| Summary: | 碩士 === 世新大學 === 資訊管理學研究所(含碩專班) === 106 === This research analyzed the database of Japanese Trendy dram (2009 ~ 2016) to explore the factors that affect TV ratings.
There are two key contributions of this study. Firstly, constructing the measurement of latent variables such as the popularity of producers, directors…and so on. Secondly, exploring an incomplete mediation model of “director popularity - leading actor popularity - TV ratings”.
The results found the relationship between TV ratings and five factors that are weekdays, time frames, producer popularity, director popularity and leading actor popularity. After further examinations, this research filtered above variables into a concise multivariate model and initiated a formula to predict TV ratings.
Using this formula to predict the real TV ratings of 2016, its Criterion-related Validity reached 0.413.
|
|---|
