Adaptation à la maladie de Lyme : comparaison des modes de collectes de données par téléphone et web

Authors: Ngambo Domche, Grace
Advisor: Valois, Pierre
Abstract: The web data collection mode has been increasing in popularity since its creation, primarily due to its cost and time effective benefits. However, its non-probabilistic tendencies often bring doubts concerning the representativity of its results in comparison to a typical mode such as the telephone. This master’s thesis compares the non-probability web and probability telephone modes, often suggested by polling firms in the province of Quebec. It principally analyzes the eventual differences between both samples on their adaptation to the Lyme disease. More specifically, the three main goals of this study are: (1) to compare the representativeness of both samples with the census data in terms of their socio-demographic characteristics, (2) to compare those two survey modes regarding their respondents’ preventive behaviour estimates against the Lyme disease, and (3) to test the measurement invariance of the nonprobability (Web panel) and probability (telephone) samples’ preventive behaviour. The data comes from a study on the adaptation of prevention behaviors against Lyme disease. The web (n=956) and telephone samples (n=1003) are made of individuals, 18 years old and above, from low and significant risk regions. Results firstly reveal that neither survey is representative of the Canadian census of 2016. Secondly, results show that the non-response rate for Lyme disease prevention adaptive behaviors and other associated variables is higher in the web survey. However, the magnitude of this significance is nil for 19 out of 30 the items, and small for the other 11. Moreover, no important dissimilarities were found in comparing both samples’ estimates in terms of the behaviours’ prevalence and their determinants. Finally, the measurement invariance analyses on the index of adaptation to the Lyme disease show a similar structure of the index for both samples. In conclusion, although the web and telephone samples are different in some respects, their respondents tend to have similar behaviors of adaptation to the disease.
Document Type: Mémoire de maîtrise
Issue Date: 2021
Open Access Date: 1 March 2021
Grantor: Université Laval
Collection:Thèses et mémoires

Files in this item:
Description SizeFormat 
36947.pdf1.31 MBAdobe PDFThumbnail
All documents in CorpusUL are protected by Copyright Act of Canada.