An investigation into fixed response questions in science at secondary and tertiary levels

This thesis focuses upon the effect of some factors on pupils'/students' responses in some assessment methods mainly fixed response questions. The major aim of the research is trying to find out these factors in order to improve the assessment process. The research was carried out over three years and the samples were selected from secondary schools both in the Glasgow area and in the Central Region of Scotland and the first year University of Glasgow students who were taking biology and chemistry courses. The results of the first investigation which was carried out to find out the effect of changing the positions of the key answer and the most plausible distractor (strong distractor) in the multiple choice questions (MCQs) showed that it can have a significant effect in the facility value of the question. By placing the most plausible distractor next to the key answer, students can make an effective discrimination which leads more often to the correct answer. One factor, which may be the cause of the difference in the facility value of the MCQs, has an "optical" origin. This factor was examined at the school level. The results of this examination revealed that the differences in the facility values of the MCQs disappears or is at least reduced when suitably treated. Because of the above differences in the facility values of MCQs, another fixed response method is offered which is called Structural Communication Grids (SCGs). However, this method was also investigated like the MCQs to find out if changing the positions of the options within the grid affect the facility value of the question. The results showed that this also brings, to some extent, a significant difference in the facility value of the grid questions but not to the some extent as the multiple choice questions. In the same area (i.e. related to the grid questions), pupils' ranking in two methods of scoring the grid questions was examined. The results showed that there was no significant difference between the two methods. The research is not only looking at the effects of some editing factors on the facility value of some fixed response questions but also to investigate the effects of some psychological factors. The results showed that field - independent pupils perform better than field - dependent pupils in the field response questions based on the grid and open response questions. On the other hand, it has been found that there is no significant difference in performance between convergent and divergent pupils in multiple choice questions (MCQs) and structural communication grids (SCGs). Another investigation was carried out to find out the relationship between pupils' knowledge structures and multiple choice test structure. The results revealed that the option in the question, which is closer to the concept on which the stem of the question is based, is more likely to be selected than other options. The final investigation was conducted to find out how pupils' performances in the word association test (WAT) is related to their achievement in the multiple choice questions. The results showed that there is a positive correlation between pupils' performances in WAT and their achievement in the multiple choice questions (MCQs). The facility value of the MCQs of those pupils who are more able to associate concepts in WAT is higher than the facility value of the MCQs for those pupils who are less able to associate concepts in WAT. Furthermore, in most cases it was found that the options, which have high response frequencies in WAT, as more likely to be selected among other options in multiple choice questions