The purpose of the current study was to investigate the structural relationships among self-reported mathematics achievement, attitudes toward statistics, and statistics outcomes by testing a structural model. The current study utilized a survey design. The participants of study consisted of 247 undergraduate and graduate students enrolled in statistics courses in a university in Turkey. The participants were from different disciplines such as engineering, education, and economics. The Turkish version of the Survey of Attitudes toward Statistics-36&copy / (SATS-36&copy / ) was used to collect data. The SATS-36&copy / assessed six components of statistics attitudes: cognitive competence, value, difficulty, effort, interest, and affect. Higher scores of the six components referred to the more positive attitudes. In addition, the SATS-36&copy / involved additional items to measure students&rsquo / self-reports of mathematics achievement and statistics outcomes. Results of the descriptive statistics analyses revealed that participants of the study had positive attitudes toward statistics except that they had neutral perceptions about the difficulty of statistics and neutral interest in statistics. Statistics outcomes variable was significantly correlated with mathematics achievement, affect, value, interest, and effort variables. Structural equation modeling was used to test the hypothesized structural regression model. Results indicated that affect, value, cognitive competence, and interest variables had large total standardized effects on statistics outcomes variable. Mathematics achievement and the effort variables had small total effects on explaining statistics outcomes. Difficulty had no statistically significant total effect on explaining statistics outcomes. Overall, the hypothesized structural regression model explained 66% of the total variance in statistics outcomes, which was statistically significant.

Despite agreement among researchers about the powerful influence of attitudes and beliefs on the development of students’ mathematical knowledge base (Leder, Pehkonen, & Törner, 2002), relatively little is known about these constructs in statistics education. This study investigated the relationship between mathematics-and statistics-related attitudes and beliefs of 11 high school students in an introductory statistics course designed around a 13-week long service-learning project. Service-learning is a pedagogical approach that situates academic learning in the context of community service. The study utilized qualitative, teacher-researcher (Cochran-Smith & Lytle, 1993) methodology from an interpretivist perspective. The three primary modes of data collection were journals, narratives, and an open-ended survey (Survey of Mathematical and Statistical Affect). Observations and reflections were also recorded regularly in a researcher journal. Inquiry adhered to guidelines for trustworthiness and rigor as outlined by Lincoln and Guba (1985). Item, pattern, and structural levels of analysis were employed (LeCompte and Schensul, 1999b). Investigation into attitudes and beliefs was framed in accordance with Op t’ Eynde, De Corte, and Verschaffel’s (2002) conceptualization of the mathematics-related belief system and McLeod’s (1992) framework of the affective domain in mathematics education. Results indicate that participants’ attitudes toward mathematics and statistics tended to converge while participants’ beliefs regarding mathematics and statistics tended to diverge. Participants like mathematics and statistics that involve real-life scenarios. Participants also like mathematics and statistics that do not require complex mathematical tasks. Participants’ beliefs regarding statistics were generally more positive than beliefs regarding mathematics. Participants reported greater confidence doing statistics than mathematics and contribute this confidence, in part, to service-learning. Participants also experienced a heightened sense of social awareness and social responsibility through the service-learning project. These results provide evidence that service-learning can be utilized to solidify positive attitudes and beliefs regarding statistics among high school students, in spite of potentially less positive ones toward mathematics.