Abstract

Successive governments have over the years made some tremendous strides in encouraging students to enroll in degree programs in mathematics or statistics. However, academic performance and graduate’s outputs don't appear to be improving since entrants lack the necessary knowledge in mathematics and statistics to pursue their university and college education in Sierra Leone. This is a result of unaccredited colleges and universities, the widespread use of forged credentials like M.Sc., MA., and other degrees, as well as fake BECCE and WASSCE certificates which resultantly give a mismatch in graduate’s performance and academic outputs of students in our institutions.
It is important to underscore that, Google form was designed and administered to collect the data from five universities and one college targeting first-year undergraduate minors in mathematics or statistics who submitted 1502 responses, but 1,494 responses constituting 749 (50.1%) males and 745 (49.9%) females in Sierra Leone were ultimately analyzed. Four psychological constructs in mathematics and statistical literacy skills such as mathematics and statistics perceptions, statistical anxiety, student's self-efficacy for statistical literacy, and computer self-efficacy as factors that might be greatly responsible.
In order to understand the effects and suggesting potential solutions to enhancing academic performance and graduate’s outcomes, these factors require a thorough investigation of the relationship, interaction, and association which remain crucial in Sierra Leone. AICs, BICs, Pseudo R-square, and deviances were used in the initial investigation to observe the behavior and to determine the best model outputs from ordered multiple/multivariable, multinomial, and mixed effects logistic regression models. The relationships between the latent psychological constructs were established through confirmatory factor analysis using the mixed effects logistic regression output, which produced the following model fit indices: [P-value = 0.00, CFI= 0.914, GFI = 0.964, RMSEA = 0.062, NFI = 0.900, TLI = 0.905, Chi-square/degree of freedom=0.1504]. Findings according to confirmatory factor analysis indicated that Computer self-efficacy is positively related and acted as both endogenous and exogenous variables to statistical anxiety and students’ self-efficacy for statistical literacy. Findings further revealed that Students’ self-efficacy for statistical literacy had a positive mediating relationship that acted as endogenous and exogenous variables to mathematics and statistics perceptions, while statistical anxiety and computer self-efficacy were used to explain the variations of mathematics and statistics perceptions of students.
From a sample of 364 first-year undergraduate mathematics and statistics minor students at the university of Makeni (UniMak), using structural equations modeling to determine a better predictor between mathematics and statistics perceptions, and statistical anxiety to academic performance(GPA). It was established that mathematics and statistics perceptions have a direct relationship with statistical anxiety as the significant predictor of academic performance in Sierra Leone with the model fit indices: GFI=0.994, RMSEA=0.067.
Mathematics and statistics academics, students, the Government of Sierra Leone, the private sector, and mathematics and statistics support services could benefit from the research to improve students learning and instill greater confidence in their understanding of computers and statistical applications to improve the already watered education system in this global data age to foster economic prosperity in Sierra Leone.