- 學位論文
以Ordinal logistic regression model探討影響數位電視「隨選視訊」服務滿意度之關鍵因素
Applying ordinal logistic regression model to study key factors of enhancing customer satisfaction in video on demand service of the digital cable television
摘要
雖然經濟部工業局2011年數位電視加值服務推動計畫調查報告顯示，「隨選視訊」服務使用率僅次於「高畫質節目」，且消費者對「隨選視訊」的需求度也高於其他數位電視加值服務，然而在台灣推動電視數位化逾十年的今日，「隨選視訊」的顧客採用率始終難有起色。為協助業界及政府了解提升數位電視「隨選視訊」服務滿意度之關鍵因素，本研究分析資策會所提供之「數位電視加值服務需求度調查」的部分資料，探討「品質滿意度」及「知覺價值」對顧客整體滿意度的影響。因為本研究所取得之實證資料為順序型資料，且不符合迴歸模型的常態假設，因此本研究以次序羅吉斯迴歸模型，進行資料分析。本研究結果證實，品質滿意度與知覺價值會影響顧客整體滿意度，即業者應努力提升「隨選視訊」之品質並加強顧客的知覺價值。因為實證資料的特質，次序羅吉斯迴歸模型也比多元迴歸擁有較佳的估計準確率。
並列摘要
Although a report of Industrial Development Bureau, Ministry of Economic Affairs, shows that, “video on demand (VOD)”, next to “high definition program”, is the value-added service of digital television consumers most frequently used, consumer demand for VOD is also higher than other value-added service of digital television, the consumer adoption of VOD is not improved after more than ten years of government promotion efforts. To help digital television service providers and Taiwan government enhances VOD adoption, this research analyzed data provided by Institute for Information Industry under the assumption that quality satisfaction and perceived value will impact customer satisfaction. In addition, due to the ordinal nature of empirical data and its violation of normal assumption, ordinal logistic regression model is applied in data analysis. The empirical result shows that quality satisfaction and perceived value affect customer overall satisfaction. When the normal assumption is not met, parameter estimates of linear regression model are biased.