DE eng

Search in the Catalogues and Directories

Hits 1 – 1 of 1

1
Cluster verification for listening learner strategies
Dobbie, Stacey. - : Victoria University of Wellington, 2015
Abstract: This thesis presents a comparison of statistical methodologies for cluster verification on ordinal response variables. Methodologies will be applied to a Listening Strategy dataset collected by the Language Learner Strategies research team at the National Institute of Education in Singapore. From this listening dataset, eight clusters suggested by Linguistics theory require verification. The methodologies undertaken is to find which listening strategies have been formed well. Methods used includes the proportional odds model, confirmatory factor analysis and ordinal agreement model. The proportional odds model is used to establish how well each cluster of questions is built. This is established by checking how similar questions within clusters are. The confirmatory factor analysis is used to verify how well the overall listening clusters have been built. This will be compared to clusters proposed by a statistical method. Lastly, the ordinal agreement model is applied to see how much agreement there is within each of the listening clusters. This will be able to show us which clusters is built better than the other clusters for this listening questionnaire. Results show that the prediction listening strategy has the highest level of agreement as well as no difference between questions within this cluster. The Socio-affective listening strategy has the lowest level of agreement and very strong evidence of a difference between questions within the cluster. This suggests that the prediction cluster has been formed better than the Socio-affective cluster.
Keyword: Cluster; Linguistics; Ordinal agreement model; Proportional odds model; Verification
URL: http://hdl.handle.net/10063/4725
BASE
Hide details

Catalogues
0
0
0
0
0
0
0
Bibliographies
0
0
0
0
0
0
0
0
0
Linked Open Data catalogues
0
Online resources
0
0
0
0
Open access documents
1
0
0
0
0
© 2013 - 2024 Lin|gu|is|tik | Imprint | Privacy Policy | Datenschutzeinstellungen ändern