Maths anxiety is common and refers to feelings of anxiety, fear and other negative emotions and thoughts in individuals when confronted with mathematical tasks or numerical information. Selfreport measures of maths anxiety have been created, but the majority are in English and are not culturally relevant to all countries. This study aimed to translate and validate existing measures for future use in Hebrewspeaking adult populations. The Mathematics Anxiety Scale – UK (MASUK) was translated to Hebrew and adult participants completed it alongside the Mathematics Anxiety Rating Scale – Revised (MARSR) and a general anxiety measure. Factor structures were explored for both the translated MASUK and a Hebrew version of the MARSR, as well as being checked for reliability and convergent and discriminant validity. Results from a final sample of 213 participants, indicated the shortened, Hebrew version of the MASUK and the MARSR are internally consistent and suitable for use in future maths anxiety research in adult Israeli populations. Findings regarding sex differences in maths anxiety are also discussed.
Maths anxiety can be described as the feelings of tension and apprehension when manipulating numbers or completing mathsbased tasks (
The Mathematics Anxiety Rating Scale (MARS;
The MARSR (Plake & Parker) and its revision by Hopko, are two different instruments, which are often mixed up in the literature due to naming proposed by Hopko, who used the term “Revised MARSR”.
), and AMAS (There is much evidence to suggest that maths anxiety is a multidimensional construct (
Organisation for Economic Cooperation and Development (
It is important that educators and researchers adopt valid measurement tools for assessing maths anxiety in students and participants. To the best of the current authors’ knowledge, there have been few attempts to assess the validity of existing maths anxiety scales in Hebrew.
Following inclusion of alternative terms and items pertaining to maths situations that would be more familiar to individuals in the U.K.,
The MASUK has been translated and modified for use with University students in Iran (
Another measure of maths anxiety, the MARSR (
The aim of this study was to validate a Hebrew version of the MASUK (
Participants were 239 adults in Israel. Twentysix participants were removed as they had missing data on scale items. Of the remaining 213 participants, 1 nonstudent did not report their sex. Participants comprised 155 students enrolled at the Reichman University in central Israel (123 [79.35%] Women, Mean age = 23.92,
Student participants were recruited via the university’s online research participation system and nonstudents were recruited via social media (online completion of the scales was not anticipated to influence participants’ maths anxiety scores nor the factorial structure of maths anxiety; c.f.
The 23item translated version of the MASUK (
In addition to the translated version of the MASUK, a translated version of the MARSR (
The BAI (
Internal consistency checks, initial exploratory factor analyses and scale and subscale correlations were conducted in SPSS. Bayesian analyses and betweengroup comparisons were obtained using JASP. Parallel analysis (used alongside initial factor analyses) was conducted using Monte Carlo PCA for Parallel Analysis (Computer Software;
Responses to all the items of the MASUK are presented in
There was no significant difference in maths anxiety according to the MARSR,
Participant Group  Anxiety Measure 


MARSR 
MASUK 
BAI 

Total  Total  Total  
Student  21.79 (11.39)  1.82 (0.95)  53.17 (18.60)  2.31 (0.81)  13.80 (10.94)  0.66 (0.52) 
NonStudent  21.14 (11.31)  1.77 (0.95)  53.75 (19.92)  2.34 (0.87)  12.58 (9.14)  0.60 (0.44) 
Total  21.61 (11.34)  1.80 (0.95)  53.31 (18.88)  2.32 (0.83)  13.48 (10.45)  0.64 (0.50) 
Cronbach’s alpha was used to measure the internal consistency of all scales and subscales. However, to account for the ordinal nature of Likert scale data, ordinal alpha was also reported (
All MASUK items were correlated with the overall 23item scale total (minimum
Exploratory factor analysis was conducted on the 23item MASUK using principal axis factoring as the extraction method, with a direct oblimin rotation. Factors were determined using eigenvalues greater than 1. A high KaiserMeyerOlkin measure (KMO = .94) indicated sampling adequacy was met (
The analysis was rerun specifying the extraction of three factors. The pattern matrix was explored for factor loadings of .45 or more, as this is considered to be a good factor loading (
Negative correlations between factors originate from the analytical solution implemented in SPSS, while weighting of items shows they are correlated positively. This means that for interpretation of the correlations between scores one should ignore the sign of correlation value and item loadings.
The above analysis was conducted in SPSS to enable comparison to previous factor analyses typically reported in the extant literature. However, to account for the ordinal nature of the data, a further exploratory factor analysis was carried out in FACTOR, a standalone exploratory factor analysis package that has the ability to run polychoric correlations, rather than the standard Pearson correlations that are the default in SPSS. The procedure was followed using the recommendations of
Reliability analysis was conducted on the total scores for the Hebrew revised MASUK (HMASUKR), which showed that all 20 items were still found to be correlated with the overall scale (minimum
The same exploratory factor analysis method was applied to the Hebrew MARSR (KMO = .94). The correlation between extracted factors was .74 indicating nonorthogonality, justifying the decision to apply a direct oblimin rotation. Two factors were extracted, explaining a total of 65.80% of the variance, with 58.67% and 7.12% of the total variance being explained by both factors. All 20 items loaded onto one factor each (see
It should be noted that when determining the number of factors to extract, FACTOR recommended extracting one or two factors for the MASUK and one factor for the MARSR. We have kept the original three and two factor solutions to enable comparisons to previous research however, we do note that using polychoric correlations to account for the ordinal nature of the data is preferable moving forward.
To justify using total scores for subsequent analyses, factor scores were obtained for each maths anxiety scale and correlated with subscale total scores. There was a strong significant correlation between HMASUKR Everyday/Social Maths Anxiety total and factor scores,
There was a strong, positive, significant relationship between the 20item HMASUKR and MARSR totals,
Scale/Subscale  1  2  3  4  5  6  7  8 

1. HMASUKR  —  
2. HMASUKR Evaluation  .91 
—  
3. HMASUKR Social/Everyday  .89 
.68 
—  
4. HMASUKR Observation  .84 
.69 
.61 
—  
5. MARSR  .81 
.82 
.58 
.79 
—  
6. MARSR Evaluation  .81 
.76 
.60 
.84 
.97 
—  
7. MARSR Learning  .68 
.79 
.43 
.58 
.90 
.76 
—  
8. BAI  .29 
.31 
.20 
.28 
.32 
.29 
.31 
— 
^{a}Credible intervals were not estimated as BF was infinity.
Sex differences in MARSR, HMASUKR, and BAI total scores were explored. A significant difference in general anxiety was found,
Subscale  BF_{10}  Men 
Women 


HMASUKR Evaluation  6.16 (210)  < .001  >10^6  15.00 (6.28)  21.18 (6.01) 
HMASUKR Social/Everyday  3.63 (210)  < .001  67.24  13.79 (5.86)  18.02 (7.37) 
HMASUKR Observation  2.52 (210)  .013  3.18  8.81 (4.26)  10.82 (4.98) 
MARSR Evaluation  3.52 (210)  .001  48.02  8.17 (7.00)  12.51 (7.57) 
MARSR Learning  4.72 (210)  < .001  3635.61  7.53 (4.57)  10.82 (4.10) 
This study aimed to assess the validity and factor structure of a Hebrew version of the MASUK (
The strong positive correlation between the HMASUKR and the MARSR demonstrated high convergent validity, and the moderate positive correlation between the maths anxiety scales and the BAI showed high discriminant validity. The strength of the correlations between the HMASUKR and the BAI (
Significant differences according to sex were observed for both maths anxiety (including subscales) and general anxiety. Whilst there are some mixed findings in the literature, it is commonly shown that women report higher levels of maths anxiety than men (
The mean HMASUKR score was 2.36, which represents ‘slightly’ to ‘a fair amount’ of maths anxiety according to the corresponding response labels. This compares to a mean of 3.43 reported by
This is further supported by the lower mean MARSR maths anxiety score in our nonstudent sample (1.76) compared to nonstudent adults MARSR scores (2.30) in more Western countries (
To further validate the scale, future work should investigate the extent to which scores on the HMASUKR and MARSR predict behavioural outcomes in an Israeli population, such as maths performance. Reasons for the observed sex difference should be explored and testing should occur within the wider adult population. In addition, further research should assess the validity of Hebrew versions of existing selfreport measures of maths anxiety in a younger population. This study has provided evidence that the HMASUKR and MARSR are suitable measures of maths anxiety in an Israeli adult population. It is recommended that the scales are implemented in empirical work with this population. Furthermore, educators may find the tools useful to identify students who may require support in overcoming maths anxiety.
Item  Factor/Loading 


Everyday/ Social Maths Anxiety  Maths Evaluation Anxiety  Maths Observation Anxiety  
Adding up a pile of change  .104 (.128)  .066 (.079)  
Being asked to add up the number of people in a room  .001 (.081)  .078 (.070)  
Calculating how many days until a person’s birthday  .064 (.082)  .060 (.077)  
Being given a telephone number and having to remember it  .062 (.038)  .038 (.036)  
Working out how much time you have left before you set off to work or place of study  .080 (.673)  .188 (.185)  
Working out how much change a cashier should have given you in a shop after buying several items  .013 (.081)  .079 (.105)  
Deciding how much each person should give you after you buy an object that you are all sharing the cost of  .160 (.020)  .166 (.159)  
Being asked to calculate three fifths as a percentage  .126 (.295)  .297 (.163)  
Working out how much your shopping bill comes to  .031 (.039)  .117 (.120)  
Having someone watch you multiply 12 x 23 on paper  .438 (.454)  .049 (.067)  
Being asked to write an answer on the board at the front of a maths class  .128 (.214)  .049 (.008)  
Taking a maths exam  .038 (.011)  .237 (.359)  
Being asked to calculate £9.36 divided by four in front of several people  .385 (.465)  .019 (.016)  
Being given a surprise maths test in a class  .093 (.089)  .210 (.350)  
Being asked a maths question by a teacher in front of a class  .120 (.195)  .157 (.291)  
Listening to someone talk about maths  .112 (.133)  .019 (.004)  
Reading a maths textbook  .125 (.138)  .103 (.043)  
Watching someone work out an algebra problem  .042 (.066)  .016 (.078)  
Sitting in a maths class  .099 (.085)  .288 (.237)  
Watching a teacher/lecturer write equations on the board  .028 (..018)  .315 (.289)  
Removed Items  
Calculating a series of multiplication problems on paper  .390 (.381)  .291 (.352)  .296 (.306) 
Being asked to memorize a multiplication table  .429 (.396)  .240 (.350)  .196 (.189) 
Reading the word “algebra”  .379 (.469)  .117 (.307)  .344 (.548) 
^{a}As noted in the results section, negative correlations between factors originate from the analytical solution implemented in SPSS, while weighting of items shows they are correlated positively. Negative signs should be ignored for interpreting item loadings in this and following tables.
Item  Factor/Loading 


Everyday/ Social Maths Anxiety  Maths Evaluation Anxiety  Maths Observation Anxiety  
Adding up a pile of change  
Being asked to add up the number of people in a room  
Calculating how many days until a person’s birthday  
Being given a telephone number and having to remember it  
Working out how much time you have left before you set off to work or place of study  
Working out how much change a cashier should have given you in a shop after buying several items  
Deciding how much each person should give you after you buy an object that you are all sharing the cost of  
Being asked to calculate three fifths as a percentage  
Working out how much your shopping bill comes to  
Having someone watch you multiply 12 x 23 on paper  
Being asked to write an answer on the board at the front of a maths class  
Taking a maths exam  
Being asked to calculate £9.36 divided by four in front of several people  
Being given a surprise maths test in a class  
Being asked a maths question by a teacher in front of a class  
Listening to someone talk about maths  
Reading a maths textbook  
Watching someone work out an algebra problem  
Sitting in a maths class  
Watching a teacher/lecturer write equations on the board 
Item  Factor/Loading 


Maths Learning Anxiety  Maths Evaluation Anxiety  
Looking through the pages in a math text  .217 (.162)  
Having to use the tables in the back of a math book  .116 (.214)  
Thinking about an upcoming math test one day before  .095 (.099)  
Watching a teacher work an algebraic equation on the blackboard  .023 (.005)  
Being told how to interpret probability statements  .074 (.122)  
Picking up a math textbook to begin working on a homework assignment  .099 (.040)  
Taking an examination (quiz) in a math course  .028 (.065)  
Reading and interpreting graphs or charts  .282 (.355)  
Signing up for a course in statistics  .279 (.296)  
Waiting to get a math test returned in which you expected to do well  .036 (.016)  
Being given a “pop” quiz in math class  .044 (.076)  
Walking on campus and thinking about a 
.139 (.269) 
The authors would like to thank the three expert translators, Ortal Michaelashvily, Dr. Limor Shtoots and Professor Daniel Levy for their assistance with translating the MASUK. Thanks also go to Professor Gregory R. Hancock for advice regarding the exploratory factor analyses. The authors are also grateful for feedback provided by reviewers of an earlier version of this paper.
For this article, a data set is freely available (
Supplementary materials include the SPSS data and syntax files used for primary analyses, JASP outputs for Bayesian analyses, R scripts for figure creation and ordinal alpha calculation and a guide for using FACTOR factor analysis software (for access see
The authors have no funding to report.
The authors have declared that no competing interests exist.