| Summary: | 碩士 === 國立臺灣大學 === 經濟學研究所 === 105 === This paper discussed the effect of movies to their originals. Novel-based screenplays have been important to the movie industry in recent years. Some of the novel were bestselling, some of them became bestselling because of the movies, and the rest of them remained unpopular. My paper used the data of movies and novel in Taiwan’s market from 2013/8 to 2016/7 to answer whether movies could give positive impact on their originals. Intuitively, novel with good quality would become bestselling and thus reflected in the content of the corresponding movies, which would bring the sales of the novel better. Also, movies could be regarded as advertisements of their originals’ quality check; in addition, the exposure movies brought for their originals would make the sales of novel greater. In this paper, I tried to prove that this phenomenon was true.
By tw.dorama.info, I scrapped the movies’ information from 2013/8 to 2016/7 of Taiwan’s market. Besides, through IMDB and books.com, I checked if movies did have originals and made sure that novel had Chinese versions in Taiwan’s market. Last, I obtained 112 movies corresponding to 201 novel as my sample. Since I couldn’t get the exact sales number, I use the rankings every week on books.com as substitutes of real sales number. The shape of the ranking pattern was like a hill, starting from low rankings, climbing up to high rankings, staying for a while, descending to low rankings, and then dropping out of the sales charts. Here I refer high ranking to small ranking number and vice versa. In addition, the sales in the range of 100 to 50 were quite similar and the rankings around 100 even shared little difference in sales of those not on the sales charts. I set the bottom line at rank 50 due to this observation; however, it didn’t really make great differences when the bottom line was set at a different rankings, which could be seen as robustness. As I mentioned before, I couldn’t get the actual number of sales so I came up with a method of using the total number of weeks on sales charts to be a proxy as the sales. For instance, a novel with rankings of 54, 36, 12, 43, and 86 will got 3 weeks on sales charts, which was based on total number of weeks on sales charts with bottom line at rank 50. Furthermore, I regarded a novel as good quality by the condition of whether it was on sales charts for at least two weeks with bottom line at rank 50 within the period of advertisement, which was 4 weeks before the opening week of corresponding movies. I found this standard reasonable but the standard could be changed and the regression results showed that changing standard wouldn’t make a big impact to the conclusion. Based on the knowledge of total number of weeks on sales charts, what I concern the most was the effect of movies, represented by box office, to their originals, counted in total number of weeks on sales charts, and the interaction between quality of novel and movies’ box office. The regression result showed that, compared to the average total number of weeks on sales charts, 2.2 weeks, with the bottom line at rank 50, 1% increase in box office would lead to 0.69 weeks more on sales charts for novel, which was quite significant. As to interaction, the result indicated that good quality novel could bring in movies with positive impact on the sales of themselves. Though there were flaws in my estimations, the result suited the intuition quite nice. In conclusion, novel with good quality was important to the sales of themselves but the appearance of movies could bring in positive impact to sales of novel whether the quality was good or bad.
|
|---|
