Faculty of Education - Theses
Permanent URI for this collection
Search Results
1 - 8 of 8
-
ItemThe measurement of curiosity : development of the state trait model of curiosity for children( 1976)The State Trait model of Curiosity developed by Drs. Naylor and Gaudry was modified for use with seventh grade subjects. Extra items were written, making a thirty--six item four choice Likert scale for trait and for state curiosity. 150 boys and 159 girls completed the trait scale, a word knowledge test, a curiosity stimulating task, and the state scale. The scales were analysed using the computer program Testat. Items were selected on the basis of their point biserial correlations. Splitting the subjects on their sex and vocabulary scores showed that many items had considerable bias. Reducing the reversed and the non-reversed subscales by removing small groups of items produced scales almost identical to those containing the items with the highest correlations. A factor analysis justified the distinction between state and trait curiosity and between reversed and non-reversed items. The loadings on the three factor Varimax rotation were highly correlated with the items' point biserial correlations. Sample-free item analysis attempts to analyse data so that the resulting statistics are independent of the sample used. Similarly it enables comparisons between individuals independently of the tests used. The program Calfit, based on this model, was used to analyse the data rescored at the three possible levels of curiosity. The trait scale fitted the model quite well but the state scale fitted poorly. This was due to the different behaviour of the three curiosity levels and the interaction of a subject's curiosity and the response on many items. Testat proved more effective for item selection but Calfit enabled the linking to previous research. Scales for future use are recommended.
-
ItemThe effects of trait anxiety, mathematical ability, method of testing, task difficulty, and their interactions, on state anxiety and performance in mathematics at primary school( 1979)A review was made of the theoretical and research literature related to the trait-state concept of anxiety, the relationship of anxiety to academic performance; and some factors confounding the anxiety, performance relationship. A study investigating the differential effect of anxiety on performance under two different methods of testing was reviewed. A study to replicate the findings related to anxiety, performance and varying test precedures was proposed. Data collected in the study made possible an investigation of Spielberger's Trait-State concept of anxiety and Spence's Drive Theory. One hundred and seventy six male, State primary school children drawn from sixteen grade five and six classes took part in the study. All testing was carried out in term I of 1978. As hypothesized the formal method of testing proved to be more anxiety arousing than the informal method of testing and students performed better under the latter testing procedure. Low anxious students performed better than high anxious students under both testing conditions. The study provided further support for Spielberger's Trait-State anxiety theory with high trait anxious students displaying higher levels of state anxiety than law trait anxious students, confirming that the State-A scale is a sensitive device for measuring differential anxiety arousal. The hypotheses related to Spence's Drive Theory were only partially supported by the results. Explanations as to why the interaction between task difficulty, ability and anxiety might not have been significant in this study were discussed. In general this study found evidence confirming the notion that high anxiety hinders performance, particularly for low ability students.
-
ItemThe effects of item wording on item characteristics on the state-trait anxiety inventory and the state-trait curiosity inventory( 1977)The state/trait distinction originally identified by Cattell ard Scheier has been used as a basis for the development of the State Trait Anxiety Inventory (STAI) and the State Trait Curiosity Inventory (STCI). Several studies have shown the state-trait distinction to be operationally meaningful in both anxiety and curiosity research. However an unexpected finding of factor analytic studies using these instruments has been the consistent emergence of a factor defined by the reversed items of the scales. Previous studies looking at this reversed items factor have offered explanations in terms such as stress level, which are external to the actual measuring instruments. Whilst the explanations offered for this occurrence were usually consistent with the data of the study, they were usually inconsistent with the explanations and findings of other studies. It seems that the only common element in studies examining the reversed items factor is the items themselves. Thus the present study has examined the characteristics of the STAI and STCI in terms of the items of the scales rather than in terms external to the instruments. Previous studies have noted that the reversed and non-reversed items characteristically display different means and variances, usually with higher means and variances associated with reversed items. Item means and variances were examined in the present study as the dependent variables, with the independent variables being the direction of item wording and the item content. The separation of direction of item wording and item content was achieved by rewording each item of the original scales to point in the opposite direction, thus providing for each item, one reversed and one non-reversed form. Ninety-six teachers' college students were randomly assigned to one of the four combinations of direction and content so produced, and their responses to the scale) provided item means and variances for each of the combinations of treatment level. A two-way analysis of variance was carried out for each scale on both the item means and the item variances so derived. The experimental design also provided for the incorporation of a validation study identical in method to the experimental study but based on a sample independent of the experimental study. It was found in two separate sets of data, that direction of item wording and item content significantly affect item means and variances, and that these factors have a significant interactive effect over and above their significant individual effects. These findings were viewed as throwing considerable doubt on the interpretation of the scales of the STAI and STCI in terms of anxiety and curiosity respectively, since these scores have been shown in the present study to be significantly affected by the actual wording of the items themselves.
-
ItemAn investigation of some educational implications of Holland's theory of vocational choice( 1976)The project set out to investigate the value of Holland's Vocational Preference Inventory (V.P.I.) as a measure of tertiary course/student compatibility and whether this compatibility had a bearing on subsequent course achievement. The sample consisted of 230 first year students at the Gordon Institute of Technology - a vocationally oriented College of Advanced Education. In addition to the course guidance investigation, a number of aspects of Holland's theory of vocational choice were examined. These included a test of Holland's proposed hexagonal pattern of relationships between the six personality types used in the interpretation of V.P.I. scores. Also, Holland's theory predicts, in relation to educational behaviour, that the choice of, stability in, satisfaction with and achievement into a field of study depend upon the degree of congruence between students and their educational environment. With a questionnaire measure of satisfaction and G.P.A. as an academic achievement measure, these predictions were tested. The results of a multiple discriminant analysis of scale scores on the V.P.I. indicated that meaningful distinctions between the course groups in the sample could be made on the- basis of scores on three significant discriminant functions. A series of planned contrasts carried out using a multivariance analysis also yielded significant results - again indicating that distinctions could be made between the course groups -using scores on the V.P.I. The values for the mean intercorrelations between scale scores for the sample showed that Holland's hexagonal ordering of the personality types was consistent with the results for this sample. As regards the congruence and satisfaction prediction, no significant differences were found on the satisfaction measures between those students who were congruent with their course and those who were not. Similar non-significant differences were found for the congruence/incongruence and academic achievement results. In both cases, however, satisfaction and achievement scores were higher for those who were congruent i.e. the differences were in the predicted direction. In -summary, the project found the V.P.I. to be able to discriminate between students in different courses and thus to offer possibilities for course guidance, and weak but not significant support for Holland's predictions relating to the effect of congruence of students and courses on educational behaviour.
-
ItemThe development and validation of an instrument to assess environmental attitudes( 1976)This study has attempted to produce valid and reliable scales for the measurement of environmental attitudes. There have already been a great many studies published which were concerned with the assessment of various. aspects of environmental attitudes. However, few of these studies seem to have undertaken any field testing of the scales used and consequently, the , validity of the scales used may be questioned. Furthermore, very few of the tests were published, which makes it difficult for other workers to confirm the published results. A group of four scales were developed in this study and details of their validity and reliability are presented. The attitude scales were derived from a pool of Likert - type items using an oblique rotation factor analysis of the item responses of senior high school and college students. The first order factor correlation matrix was subjected to a second oblique factor analysis and the factor attitude scales were constructed on the basis of the second order factors.. These scales were then used to investigate the attitudes of students at Rusden State College. Overall it was shown that Environmental Studies students had more favourable attitudes towards the environment than non - Environmental Studies students. No significant differences could be found between the-first and third year students. This could possibly reflect the increased incidence of environmental studies programmes in schools. The scales developed in - this study could prove to be useful in the measurement of environmental attitudes. Future research work in this. area should concentrate more on the relative effects on. environmental attitudes of influences other than environmental education programmes, such as parental and home influences and personal attributes. Thus it would be possible to determine the nature of individuals who need environmental education and the relative impact of different environmental education courses on different groups of students. Also this information would aid the design of environmental education programmes so that each group of students would be presented with the most effective programme.
-
ItemThe measurement of readability: theory and research in the measurement of the difficulty level of reading materials( 1973)The present submission consists of three papers. The first paper is a review of theory and research in the field of readability. The emphasis is upon approaches to the measurement of readability, and the theoretical bases of such approaches. Two broad types of approach are considered: analytic or formula approaches, and synthetic approaches. The second paper is concerned with the reliability and validity of the Readability Reference Scale (developed by Anderson). This paper reports research which focusses upon the Readability Reference Scale in comparison with formula approaches. The third paper, also a research report, describes research which seeks to assess the applicability of the Rasch model (a probabilistic model for item analysis) to the measurement of readability.
-
ItemChildren's understanding of literature( 1977)Sixty primary school children aged seven, nine and eleven years listened to and read a selection of six pieces of literature at three different difficulty levels. Children were interviewed individually about the literature. In interview, a flexible individualized approach was combined with some standardization to the extent that each child was asked seven key questions: two were literal questions, the remaining five were different types of inference as listed in the Barrett taxonomy. Responses were scored according to Piagetian stages of cognitive growth indicated on a scale developed by Hallam (1966, 1967). Data were analyzed in a mixed model design, with two factors on the subjects - age and sex, and three factors in the repeated measures - difficulty level, mode and cognitive demand. A multivariate analysis approach to the repeated measures was utilized (Finn, 1972). That there is a progression of cognitive growth in response to literature was established both statistically and descriptively. For literal and inferential cognitive demand a significant interaction was found. Mode of presentation of the literature was found not to be significant for inferential cognitive demand but was significant at the level of literal demand. No significant sex differences were established. Educational implications of the results are discussed.
-
ItemHome environment, spelling ability and types of spelling errors( 1978)This study sought to investigate spelling ability and types of spelling errors, and their relationships with home environment, as a means to finding ways in which parents and teachers can improve the spelling achievement of children. The sample consisted of 230 Grade 6 pupils in Canberra, who were given tests relating to spelling, reading and intelligence, and whose parents were interviewed to obtain information on their home environment. These test and questionnaire data were collected by Keeves (1972), but re-analysed and refined as part of the present study. For example, errors on the spelling test were classified into six types, such classification proving highly valid and reliable. Path analysis models were the basic tool used to investigate propositions. Each model included sex, three dimensions of home environment - home structure, parental attitudes and home processes, and a criterion drawn from spelling ability and the six spelling error types. To investigate other propositions groups were formed whereby spellers of different ability were grouped according to the extent of their reading or intellectual abilities. A major result was that the home environment dimension of home processes was shown to have greater effect on spelling ability than either home structure or parental attitudes. A number of positive methods were revealed by which parents and teachers could influence home processes, and hence spelling achievement. For example, parents could 'hear' their children read regularly, (say) nightly from their school readers, whilst teachers could make sure children borrowed school library books to read at home. Significant group differences were revealed for the five major spelling error types, and for a number of home environment variables. A sex difference was found showing that not only did girls appear to be superior spellers to boys on total spelling errors, but that they made less errors of each spelling error type.