The tripartite theory of attitudes or ABC Model of attitudes is underpinning this study. Every attitude has three components: Affect, Behaviour and Cognition.^18,19,20 Affect involves a person’s feelings, emotions about the attitude object; for example, I am afraid of scorpions. Behavioural (conative) involves the way we respond to the attitude object; for example, I will avoid scorpions and scream and run away if I see one. Cognitive component involves a person’s belief and knowledge about the attitude object; for example, I believe that scorpions are dangerous reptile.^21,22 According to Fennema and Sherman²⁰ in mathematics, affect (A) factor measures the following: attitude towards success – I would be proud and happy to do well in mathematics, confidence – I am confident in doing mathematics, anxiety – I am not anxious about doing mathematics. Behaviour (B) factor measure the effectance motivation of mathematics – the challenges of mathematics (problem solving) motivate me. Cognitive (C) factor measures the perceived usefulness of mathematics – I regard mathematics as useful in my life.

Home environment and school environment play an important role in shaping students’ attitudes towards mathematics^20,23 The mother or father component attempts to measure students’ perception of their mother’s and father’s influence in their mathematical ability. In the home factors, mother and/or father may encourage and affirm students’ efforts in mathematics. The teacher component attempts to measure students’ perception of their teacher’s influence in their mathematical ability. The school factors, such as mathematics teacher, play a significant role in the development of attitudes towards mathematics. Teachers influence their students in the classrooms.²⁰

Males versus Females: A study conducted by Farooq and Shah⁴ in Pakistan on high school students attitude towards mathematics found no gender significant differences in the confidence towards mathematics, usefulness of mathematics, mathematics as a male domain and mathematics teachers’ perception (p > 0.050). They concluded that gender differential has no impact on the attitude towards mathematics in Pakistan. Kosgey et al.²⁴ in their research on gender differences and attitudes towards learning of mathematics amongst secondary school students in Keiyo District, Kenya, found that both males and females have positive attitudes towards learning of mathematics, although males were more positively inclined than females. Students’ perceptions of parental, teachers’ and peer expectations were found to significantly influence gender differences and attitudes towards learning of mathematics (p < 0.050). However, a study performed by Asante²³ revealed a significant difference in attitudes towards mathematics between males and females (p < 0.050). High school male students showed a more positive attitude towards mathematics than female students Woodard²⁵ who examined the effects of mathematics anxiety on post-secondary development students’ gender, found that female students are significantly more anxious about mathematics than male students (p < 0.050). Correspondingly, a study by Yeo (2013)²⁶, found that mathematics anxiety exists amongst the high ability students in two secondary schools. Girls exhibited a higher level of mathematics anxiety than boys. Research by Arslan et al.²⁷ revealed a significant difference between female and male students in terms of students’ success in mathematics (p < 0.050). Female students showed a positive attitude towards success in mathematics than male students. Younger students versus older students: Frazier-Kouassi²⁸ discovered significant difference between younger and older students on the mathematics as a male domain (p < 0.050). The older students had a more positive male domain attitudes than the younger students. No further significant differences were found²⁸ in mathematics anxiety of traditional-aged students (< 25 years) and non-traditional aged students (≥ 25 years) (p > 0.050). Rural students versus urban students: The results of a study conducted by Frazier-Kouassi²⁸ revealed a significant difference between rural and urban students on the mother’s mathematics attitude (p < 0.050). No further differences were found on the other mathematics attitudes domains. Low-achieving students versus high-achieving students: Frazier-Kouassi²⁸ discovered significant differences between low-achieving and high-achieving students were found on the mathematics anxiety, effectance motivation (attitude towards problem solving) and usefulness mathematics domains (p < 0.050). No other significant differences were found.

A quantitative research design in the form of surveys was adopted for this study. Research design is a plan or blueprint of how one plans to carry out the research.²⁹ This study is located to a post-positivist paradigm. The study targeted all pre-service accounting teachers at a higher education institution with a population of 299 pre-service accounting teachers (first year, 179; second year, 85 and third year, 35). Convenience sampling method was used because the pre-service accounting teachers were based at the HEI and were easy to reach and easy to contact. Convenience sampling involves choosing participants that are easily accessible to the researcher.³⁰

The questionnaires were administered to the pre-service accounting teachers at a higher education institution in KwaZulu-Natal, South Africa. To facilitate maximum response and to analyse data with ease, I divided the questionnaire into two sections as follows: Section A: Biographical variables – This section enabled the researcher to obtain a personal profile of the participants such as age, gender, race, year of study, mother tongue, area of schooling, education level and previous education background and Section B: Fennema-Sherman Mathematics Attitude Scale (F-SMAS) – This section was divided into eight sections, each seeking to elicit pre-service accounting teachers’ attitudes to an aspect of mathematics.

The F-SMAS, with established reliability and validity was used for the study. The F-SMAS used are as follows: The attitude towards success in mathematics subscale, the mother subscale, father subscales, the teacher subscale, the confidence in learning mathematics subscale, the mathematics anxiety subscale, the effectance motivation in mathematics subscale, and the mathematics usefulness subscale.²⁰ Each of these scales contained 12 items, of which 6 measure positive attitudes towards mathematics and 6 measure negative attitudes towards mathematics. The five-point Likert scale was used: strongly agree, agree, not sure, disagree and strongly disagree. Each of the Likert responses was given a value of 5 to 1, respectively, for the positively stated questions and 1 to 5, respectively, for the negatively stated questions. The minimum possible score was 12, neutral possible score was 36. and the maximum possible score was 60. A higher score indicates more positive attitudes towards mathematics and a lower score indicates more negative attitudes towards mathematics. A questionnaire is a research tool consisting of a series of questions designed to generate a collection of written answers (by the respondents) to underlying issues, which the research seeks to investigate.³¹ The main aim of a questionnaire is to acquire facts and opinions about a phenomenon from participants.³² All allotted questionnaires were received on the same day. Some students opted not to participate in the research. The 255 received responses are as follows: first year, 143 (80%); second year, 77 (91%) and third year, 35 (100%). Incomplete questionnaires were removed to avoid bias. The objectives of the study were explained, and informed consent was signed by the participants. I stressed that responses will be treated as confidential, and only pseudonyms would be used.

Descriptive and inferential statistics were generated from the questionnaires with the assistance of the Statistical Package for Social Sciences (SPSS version 23). The results were interpreted by comparing the means with the domains and using t-test and analysis of variance (ANOVA) at p < 0.050, level of significance. The Cronbach’s coefficient alpha was used to establish internal consistency reliability. The reliability for the overall questionnaire was 0.961. The Cronbach’s alpha value of 0.961 is considered excellent. This study’s Cronbach’s alpha for each used domain of F-SMAS is as follows: The attitude towards success in mathematics (0.760), mother’s mathematics attitude (0.842), father’s mathematics attitude (0.913), mathematics anxiety (0.912), effectance motivation in doing mathematics (0.847), usefulness of mathematics (0.893), teacher’s mathematics attitude (0.824) and confidence in learning mathematics (0.922).

There were significant differences between mathematics and mathematical literacy students. In comparison with mathematical literacy students, the mathematics students reported a more positive attitude towards success in mathematics; a more positive perception of their father as a source of encouragement, support and affirmation of their efforts in mathematics; a more positive attitude towards usefulness of mathematics; a more positive perception of their teacher as a source of encouragement, support and affirmation of their efforts in mathematics and a more positive overall (F-SMAS total) mean scores of mathematics and mathematical literacy students’ attitudes towards mathematics (p < 0.050). No further significances were found amongst the two groups on remaining F-SMAS subscales (p > 0.050) (see Table 5).

No significant differences were found amongst first-, second- and third-year students (p > 0.050) (see Table 6).

The results of the ANOVA showed significant differences between rural, township and suburban students. The suburban students reported a more positive attitude towards success in mathematics; a more positive perception of their mother and father as sources of encouragement, support and affirmation of their efforts in mathematics; a more positive attitude towards effectance motivation in doing mathematics; a more positive perception of their teacher as a source of encouragement, support and affirmation of their efforts in mathematics; and a more positive overall mean scores of suburban area students attitudes towards mathematics than township and rural students (p < 0.050). No further significances were found amongst the two groups on remaining F-SMAS subscales (p > 0.050) (see Table 7).