UMD Theses and Dissertations
Permanent URI for this collectionhttp://hdl.handle.net/1903/3
New submissions to the thesis/dissertation collections are added automatically as they are received from the Graduate School. Currently, the Graduate School deposits all theses and dissertations from a given semester after the official graduation date. This means that there may be up to a 4 month delay in the appearance of a given thesis/dissertation in DRUM.
More information is available at Theses and Dissertations at University of Maryland Libraries.
Browse
2 results
Search Results
Item The Impact of Model Selection on Loglinear Analysis of Contingency Tables(2009) Gao, Jing; Dayton, C. Mitchell; Measurement, Statistics and Evaluation; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)It is common practice for researchers in the social sciences and education to use model selection techniques to search for best fitting models and to carry out inference as if these models were given a priori. This study examined the effect of model selection on inference in the framework of loglinear modeling. The purposes were to (i) examine the consequences when the behavior of model selection is ignored; and (ii) investigate the performance of the estimator provided by the Bayesian model averaging method and evaluate the usefulness of the multi-model inference as opposed to the single model inference. The basic finding of this study was that inference based on a single "best fit" model chosen from a set of candidate models leads to underestimation of the sampling variability of the parameters estimates and induces additional bias in the estimates. The results of the simulation study showed that due to model uncertainty the post-model-selection parameter estimator has larger bias, standard error, and mean square error than the estimator under the true model assumption. The same results applied to the conditional odds ratio estimators. The primary reason for these results is that the sampling distribution of the post-model-selection estimator is, in actuality, a mixture of distributions from a set of candidate models. Thus, the variability of the post-model- selection estimator has a large component from selection bias. While these problems were alleviated with the increase of sample size, the interpretation of the p-value of the Z-statistic of the parameters was misleading even when sample size was quite large. To avoid the problem of inference based on a single best model, Bayesian model averaging adopts a multi-model inference method, treating the weighted mean of the estimates from each model in the set as a point estimator, where the weights are derived using Bayes' theorem.Item ESSAYS ON MARKETING MODEL APPLICATIONS FOR ONLINE AND OFFLINE COMMUNITIES(2011) Gao, Jing; Kannan, P.K.; Zhang, Jie; Business and Management: Marketing; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Social interactions in a community influence perceptions and values of members of the community. Recently Web 2.0 technologies have stimulated rapid growth of online communities, where communications between participants are made much easier. It is important to study how participants' behaviors and preferences are affected by their communities. In my dissertation, I develop quantitative marketing models to empirically study perceptions and attitudes of participants in online and offline communities. Essay 1 examines an offline community, distributor community in multi-level marketing organizations. We propose a spatial model to understand the determinants of distributor satisfaction and simultaneously account for biases in measures in the context of cross-country marketing operations. We define an attribute-space using measures such as sales momentum and effort expended on business by distributors. The relationship between distributor satisfaction and its drivers varies within this attribute-space and across markets. Based on survey data from a large multi-national multilevel marketing firm, we empirically illustrate how marketing control variables impact distributor satisfaction scores across countries after controlling for biases. We also discuss the resource allocation implications based on the study. Essay 2 studies an online community, online bargain hunting forum. We investigate whether and how online discussions posted by active participants affect the interest and preference of the silent majority. We collect data from a major bargain hunting forum. Our analysis of the online discussions goes beyond measures of volume and valence, and delves into the specific contents of discussions posted in the forum. We classify the contents into a range of specific categories, and develop a Bayesian Poisson-Binomial model to examine how silent viewers' interest in and preference for a featured deal are influenced by the discussions, while controlling for many other factors. Our results show that the content of discussions posted by active participants indeed affects the silent viewers' interest in and preference for a featured deal, and that the effects are different across the specific categories of content. Our findings demonstrate that marketers can benefit from monitoring activities in online bargaining hunting forums, and suggest ways for them to participating in these forums.