Glossophobia or fear of public speaking is commonly experienced by tertiary learners during  classroom presentations. This study intends to examine the speech anxiety level experienced by engineering learners and the relationship between the learners’ English language proficiency and their speech anxiety. The study will also identify the causes of public speaking anxiety among these engineering learners. The samples were 65 Engineering Diploma students who responded to McCroskey’s Personal Report of Public Speaking Anxiety (PRPSA) questionnaire. The students’ English language grades at SPM level were used as the measurement of their English language proficiency. The findings showed the students’ speech anxiety is at moderate level (M= 105.9) and there was a weak negative relationship between total PRPSA scores and SPM English language grades (r = -.149, n = 65, p > 0.05). Mere thoughts of presenting a speech with undue worries of being asked questions were the causes of their speech anxiety. Therefore, it is proposed engineering learners be given ample public speaking trainings through planned programmes and engineering curriculum planners include elements of public speaking in engineering subjects. Public speaking skills will be an added value and increase engineering graduate employability in future. Keywords: Communication apprehension, Glossophobia, Public speaking, Speech anxiety, PRPSA

This study focuses on testing the moderating effects based on gender, academic performance, faculty and residential setting based on Self-Action-Outcome (SAO) model that was developed by the researchers. This model examined the relationship between self-esteem, job-search intensity and career decision-making self-efficacy among undergraduate students. Several instruments were used to develop this model were, the Rosenberg Self-Esteem Scale, Job-Search Intensity Scale, and Career Decision-Making Self-Efficacy Scale. Cluster sampling procedure was used in recruiting the participants. Data were collected using a self-reported questionnaire from 678 undergraduate students and were tested with confirmatory factor analysis to obtain three best-fit measurement models from the three latent variables. The results showed that gender, academic performance, faculty and residential setting did not moderate the hypothesized SAO model which means that the hypothesized SAO model is a good model for undergraduate students. Implications of the findings for social work practice such as guidance counsellor, research, theory, policy and education were also discussed.