Individual Submission Summary
Share...

Direct link:

Poster #83 - The role of cognitive, non-cognitive and situational factors on mathematic achievement

Thu, March 21, 4:00 to 5:15pm, Baltimore Convention Center, Floor: Level 1, Exhibit Hall B

Integrative Statement

Solving mathematic tests is a complex task that involve several abilities. There is increasing evidence that several factors, including cognitive, non-cognitive and situational factors play an important role on mathematic achievement (Giofrè, Borella & Mammarella, 2017; Pekrun, Elliot, & Maier, 2009). Among these factors, intelligence, math anxiety, self-esteem and teacher-student relationship quality have shown to be consistently related to mathematic achievement (Chang & Beilock, 2016). Nevertheless, to be best of our knowledge, no study has investigated jointly the role of these factors on mathematic achievement, by examining their relations at the latent level (using structural equation modeling). The present study addresses this gap, by examining the joint role of cognitive (intelligence), non-cognitive (math anxiety and self-esteem) and situational (teacher-student relationship quality) factors on mathematic achievement.
The sample consisted of 219 participants attending secondary schools, sixth grade. Children with clinical diagnoses and that belonged to disadvantaged socio-cultural group were not included in the study (4 children). Thirty-six children were absent in at least one session and they were excluded from the analysis Therefore, the final sample included a total of 181 children (Females = 49%, Age = 10.65, SD = 0.49). All children completed in three collective sessions a battery, as following: Cattell Culture Fair Intelligence Test (Cattell & Cattell, 1981); Primary Mental Abilities (PMA; Thurstone & Thurstone, 1963); The MeMa Test (Caponi, Cornoldi, Falco, Focchiatti et al., 2012); Multidimensional Self Concept Scale (Bracken, 2003); Student-Teacher Relationship Questionnaire (Tonci, De Domini, & Tomada, 2012).
The R program (R Core Team, 2016) with the “lavaan” library (Rosseel, 2012) was used to perform CFA (confirmatory factor analysis) and SEM (structural equation modeling) models. Model fit was assessed using various indexes according to the criteria suggested by Hu and Bentler (1999). In the first CFA model, we hypothesized the presence of 6 latent variables: intelligence (g), anxiety (ANX), self-esteem (SE), teacher-student relationship quality (TSRQ), mathematic achievement (MAT). The fit of this model was adequate, χ2(109)=178.71, p<.001, RMSEA=.06, SRMR =.07, CFI=.93, NNFI=.92, AIC=18191. A subsequent models showed that g and ANX were explaining a unique portion of the MAT variance, while TSRQ was not. In this model, ANX was mediating the effects of both SE and TSRQ on mathematic achievement. This models was more parsimonious compared to the previous one, χ2(113)=178.99, p<.001, RMSEA=.06, SRMR =.06, CFI=.94, NNFI=.92, AIC=18192, Δχ2(4)=0.28, p=.991, and was retained as our final model. In agreement with previous evidence, we found that mathematic achievement was primarily predicted by cognitive factors but it also depended on other non-cognitive factors (Chang & Beilock, 2016). We also found that others situational factors are involved in the explanation of mathematic achievement and that the effect of math anxiety can be mitigated by other protective factors, such as the teacher-student relationship quality.

Authors