Regression Analysis of American Hotels Having Price as Dependent Variable
ABM PROJECT 2013
1. INTRODUCTION AND SUMMARY OF RESEARCH HYPOTHESIS
It is common knowledge that the prices people have to pay for accommodation in hotels vary enormously. Furthermore, hotel revenue managers probably posses or more or less accurate intuition of what causes room rates to diverge. However, they do not know how Online Travel Agent sites select the leading hotels to be placed on their first search page. In this respect, some determinants are expected to be associated with hotel prices in a more or less linear way. To say it differently, price differences between hotels underscore the presence or not of some variables expected to influence the latter. It is essential for hotels to understand how they can price their rooms and maximize yield while remaining competitive. Therefore, we conducted an extensive analysis to help hotel revenue managers find out what key variables influence price on Orbitz. The data were gathered from Orbitz.com directly. The data is about 1623 hotels that are located in 8 different geographical markets in the United States: Atlanta, Chicago, Los Angeles, Las Vegas, New York, Orlando, Honolulu, and Myrtle Beach. In section 1, we will present a Causal Relationship Scheme (CRS) that will be used as the framework of our research establishing the different relationships among our selected variables. Section 2 of the report contains the results of our univariate analyses. The distribution, mean and amount of variation of the data will be discussed. The subsequent section then introduces our findings from the bivariate analyses, investigating the presence, nature and strength of the relations between the pairs of variables described in the CRS. The fourth section presents the results of three multivariate analyses consisting first of a 2-factor ANOVA analysis that involves the response variable price and two other factors belonging to the CRS of the report. Secondly, we will formulate a regression model based
1. INTRODUCTION AND SUMMARY OF RESEARCH HYPOTHESIS
It is common knowledge that the prices people have to pay for accommodation in hotels vary enormously. Furthermore, hotel revenue managers probably posses or more or less accurate intuition of what causes room rates to diverge. However, they do not know how Online Travel Agent sites select the leading hotels to be placed on their first search page. In this respect, some determinants are expected to be associated with hotel prices in a more or less linear way. To say it differently, price differences between hotels underscore the presence or not of some variables expected to influence the latter. It is essential for hotels to understand how they can price their rooms and maximize yield while remaining competitive. Therefore, we conducted an extensive analysis to help hotel revenue managers find out what key variables influence price on Orbitz. The data were gathered from Orbitz.com directly. The data is about 1623 hotels that are located in 8 different geographical markets in the United States: Atlanta, Chicago, Los Angeles, Las Vegas, New York, Orlando, Honolulu, and Myrtle Beach. In section 1, we will present a Causal Relationship Scheme (CRS) that will be used as the framework of our research establishing the different relationships among our selected variables. Section 2 of the report contains the results of our univariate analyses. The distribution, mean and amount of variation of the data will be discussed. The subsequent section then introduces our findings from the bivariate analyses, investigating the presence, nature and strength of the relations between the pairs of variables described in the CRS. The fourth section presents the results of three multivariate analyses consisting first of a 2-factor ANOVA analysis that involves the response variable price and two other factors belonging to the CRS of the report. Secondly, we will formulate a regression model based