Analysis of statistics study-units followed by non-statistics students

Abstract

The objective of this dissertation is to examine some of the problems students taking statistics courses encounter with a special emphasis on those who are not majoring in statistics. The two main tools used were statistical analysis of a questionnaire and examination marks. The questionnaire designed to gather students' opinion regarding the subject was administered to students not majoring in statistics but having statistics study-units as compulsory units. The responses from the questionnaires were analyzed using descriptive statistics, crosstabulation and log-linear modeling techniques. Internal Consistency and reliability of both questionnaires were statistically tested to check for consistent major attitudinal behavior. Examination marks of the ST021 students were analyzed through principal component factor analysis in an attempt to give a meaningful interpretation to the examination performance broken down by marks in the four questions. Contrasting students with more mathematical environment in their course with those with less (e.g. engineering and architecture as compared with social science students) reveals a poorer response from the latter. Results suggested that though classification of students into mathematically and nonmathematically inclined gives an initially important factor, limiting oneself to such a classification is simplistic. Appreciation of how statistics fits in their studies courses and eventually figures in their profession is generally low. This somehow confuses their perspectives. Performance in examinations confirmed that science students achieved the best performance session. FEMA and Social work students performed poorly. STOs disapprove of statistics study-units while ST As appreciate these credits. Science students, in particular show a much more positive attitude towards the subject. While it is clear that science students are experiencing minute problems, it is equally clear that a number of students with limited scientific background have psychological problems when studying statistics at university level. One suggestion, which transpires from, related consideration is that statistics should not be made compulsory.