G. (2007). Fearing the unknown: A short version
of the Intolerance of Uncertainty Scale. Journal of Anxiety Disorders, 21, 105–117.
https://doi.org/10.1016/j.janxdis.2006.03.014
Chen, G., Gully, S. M., & Eden, D. (2001). Validation of a New General Self-Efficacy Scale.
Organizational Research Methods, 4(1), 62–83.
https://doi.org/10.1177/109442810141004
Chew, P. K. H., & Dillon, D. B. (2014). Statistics Anxiety Update: Refining the Construct and
Recommendations for a New Research Agenda. Perspectives on Psychological
Science, 9(2), 196–208. https://doi.org/10.1177/1745691613518077
Cruise, R., J., Cash, R. W., & Bolton, D., L. (1985). Development and Validation of an
Instrument to Measure Statistical Anxiety. American Statistical Association
Proceedings of the Section on Statistical Education, 4, 92–97.
Curran, P. G. (2016). Methods for the detection of carelessly invalid responses in survey
data. Journal of Experimental Social Psychology, 66, 4–19.
https://doi.org/10.1016/j.jesp.2015.07.006

Daker, R. J., Cortes, R. A., Lyons, I. M., & Green, A. E. (2020). Creativity anxiety: Evidence
for anxiety that is specific to creative thinking, from STEM to the arts. Journal of
Experimental Psychology: General, 149(1), 42–57.
https://doi.org/10.1037/xge0000630
D’Urso, E. D., Maassen, E., van Assen, M. A. L. M., Nuijten, M. B., De Roover, K., &
Wicherts, J. M. (2022). The Dire Disregard of Measurement Invariance Testing in
Psychological Science [Preprint]. PsyArXiv. https://doi.org/10.31234/osf.io/n3f5u
Field, A. P. (2014). Skills in Mathematics and Statistics in Psychology and Tackling
Transition (Higher Education Academy STEM Series).
Flake, J. K. (2021). Strengthening the foundation of educational psychology by integrating
construct validation into open science reform. PsyArXiv.
https://doi.org/10.31234/osf.io/vqjb8
Frederick, S. (2005). Cognitive Reflection and Decision Making. Journal of Economic
Perspectives, 19(4), 25–42. https://doi.org/10.1257/089533005775196732
González, A., Rodríguez, Y., Faílde, J. M., & Carrera, M. V. (2016). Anxiety in the statistics
class: Structural relations with self-concept, intrinsic value, and engagement in two
samples of undergraduates. Learning and Individual Differences, 45, 214–221.
https://doi.org/10.1016/j.lindif.2015.12.019
Gonzalez, O., MacKinnon, D. P., & Muniz, F. B. (2020). Extrinsic Convergent Validity
Evidence to Prevent Jingle and Jangle Fallacies. Multivariate Behavioral Research,
0(0), 1–17. https://doi.org/10.1080/00273171.2019.1707061
Grahe, J. E., Chalk, H. M., Alvarez, L. D. C., Faas, C. S., Hermann, A. D., & McFall, J. P.
(2018). Emerging Adulthood Measured at Multiple Institutions 2: The Data. Journal of
Open Psychology Data, 6(1), 4. https://doi.org/10.5334/jopd.38
Hanna, D., Shevlin, M., & Dempster, M. (2008). The structure of the statistics anxiety rating
scale: A confirmatory factor analysis using UK psychology students. Personality and
Individual Differences, 45(1), 68–74. https://doi.org/10.1016/j.paid.2008.02.021

Huang, J. L., Curran, P. G., Keeney, J., Poposki, E. M., & DeShon, R. P. (2012). Detecting
and Deterring Insufficient Effort Responding to Surveys. Journal of Business and
Psychology, 27(1), 99–114. https://doi.org/10.1007/s10869-011-9231-8
Hunt, T. E., Simms, V., Cahoon, A., & Muwonge, C. M. (2021). Socio-Cognitive-Affective
Barriers to Mathematics Education in Developing Nations. In W. Leal Filho, A. M.
Azul, L. Brandli, P. G. Özuyar, & T. Wall (Eds.), Quality Education (pp. 1–11).
Springer International Publishing. https://doi.org/10.1007/978-3-319-69902-8_128-1
Ioannidis, J. P. A. (2005). Why Most Published Research Findings Are False. PLOS
Medicine, 2(8), e124. https://doi.org/10.1371/journal.pmed.0020124
Kelley, T. L. (1927). Interpretation of educational measurements. New York, NY: World
Book.
Lawson, K. M., & Robins, R. W. (2021). Sibling Constructs: What Are They, Why Do They
Matter, and How Should You Handle Them? Personality and Social Psychology
Review, 25(4), 344–366. https://doi.org/10.1177/10888683211047101
Leary, M. R. (1983). Brief Fear of Negative Evaluation Scale [Data set]. American
Psychological Association. https://doi.org/10.1037/t00048-000
Liebowitz, M. R. (1987). Social Phobia. Modern Problems in Pharmacopsychiatry, 22, 141–
173.
Long, O., & Chalk, H. M. (2020). Belonging and Marital Perception Variances in Emerging
Adults With Differing Disability Identities. Psi Chi Journal of Psychological Research,
25(1), 22–29. https://doi.org/10.24839/2325-7342.JN25.1.22
Meade, A. W., & Craig, S. B. (2012). Identifying careless responses in survey data.
Psychological Methods, 17(3), 437–455. http://dx.doi.org/10.1037/a0028085
Miller, R. B., Greene, B. A., Montalvo, G. P., Ravindran, B., & Nichols, J. D. (1996).
Engagement in Academic Work: The Role of Learning Goals, Future Consequences,
Pleasing Others, and Perceived Ability. Contemporary Educational Psychology,
21(4), 388–422. https://doi.org/10.1006/ceps.1996.0028

Paechter, M., Macher, D., Martskvishvili, K., Wimmer, S., & Papousek, I. (2017).
Mathematics Anxiety and Statistics Anxiety. Shared but Also Unshared Components
and Antagonistic Contributions to Performance in Statistics. Frontiers in Psychology,
8. https://doi.org/10.3389/fpsyg.2017.01196
Papousek, I., Ruggeri, K., Macher, D., Paechter, M., Heene, M., Weiss, ElisabethM.,
Schulter, G., & Freudenthaler, H. H. (2012). Psychometric Evaluation and
Experimental Validation of the Statistics Anxiety Rating Scale. Journal of Personality
Assessment, 94(1), 82–91. https://doi.org/10.1080/00223891.2011.627959
Rad, M. S., Martingano, A. J., & Ginges, J. (2018). Toward a psychology of Homo sapiens:
Making psychological science more representative of the human population.
Proceedings of the National Academy of Sciences of the United States of America,
115(45), 11401–11405. https://doi.org/10.1073/pnas.1721165115
Ree, M. J., French, D., MacLeod, C., & Locke, V. (2008). Distinguishing Cognitive and
Somatic Dimensions of State and Trait Anxiety: Development and Validation of the
State-Trait Inventory for Cognitive and Somatic Anxiety (STICSA). Behavioural and
Cognitive Psychotherapy, 36(03). https://doi.org/10.1017/S1352465808004232
Richardson, F. C., & Suinn, R. M. (1972). The Mathematics Anxiety Rating Scale:
Psychometric data. Journal of Counseling Psychology, 19(6), 551–554.
http://dx.doi.org/10.1037/h0033456
Rodebaugh, T. L., Woods, C. M., Thissen, D. M., Heimberg, R. G., Chambless, D. L.,
Rapee, R. M., & Link to external site, this link will open in a new window. (2004).
More Information From Fewer Questions: The Factor Structure and Item Properties
of the Original and Brief Fear of Negative Evaluation Scale. Psychological
Assessment, 16(2), 169–181. http://dx.doi.org/10.1037/1040-3590.16.2.169
Schwab-McCoy, A. (2019). The State of Statistics Education Research in Client Disciplines:
Themes and Trends Across the University. Journal of Statistics Education, 27(3),
253–264. https://doi.org/10.1080/10691898.2019.1687369

Shenhav, A., Rand, D. G., & Greene, J. D. (2012). Divine intuition: Cognitive style influences
belief in God. Journal of Experimental Psychology: General, 141(3), 423–428.
https://doi.org/10.1037/a0025391
Spielberger, C. D. (1983). State-Trait Anxiety Inventory for Adults (STAI-AD) [Database
record]. APA PsycTests. https://psycnet.apa.org/doi/10.1037/t06496-000
Terry, J., & Field, A. P. (2022). A review of the theoretical explanations of the relationship
between statistics anxiety and statistics course outcomes. Manuscript in Preparation.
Terry, J., Lea, C., & Field, A. P. (2022). Statistics Anxiety and Maths Anxiety: Evidence of
another Jangle Fallacy. Manuscript in Preparation.
Tindall, I. K., Curtis, G. J., & Locke, V. (2021). Dimensionality and Measurement Invariance
of the State-Trait Inventory for Cognitive and Somatic Anxiety (STICSA) and Validity
Comparison With Measures of Negative Emotionality. Frontiers in Psychology, 12.
https://www.frontiersin.org/articles/10.3389/fpsyg.2021.644889
Trakulphadetkrai, N. V. (2017, February). Undergraduate students’ perceptions of learning
introductory statistics through producing a statistical picture book. CERME 10.
https://hal.archives-ouvertes.fr/hal-01927710
Trassi, A. P., Leonard, S. J., Rodrigues, L. D., Rodas, J. A., & Santos, F. H. (2022).
Mediating factors of statistics anxiety in university students: A systematic review and
meta-analysis. Annals of the New York Academy of Sciences, 1512(1), 76–97.
https://doi.org/10.1111/nyas.14746
Zeidner, M. (1991). Statistics and Mathematics Anxiety in Social Science Students: Some
Interesting Parallels. British Journal of Educational Psychology, 61(3), 319–328.
https://doi.org/10.1111/j.2044-8279.1991.tb00989.x