Relations of previous English achievement with college academic achievements as mediated by English learning beliefs

Learning is the result of many factors that interplay within individuals. That is why most learning variables affect academic performances of learners in one way or another though the strength of their predictive power could vary. The present study aimed to examine the effects of previous English achievement on English learning beliefs (ELBs) that would consequently affect college academic performances. Four hundred and seventy three re-spondents participated in the study through multistage sampling. A 15 items questionnaire with a 5-point Likert Scale was used to measure students' English learning beliefs. Self-reported Ethiopian Higher education Entrance Examination result of English achievement of students was used to represent their previous English results, while a 25 items teacher-made test was used to assess students' English achievement. First year first semester Grade Point Average was also used from students themselves. The principal factor analysis method produced three dimensions of English learning beliefs: combined beliefs about authority and difficulty, nature of English language learning, and risk-taking beliefs. Coefficient alpha values of the components ranged from 0.77 to 0.92. Multiple regression and structural equation modeling were employed to analyze the collected data. The results indicated that past English achievement had significant direct and indirect effects on English test scores and semester grade point average (SGPA) at college. The study discussed the results and educational implications were forwarded.