Measuring math anxiety in Italian college and high school students: Validity, reliability and gender invariance of the Abbreviated Math Anxiety Scale (AMAS) Caterina Primi a, ⁎, Chiara Busdraghi a , Carlo Tomasetto b , Kinga Morsanyi c , Francesca Chiesi a a Neurofarba, Section of Psychology, University of Florence, Italy b Department of Psychology, University of Bologna, Italy c School of Psychology, Queen's University Belfast, UK abstract article info Article history: Received 26 July 2013 Received in revised form 5 April 2014 Accepted 22 May 2014 Keywords: Math anxiety Validity Reliability Educational level invariance Gender invariance Given that the ability to manage numbers is essential in a modern society, mathematics anxiety – which has been demonstrated to have unfortunate consequences in terms of mastery of math – has become a subject of increasing interest, and the need to accurately measure it has arisen. One of the widely employed scales to measure math anxiety is the Abbreviated Math Anxiety Scale (AMAS) (Hopko, Mahadevan, Bare & Hunt, 2003). The first aim of the present paper was to confirm the factor structure of the AMAS when administered to Italian high school and college students, and to test the invariance of the scale across educational levels. Additionally, we assessed the reliability and validity of the Italian version of the scale. Finally, we tested the invariance of the AMAS across genders. The overall findings provide evidence for the validity and reliability of the AMAS when administered to Italian students. © 2014 Elsevier Inc. All rights reserved. 1. Introduction Math anxiety is commonly defined as an adverse emotional reaction to math or the prospect of doing math (Hembree, 1990). It is a state of nervousness and discomfort brought upon by the presentation of math- ematical problems and may impede mathematics performance irrespective of true ability (Aiken, 1970; Ashcraft, 2005; Ashcraft & Moore, 2009). Math anxiety has been demonstrated to have unfortunate consequences on the mastery of math. Across a number of studies, individuals high in math anx- iety have been shown to perform more poorly than their low math anxious peers on a range of numerical and mathematical tasks (Ashcraft & Faust, 1994; Maloney, Ansari, & Fugelsang, 2011), and they showed lower levels of mathematics achievement (Aschcraft, 2002; Hembree, 1990). Further- more, individuals with mathematics anxiety have shown to avoid environ- ments and careers that require the use of math skills. From a developmental point of view, while children begin formal education with a very positive view of mathematics (Stevenson et al., 1990), both interest and motivation decline as they grow older, and math anxiety is thought to develop in junior high school, coinciding with the increasing difficulty of the math curriculum (Ashcraft, 2005). Given that being able to manage numbers is essential in a modern society, mathematics anxiety has become a subject of increasing interest (Ashcraft, Krause, & Hopko, 2007; Ashcraft & Ridley, 2005), and thus it is crucial to develop valid and reliable instruments to mea- sure math anxiety (Hopko, 2003; Nunez-Pena, Suarez-Pellicioni, Guilera, & Mercadè-Carranza, 2013). Pioneering work on the assess- ment of math anxiety involved the construction of the Math Anxiety Rating Scale (MARS; Richardson & Suinn, 1972). The MARS is a 98- item rating scale where respondents rate on a 1 to 5 Likert-type scale how anxious they would feel in situations which include both formal math settings (e.g., taking a pop quiz in a math class) and informal, ev- eryday situations (e.g., totaling a dinner bill to check if you were overcharged). Whereas several studies have documented its reliability and validity (for a review, see Capraro, Capraro, & Henson, 2001; McMorris, 2004), the 98-item MARS has some limitations arising from the length of the scale. Indeed, it is burdensome, time-consuming to ad- minister, and cumbersome to score (Ashcraft & Moore, 2009). Thus, fol- lowing Richardson and Suinn's pioneering work, researchers have engaged in the creation of several abbreviated instruments, such as the Fennema–Sherman Mathematics Anxiety Scale (MAS, Fennema & Sherman, 1976), the Sandman Anxiety Toward Mathematics Scale (ATMS; Sandman, 1980), the Math Anxiety Rating Scale — Revised (MARS-R; Plake & Parker, 1982), and the Abbreviated Math Anxiety Rating Scale (sMARS, Alexander & Martray, 1989). All these scales have the advantage of demanding less time than the original MARS, but, apart from a few recent studies, the psychometric properties of these scales have not been adequately tested (Hopko, 2003). For exam- ple, Nunez-Pena et al. (2013) have assessed the reliability and validity Learning and Individual Differences 34 (2014) 51–56 ⁎ Corresponding author at: Neurofarba -Section of Psychology Via di S.Salvi 12- Padiglione 26 50135 Firenze Italy. E-mail address: primi@unfi.it (C. Primi). http://dx.doi.org/10.1016/j.lindif.2014.05.012 1041-6080/© 2014 Elsevier Inc. All rights reserved. Contents lists available at ScienceDirect Learning and Individual Differences journal homepage: www.elsevier.com/locate/lindif