Web Survey Bibliography

Title Satisfaction Scales in a CAWI Survey on University Teaching Evaluation
Author Fabbris, L.
Source International Conference on Questionnaire Development, Evaluation, and Testing Methods (QDET), 2002
Year 2002
Access date 26.05.2004
Abstract Computer assisted questionnaires are going to upheld the paper-and-pencil questionnaires in large scale and lasting surveys. The Internet may be the medium for questionnaire administration and data storage. Questionnaires need to be specifically designed for Computer Assisted Web-based Interviewing (CAWI) and methodological choices may differ even according to software facilities. We refer, in particular, to the availability of software for dynamic questionnaire programming and graphic presentation of questions and answers. A set of experiments was carried out at the University of Padua on Web-based questionnaires put forward for course evaluation by attendants. The experiments focused, among other trials, on * Scale construction for the measurement of students' satisfaction about several features of teaching quality, for each course currently attended. Scales experimented were the 4-point (two negative, two positive) ordinal scale, the same scale with a "neutral" category in the middle, the 1-7 scale, the 1-10 scale. * Approach for the evaluation of features of didactic activities and the interaction between types of scale and evaluative approaches. Two approaches were evaluated: the traditional "efficiency" approach, based on judgements offered by students on teacher and didactic environment, and the "effectiveness" approach, based on self-evaluation of learning as a reflection of teaching. The experiment informed about the following: 1) Experimented scales interact heavily with teaching evaluative approaches 2) Scales based on 4-point ordinal categories, with an equal number of positive and negative categories, are to be preferred to other scales if non responses may bias the distribution of responses. 3) The neutral category on an ordinal balanced scale is considered by students as an alternative to complaisant marks. Hence it is not to be considered "neutral", in the sense of neither positive nor negative, but a no-choice item. 4) 1-7 and 1-10 scales, presented as sets of equally-spaced within-square numbers, are preferable for almost all considered indicators (stability of averages and of gaps from maximum satisfaction levels, absence of no-choice category, independence from evaluative approach).
Access/Direct link Homepage - conference (abstract)
Year of publication2002
Bibliographic typeConferences, workshops, tutorials, presentations
Print

Web Survey Bibliography - International Conference on Questionnaire Development, Evaluation, and Testing Methods (QDET), 2002 (8)