This thesis examines the rational basis for how people search online consumer reviews in service of purchasing decisions. It considers the questions of (1) how prior uncertainty about the relative value of alternatives affects the depth of opinion search (2) whether people have a preference to obtain more reviews for one of the alternatives under consideration, and (3) whether people have a tendency to obtain specific reviews, and take varying time to read reviews, depending on reviews rating. Two sequential sampling probabilistic models are proposed to describe opinion seeking behaviour. The central assumption in both models is that people consult online opinions to discriminate between the choice's alternatives. The first model examines only when opinion seeking ceases. The second model, extends the first, and additionally examines what information people gather, by proposing an Optimal Experimental Design theory which operationalises the value of information as the extent that it discriminates between alternatives. It is also sensitive to the shape of the distribution of review ratings in the world. Model predictions are generated, and tested in five controlled experiments. The work demonstrates that viewing online opinion search as optimally increasing the discrimination between alternatives by acquiring reviews, explains the effect of prior relative uncertainty on the depth of information search, and the preferences for which alternative to seek information. It also explains the preference to obtain reviews of lower rating, and to take more time to inspect them for inferior alternatives. However, it does not explain what appears to be a similar tendency to take more time to inspect a review with lower rating for the best alternative under consideration - which is explained however by an outcome maximisation search strategy. It is suggested that a new Optimal Experimental Design theory that takes into account the value of information in both discriminating between alternatives and maximizing choice outcome may offer a comprehensive account of the data. 10
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:505482 |
| Date | January 2009 |
| Creators | Lelis, Stylianos |
| Publisher | University of Manchester |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
Page generated in 0.0111 seconds