外文翻譯--在線口碑和產(chǎn)品銷售動態(tài)——對電影行業(yè)的實證調(diào)查

1、<p><b>　　中文3310字</b></p><p><b>　　畢業(yè)論文外文翻譯</b></p><p><b>　　一、外文原文</b></p><p>　　標題：The dynamics of online word-of-mouth and product sales—An e

2、mpirical investigation of the movie industry</p><p><b>　　原文：</b></p><p>　　Introduction</p><p>　　Word-of-mouth (WOM) has been recognized as one of the most influential re

3、sources of information transmission since the beginning of human society (Godes and Mayzlin 2004; Maxham and Netemeyer 2002; Reynolds and Beatty 1999). However, conventional interpersonal WOM communication is only effect

4、ive within limited social contact boundaries, and the influence diminishes quickly over time and distance (Bhatnagar and Ghose 2004; Ellison and Fudenberg 1995). The advances of information technology and th</p>&

5、lt;p>　　Online WOM presents both challenges and opportunities to retailers. On the one hand ,WOM provides an alternative source of information to consumers, thus reducing retailers’ ability to influence these consumers

6、 through traditional marketing and advertising channels. Prior studies show that a variety of aspects of WOM influence retail sales. Some found that WOM dispersion (Godes and Mayzlin 2004) and valence (Chevalier and Mayz

7、lin 2006; Forman, Ghose, andWiesenfeld 2008) have significant effects </p><p>　　We have chosen the movie industry as our research context because industry experts agree that WOM is a critical factor underlyi

8、ng a movie’s staying power, which leads to its ultimate financial success (Elberse and Eliashberg 2003). In addition, the movie industry has by far received the most attention in marketing literature on WOM, which allows

9、 in-depth comparison of our results with those of previous studies. We, however, note that movies are a unique type of experience goods and the results f</p><p>　　Online WOM in the movie industry takes many

10、forms, including online reviews, discussion boards, chat rooms, blogs, wikis, and others. In this study, we focus on online user reviews because statistics suggest that user reviews are more prevalent than other forms of

11、 WOM communication in the movie industry. Beyond volume, another subtle but important difference between online user reviews and other types of WOM is that user reviews usually reflect user experience and consumer satisf

12、action, which ar</p><p>　　The rest of the paper is organized as follows. The next section provides the literature review followed by the discussion of our conceptual framework and research hypo theses. We th

13、en describe our sources of data and the empirical model and estimation. Main findings are presented and discussed next, and the paper ends with a discussion of implications, limitations, and future research.</p>&

14、lt;p>　　Empirical model specification</p><p>　　The development of our empirical model is guided by the following considerations. First, as we are interested in the drivers of both box office revenue and WO

15、M, we construct a system of two interdependent equations: one equation with daily revenue as the dependent variable (the revenue equation) and the other</p><p>　　with WOM volume as the dependent variable (th

16、e WOM equation). We assume that in each time period (i.e., day), the errors in the two equations may be correlated, which implies that factors not included in our model could simultaneously influence both movie revenue a

17、nd WOM.</p><p>　　Second, recognizing that interactions between consumers’ movie-going behavior and WOM can go beyond the concurrent term (Elberse and Eliashberg 2003), we develop a system of dynamic equation

18、s. That is, in the revenue equation, we include</p><p>　　not only the contemporaneous term of daily WOM volume, but also multi-lag terms. Likewise, in the WOM equation, multi-lag revenue terms are also incor

19、porated. Such a specification also helps identify both equations for the simultaneous equation system since the lagged terms are exogenous variables in either equation. In addition, following extant research, we use alog

20、-linear formulation (e.g., Elberse and Eliashberg 2003; Liu2006) in our model. The log-linear formulation is consistent with</p><p>　　theoretical models of a multistage consumer decision-making process, wher

21、e sales of a movie can be viewed as a series of conditional probabilities applied to the consumer base. A log transformation converts the relationship into a linear form for empirical estimation. Moreover, log transforma

22、tion smoothes the distribution of variables in the linear regression, and the estimated coefficients of the log-linear form directly reflect the elasticity of independent and dependent variables. Third, to c</p>&

23、lt;p>　　fixed effects estimation also allows the error term to be arbitrarily correlated with other explanatory variables, thus making the estimation results more robust.</p><p>　　Implications, limitations

24、, and future research </p><p>　　Our model specifies the dual causal relationship and reveals the positive feedback mechanism between online WOM and product sales. Our findings strongly support the value of c

25、onsidering the endogeneity of WOM and its interdependence with consumers’ consumption behavior. The notably different results obtained from 3SLS (the statistically more robust method) and OLS suggest that extant research

26、 using simple regression techniques may have drawn biased conclusions about the direction and magnitude o</p><p>　　Our findings also bring important extensions to previous research (Basuroy et al. 2003; Elia

27、shberg and Shugan 1997; Liu 2006) on the relationship among WOM volume ,WOM valence and box office sales. Previous research has been focusing on</p><p>　　the direct impact of WOM volume and valence on box of

28、fice revenues and find that most of the explanatory power comes from WOM volume not WOM valence (Liu 2006). Our study extends this approach by considering the interaction between WOM valence and WOM volume. We find that

29、while WOM valence does not directly affect revenue, higher WOM valence indirectly increases box office revenue by generating higher volume of WOM.</p><p>　　The contributions of this research to retail litera

30、ture on WOM are multifaceted. From the methodology perspective, we bring to light the importance of separating the effect of WOM as both a precursor and an outcome of sales. Our results also highlight</p><p>

31、;　　the importance of using a dynamic system and high-frequency data in studying the effect of WOM in the digital environment. From the managerial perspective, we show that WOM valence and WOM volume play different roles

32、in influencing product sales. We also show that time-series changes in WOM valence influences WOM volume which leads to higher product sales. Our findings support the idea that the online WOM process has a significant im

33、pact on sales, suggesting that businesses should</p><p>　　embrace and facilitate WOM activities.</p><p>　　There are a number of opportunities to extend the current research. One important and in

34、teresting extension of our research will be to investigate the consumer decision process under the influence of WOM information, especially in the digital</p><p>　　environment. In addition, not all WOM is eq

35、ual. Consumers need to distinguish the “true” and “honest” opinions from all kinds of feedback and recommendations on the web. Under such circumstances, how consumers choose their information source find trusted informat

36、ion sources will be of particular interest for future research.</p><p>　　Online user reviews are only one type of consumer-generated media. The recent explosive growth of popular online social communities (e

37、.g., www.YouTube.com, www.Flickr.com, and www.Digg.com) has generated a renewed interest in the Internet as a new medium for content generation and distribution. Different from online review sites we explored here, onlin

38、e social communities encourage interaction between users, which potentially changes the dynamics of WOM distribution. The modeling approach used i</p><p>　　Our analysis is, by necessity, restricted to online

39、 users who choose to post reviews and to post them on Yahoo! Movie. Thus, our estimates are conditioned on such a user population. While such a restriction does not necessarily bias the panel data estimation results, the

40、y should be interpreted as applying to a self-selected set of online users. In addition, this paper studies only the relationship between the postre lease WOM and sales. However, WOM certainly has existed before a movie’

41、s release</p><p>　　出處：Wenjing Duan,Bin Gu, Andrew B.The dynamics of online word-of-mouth and product sales—An empirical investigation of the movie industry[J]. Journal of Retailing ,2008(6): 233-242</p>

42、;<p><b>　　二、翻譯文章</b></p><p>　　標題： 在線口碑和產(chǎn)品銷售動態(tài)——對電影行業(yè)的實證調(diào)查</p><p><b>　　譯文：</b></p><p><b>　　介紹</b></p><p>　　口碑自人類社會開始以來，已經(jīng)被公認為

43、最有影響力的信息傳遞的資源(Godes and Mayzlin 2004; Maxham and Netemeyer 2002; Reynolds and Beatty 1999).然而，傳統(tǒng)的人際間口碑交流只有在有限的社會接觸范圍內(nèi)才會起作用，其影響隨時間和空間的推移迅速減弱(Bhatnagar and Ghose 2004; Ellison and Fudenberg 1995)。信息技術(shù)的進步和在線社交網(wǎng)站的出現(xiàn)，深刻地改變了信息

44、傳播的方式，并已超過了傳統(tǒng)口碑的界限(Laroche et al. 2005)。另外，在一個或幾個朋友間傳播的短暫口碑，已經(jīng)被看作是能透過此看見世界的信息。因此，在線口碑在消費者購買決策中發(fā)揮著日益重要的作用。</p><p>　　在線口碑對零售商來說，既是機遇也是挑戰(zhàn)。一方面，口碑為消費者提供了可供選擇的信息源，從而降低了零售商通過傳統(tǒng)市場營銷和廣告渠道影響消費者的能力。先前的研究表明，口碑的各個方面會影響到零

45、售的銷量。一些研究發(fā)現(xiàn)，口碑量(Godes and Mayzlin 2004)和價數(shù)(Chevalier and Mayzlin 2006; Forman, Ghose, andWiesenfeld 2008)對產(chǎn)品銷量產(chǎn)生重大影響，而另一些研究發(fā)現(xiàn)，口碑量被看作是產(chǎn)品銷量的主要驅(qū)動力(Chen, Wu, and Yoon 2004; Liu 2006).。另一方面，在線口碑為零售商提供一個接觸到消費者和戰(zhàn)略影響消費者意見的場所。近幾年

46、出現(xiàn)的傳聞證據(jù)表明，在線口碑已作為新的營銷工具被成功應(yīng)用(Dellarocas2003).。不同于傳統(tǒng)的市場營銷效果，口碑效應(yīng)的獨特之處在于口碑和產(chǎn)品銷量的積極反饋進程，也就是說，口碑會提高產(chǎn)品銷量，反過來產(chǎn)生更多的口碑和更多的產(chǎn)品銷量。積極的反饋機制表明，口碑不僅是消費者購買的驅(qū)動力，而且是零售銷量的結(jié)果( Godes and Mayzlin 2004; Srinivas</p><p>　　我們選擇了電影產(chǎn)業(yè)

47、作為研究范疇。因為業(yè)界專家一致認為，口碑是能令電影持續(xù)發(fā)揮其功效的一個關(guān)鍵因素，從而實現(xiàn)最終財務(wù)上的成功(Elberse and Eliashberg 2003)。此外，電影界迄今為止在文學營銷口碑上受到重視，這使得我們的成果能和之前的研究進行深入的比較。然而，我們注意到電影作為一種特殊的體驗產(chǎn)品類型，其產(chǎn)業(yè)的結(jié)果并不一定能夠推廣到其他零售行業(yè)。相反，我們的目標是利用電影產(chǎn)業(yè)作為一個內(nèi)容來強調(diào)考慮零售銷量和在線口碑的動態(tài)和他們之間關(guān)系的

48、重要性，并說明設(shè)置的聯(lián)立方程方法的正確性。我們發(fā)現(xiàn)，一部電影的票房收入和口碑效價明顯地影響著口碑量，口碑量反過來會帶來更高的票房和更好的成績。我們的研究結(jié)果澄清了早期研究中用戶評分對票房收入影響的博弈。我們表明，用戶評分不會直接影響票房收入。但是，他們會通過口碑間接影響到票房收入。</p><p>　　此外，我們的研究還證實了在線口碑不僅是先導，而且是產(chǎn)品銷量的結(jié)果。我們表明， 忽視口碑的雙重性質(zhì)會導致錯誤的結(jié)果

49、。</p><p>　　電影產(chǎn)業(yè)的在線口碑有多種表現(xiàn)形式，包括在線評論，討論版，聊天室，博客，維基和其他。在此研究中，我們側(cè)重于在線用戶評論，因為數(shù)據(jù)顯示，在電影產(chǎn)業(yè)中，在線用戶評論比其他口碑交流更加普遍。除了量之外，在線用戶評論和其他形式的口碑之間的重要微妙區(qū)別在于，用戶評論通常反映了用戶的經(jīng)歷和滿意度，這被看作是產(chǎn)品的信息源之一(Chen and Xie 2004; Li and Hitt 2008).。與此

50、同時，口碑的其他類型，比如較能反映消費者的預期的在線社區(qū)網(wǎng)站的討論，深受社會結(jié)構(gòu)的影響(Gopal et al. 2006; Liu 2006)。</p><p>　　本文的其余部分組織如下。下一節(jié)將提供支撐我們概念框架和研究模型的文獻。然后，我們描述我們的數(shù)據(jù)來源和經(jīng)驗模型及其預測。之后，將描述和討論主要的調(diào)查結(jié)果，以及討論的意義、限制及未來的研究。</p><p><b>　

51、　實證模型詳述</b></p><p>　　我們的實證模型發(fā)展遵循以下考慮。首先，由于我們對電影票房收入和口碑驅(qū)動力感興趣，我們構(gòu)建了兩個相互依存的系統(tǒng)，一個是每日收入作為因變量方程（收入方程），另一個是口碑作為因變量的方程（口碑方程）。我們假設(shè)在每個階段（例如，一天），兩個方程的錯誤是相關(guān)的，這意味著不包括在我們模型的因素會同時影響著電影的收入和口碑。</p><p>　　其

52、次，認識到消費者對電影的選擇行為和口碑之間相互作用會超越同期(Elberse and Eliashberg 2003)，我們發(fā)展了一個動態(tài)方程系統(tǒng)。在收入方程中，不僅包括同期的每日口碑量，而且還有滯后的因素。同樣，在口碑方程中，滯后的收入因素也同樣存在著。這樣的規(guī)范有助于識別兩個聯(lián)立方程系統(tǒng)，因為滯后因素屬于每個方程的外在變量。此外，繼現(xiàn)存的研究，我們在模型中使用線性方程(e.g., Elberse and Eliashberg 200

53、3; Liu2006)。線性方程符合消費者決策過程的各個階段，一部電影的銷量可以看作是應(yīng)用到消費群的一系列條件概率。一個數(shù)可轉(zhuǎn)換成經(jīng)驗估計的線性關(guān)系模型，此外，線性方程中回歸變量的平滑分布，以及線性形式的估計系數(shù)可直接反應(yīng)非獨立和獨立變量的彈性。第三，為了控制更多可影響影響電影收入和口碑的特有因素，如預算、市場營銷、明星和其他(Basuroy et al. 2003; Elberse and Eliashberg 2003; Liu 2

54、006)，我們把在模型中通過增加特定虛擬變量所帶來的固定效果包括在內(nèi)。固定效果捕獲了任何變化因素，包括內(nèi)在的電影特色，</p><p>　　意義、限制和未來研究</p><p>　　我們的模型詳細說明了雙重因果關(guān)系，并揭示了在線口碑和產(chǎn)品銷量之間的積極反饋機制。我們的研究有力地支持了考慮口碑的內(nèi)在性質(zhì)和消費者消費行為之間的相互依存關(guān)系。從3SLS（更強大的統(tǒng)計方法）和OLS的明顯不同得出，

55、利用簡單回歸技術(shù)的現(xiàn)存研究可能會得出關(guān)于口碑效應(yīng)方向和大小的偏見結(jié)論。我們的結(jié)果驗證了我們關(guān)于在線用戶評論數(shù)量和零售銷量雙贏關(guān)系的斷言。</p><p>　　我們的研究也是對先前關(guān)于口碑量、口碑效價、票房銷量之間關(guān)系研究(Basuroy et al. 2003; Eliashberg and Shugan 1997; Liu 2006)的重要延伸。先前的研究一直注重口碑量的直接影響和票房收入價的效價，并發(fā)現(xiàn)主要的

56、解釋力量來自于口碑量而不是口碑效價(Liu 2006)。我們的研究拓展了考慮口碑量和口碑效價相互作用的方法，我發(fā)現(xiàn)，雖然口碑效價不能直接影響收入，但更高的口碑效價能通過產(chǎn)生更多的口碑量間接提高票房收入。此研究對口碑文獻的貢獻是多方面，從方法論的角度來看，我們揭示了分離先導和銷售結(jié)果的口碑效應(yīng)的重要性。同時，我們的結(jié)果強調(diào)了，在數(shù)字環(huán)境中，使用動態(tài)系統(tǒng)和高頻數(shù)據(jù)對研究口碑效應(yīng)的重要性。從管理角度來看，我們說明了口碑效價和口碑量在影響產(chǎn)品銷

57、量中扮演著不同的角色。我們還說明，口碑效價時間序列的變化影響著能帶來更高銷量的口碑量。我們的研究支持在線口碑進程對銷量有重要影響的想法，這表明企業(yè)應(yīng)支持和豐富口碑活動。</p><p>　　有很多機會去延伸現(xiàn)有的研究，我們研究中的一項重要和有趣的擴展是去調(diào)查消費者在口碑信息，尤其是數(shù)字環(huán)境中的決策過程。此外，并非所有的口碑都是平等的，消費者需要從網(wǎng)上所有反饋和建議中尋找到真實和誠實的意見。在這種情況下，消費者如何

58、從信息源中找到可信賴的信息，是今后研究中特別有趣的。</p><p>　　在線用戶評論只是一種，連接消費者的媒介。作為一項新媒介的生成和分布的在線社區(qū)(e.g., www.YouTube.com, www.Flickr.com, and www.Digg.com)受歡迎程度的急劇增長，在互聯(lián)網(wǎng)上引起了新的關(guān)注。與在線網(wǎng)站評論不同，我們在這里探討，在線社交社區(qū)鼓勵用戶之間的互動，這可能會改變口碑的動態(tài)分布。因此，本

59、次研究中使用的模型方法可能不能充分地反映所研究的內(nèi)容。對新媒介中在線口碑效應(yīng)的描述和確定的新研究，有利于我們理解在線連接消費者的媒介對市場營銷和零售戰(zhàn)略所產(chǎn)生的影響。</p><p>　　我們的分析，根據(jù)需要對選擇發(fā)帖評論和在雅虎評論電影的在線用戶進行限制。因此，我們的研究是對這樣一個有條件的用戶群的估計。雖然這樣的限制不一定有偏于面板數(shù)據(jù)的預測結(jié)果，他們應(yīng)該被解釋為運用到一個在線用戶的自我選擇設(shè)置。此外，本為只

60、是對張貼發(fā)布口碑和銷量之間關(guān)系的研究。不過，口碑在電影發(fā)布前是肯定存在的，電影制片廠作出各種營銷努力以促進口碑(Liu 2006)。關(guān)于雅虎電影的一項詳細調(diào)查顯示了預發(fā)布口碑和發(fā)布口碑之間重要和有趣的區(qū)別。我們注意到，在雅虎電影討論版的預發(fā)布口碑活動中心，帖子主要反映了消費者期望。同時，在雅虎電影用戶評論網(wǎng)站的發(fā)口碑貼中心，帖子主要反映消費者的滿意度和產(chǎn)品體驗。這種差別表明了在預發(fā)布口碑和發(fā)布口碑之間，確實存在著不同的機制。因此，對現(xiàn)有