初步研究:负面网络口碑如何影响消费者的品牌态度的变化

摘要：本文设计了两个实验，从易获得性与诊断力理论两方面探索了当产品涉入度和品牌熟悉度不同时，负面网络口碑对消费者品牌态度变化影响的差异。得到的结论是：当消费者对高涉入产品或对熟悉的品牌有良好的初始印象时，他们将能够对信息诊断力的高负面口碑做出理性的判断，并有强大的信息筛选能力。相比之下，当消费者对低涉入产品或对先前印象不良好的不熟悉的品牌进行判断时，信息诊断力高的负面口碑更容易影响消费者品牌态度的变化。

关键词：易获得性-诊断力；品牌态度；负面网络口碑

一、引言

互联网已经改变了人们的购物方式，消费者喜欢以网络口碑或图片的形式在网络上交流和分享他们购买和使用产品的感受。与传统口碑相比，学者们发现，匿名性、受众的广泛性、容易复制和网络口碑扩散的便利性对消费者的品牌态度有着显著的影响(Gelb amp; Sundaram, 2002; Dellarocas, 2003; Godes amp; Mayzlin, 2004)。研究结果还表明，网络口碑的心理效应会影响人们的判断(Magnus amp; Soderlund, 1998; Herr et al., 1991)。Kardes amp; Kim (1991) 则在易获得性-诊断力理论视角的口碑研究中发现，极端负面口碑作为一种诊断力较强的信息，将削弱其传播时在消费者产品判断中的作用。本文试图探讨负面网络口碑对消费者品牌态度改变的影响。

二、文献和假设

大量研究表明，消费者接受负面口碑带来的品牌态度变化远大于接受正面口碑带来的态度变化(Russell et al., 2001)，负面口碑是一种比正面口碑更有价值的信息(Mahajan, 1984)。费尔德曼和林奇（1988）的易获得性与诊断力理论模型指出，易获得性是指先前的知识在人的记忆中被迅速提取的容易程度(如有趣的、栩栩如生的、引人入胜的信息)，这类信息被认为较易获取，也容易影响人们对事物的判断。同时，理论界对此问题也有相反的研究结果。有学者认为易获得性强的信息对人的记忆存储有显著影响，而在其态度或行为改变上效应并不明显。诊断力是指人们提取的知识是否足以帮助解决问题，若一则信息较易帮助人们将产品进行简单唯一分类，或对信息相关事物的理解是排他和独一无二的，这两类信息为高诊断力信息。一般来说，极端负面口碑比一般负面、中立和正面口碑诊断力高，描述人们头脑中有先前印象的产品相关信息的诊断力也较高。Kardes amp; Kim (1991) 认为信息易获得性在消费者产品判断中具有中介作用。目前，有相当多的实证研究文献使用费尔德曼与林奇（1988）的理论框架进行研究(Patti et al, 2000; Peng Zou 2011)。品牌熟悉度相关研究显示，高熟悉度利于引起人们对某一服务或产品的积极评价，面对熟悉的品牌顾客容易感到温暖亲密。Park 和Stoel(2005) 验证了品牌熟悉度、网站信息及先前网购经验对消费者网购感知风险和购物意向的影响，发现人们对事物的态度易受过往经验影响。当面对全新事物时，人们没有足够经验去判断，第三方信息就较易左右人们； 相反地，当人们判断熟悉事物时，过往经验有助于信息甄别，以得到最有效的判断结果，这时第三方信息的影响力会减弱。从以上观点来看，品牌熟悉度影响消费者对负面口碑的反应，而负面口碑诊断力对消费者品牌评价也有影响。据此，我们设计了2个实验来验证当消费者对品牌的预先态度和对品牌的熟悉度在不同水平时，负面网络口碑对他们品牌态度变化的影响。得到如下假设:

假设1a：面对新的品牌，消费者通过接受模糊负面口碑带来的品牌态度变化比接受明确负面信息时更显著。

假设1b：面对熟悉的品牌，消费者通过接受模糊负面口碑带来的品牌态度变化比接受明确负面信息时更显著。

假设1c：面对熟悉的品牌和有好的初始印象的品牌，消费者通过接受模糊负面口碑带来的品牌态度变化比接受明确负面信息时更显著。面对品牌初始印象不好的品牌也是同样的情况。

假设2a：面对新的品牌和高涉入产品，消费者通过接受模糊负面口碑带来的品牌态度变化比接受明确负面信息时更显著。面对新的品牌和低涉入产品也是同样的情况。

假设2b：面对熟悉的品牌和高涉入产品，消费者通过接受模糊负面口碑带来的品牌态度变化比接受明确负面信息时更显著。面对熟悉的品牌和了低涉入的产品也是同样的情况。

三、方法

研究对象被分为负面清晰口碑和负面模糊口碑(高诊断力和低诊断力) 两组，通过引导完成了问卷调查材料。首先，测量明确负面信息组的成员对某品牌（问卷调查第一页）的预先态度，然后阅读从在线购物网站中提取的关于该品牌的一些极端负面口碑信息，最后，尝试重新评估他们的品牌态度(Zemborain, 2007)。模糊负面信息组成员阅读模糊的负面口碑，实验程序同上。

3.1 实验1

邀请中部地区大学的62名商学院学生（包括本科生和研究生）参加一个实验（实验对象是运动鞋），32名男生30名女生，平均年龄22岁。实验将持续30分钟，采用一个2times;3的实验设计为主体，研究对象将被随机分为2组（明确负面信息组vs模糊负面信息组），分别对三组品牌的运动鞋（瑞蛙，耐克，安踏）进行评价，实验程序同上。

操作测试：如表1所示，研究对象对瑞蛙的品牌熟悉度低（Mk = 1.04，SDk = 0.51），但对耐克的熟悉度高（Mn = 5.73，SDn = 0.48），对安踏的熟悉度也比较高（Ma = 5.30，SDA的= 0.44），t检验结果是tn-k = 52.6（P＜0.001）和49.7（P＜TA-k = 0.001），耐克预先品牌态度均值为Mnike = 5. 69(0. 56)，安踏预先品牌态度均值为Manta = 3. 09(0. 65)，tn-a = 2. 35(p ＜ 0. 01)，表明实验对品牌熟悉度和预先品牌态度操控有效。

表1 品牌熟悉度测试

表2 实验1的结果

（注：括号外数值表示品牌态度变化均值，括号内数值表示品牌态度变化标准差；组间均值出现两个“a”表示差异不显著，出现“a”和“b”表示差异显著，文中表格表述均同。）

实验结果：表2的结果表明，瑞蛙组(新品牌)被试接受负面清晰口碑后态度变化均值-0.36 大于接受负面模糊口碑后态度变化均值-1.53，但差异并不显著(tchange1＜1. 96)，假设1a没有得到验证。但是对于熟悉的品牌，模糊负面信息组的态度变化比明确负面信息组更显著，主要影响显著（P＜0.001），假设1b得到验证。

耐克的品牌态度在接收模糊负面信息后的态度变化均值-2.28大于接受明确的负面信息后态度变化均值 -0.87，这种差异是显著的tchange2 = 9.23（P＜0.001），安踏组接受负面清晰口碑后态度变化均值-1.06大于接受负面模糊口碑后态度变化均值- 0.48（P＞0.05），但差异并不显著( tchange3 ＜1. 96)，从而假设1c得到部分验证。

3.2实验2

65 名高校学生(37 男、28 女，平均21 岁) 参加一个以相机和运动鞋为标的物的实验。实验采取2 times; 4 设计，两组被试(负面模糊组和负面清晰组) 分别对先锋、瑞蛙、尼康和耐克等四个品牌进行评价。

操作测试： 如表3所示，被试对瑞蛙品牌熟悉度均值低（Mrui = 1.29，SDrui = 0.97）的对先锋相机品牌熟悉度也低（（MX = 0.56，SDX= 0.47）。但对耐克品牌熟悉度均值高（Mn2= 5.72，SDn2= 0.51）的对尼康相机品牌熟悉度均值也比较高（Mnikang = 5.33，SDnikang = 0.57）。耐克品牌熟悉度均值高于瑞蛙品牌，t的检验结果（tn-r = 31.08）非常显著。同时，对尼康品牌的品牌熟悉度均值高于先锋品牌，且差异（tn-x = 51.17）是出人意料的显著。结果表明，在本实验中对品牌熟悉度的操作是有效的。

表3 品牌熟悉度测试

表4显示，相机产品了解程度高于运动鞋，新品牌涉入度的平均值是（4.34＞2.69，t= 4.43，p＜0.001），熟悉品牌涉入度是（5.96＞3.38，T = 9.55，P＜0.001）。这表明，在实验中对产品涉入度操作是有效的。

表4 产品了解度测试

表5 实验2结果

表5显示，先锋组被试接受负面模糊口碑后态度变化均值 -1.84大于接受负面明确口碑后态度变化均值-0.73，这种差异很显著(tchange21 = 7. 24，p ＜ 0. 01)。瑞蛙组被试接受负面模糊口碑后态度变化均值-0.36小于接受负面明确口碑后态度变化均值-1.56，但差异并不显著（tchange22 = 10.3 p＞0.05）。实验数据证明当面对新的品牌、高涉入的产品时，通过接受模糊负面信息的消费者的品牌态度变化比接受明确负面信息的更显著。数据还表明，当面对新的品牌、低涉入的产品时，明确负面口碑的影响更大，假设2a得到部分验证。

尼康组被试受负面模糊口碑影响态度变化均值-1.60大于受负面明确口碑影响态度变化均值-0.41，且这种差异有显著性tchange23 = 10.92（P＜0.001）。实验数据证明，，表明熟悉品牌

剩余内容已隐藏，支付完成后下载完整资料

An Initial Study: How Negative Online WOM Influence Consumerrsquo;S Brand Attitude Change

Fang Ji, Ting-jui Chou

School of Business, RenMin University of China, 100872, Beijing, CHINA

Abstract: Two experiments are designed to examine the effect of negative online WOM（Word Of Mouth) .The conclusion is: when consumers have high product involvement or good pre-attitude on familiar brand, they would make rational judgment on the high diagnostic negative WOM, and have strong Screening capacity on information; In contrast, when judging low-involvement product and unfamiliar brand with negative prior attitude, the high diagnostic negative WOM is easier to influence consumers brand attitude change.

Keywords: Accessibility-Diagnosticity; Brand Attitude; Negative WOM

1. Introduction

Internet has changed the way people shopping, consumers like to communicate and share their feelings of purchasing and products using in form of WOM or images on the web. Compared to traditional WOM, Scholars found that the anonymity, the audience extensiveness, easy to duplicate and diffuse convenience of online WOM, which have significant impact on consumer brand attitude (Gelb amp; Sundaram, 2002; Dellarocas, 2003; Godes amp; Mayzlin, 2004). It was shown that valence of WOM would affect peoples judgment. (Magnus amp; Soderlund, 1998; Herr et al., 1991). Kardes amp; Kim (1991) revealed that extreme negative WOM would weaken the influence of WOM on Consumer Product judgment. This article attempts to explore the effects of negative online WOM on consumers brand attitude change.

2. Literature and Hypothesis

A lot of studies suggest that brand attitude change of consumer exposure to negative information is larger than positive information (Russell et al., 2001). Negative WOM is kind of more valuable information than positive WOM (Mahajan, 1984). Feldman amp; Lynch (1988) proposed that negative WOM which have symbol of extreme are more diagnostic than the generally negative, neutral, and positive word of mouth (L.AAKER, 2000). Kardes amp; Kim (1991) argued that the accessibility of information has a mediate effect on product judgment. At present, there are quite a lot of empirical research literature using Feldman, amp; Lynch (1988) theoretical framework making research (Patti et al, 2000; Peng Zou 2011).We designed 2 experiments to verify when pre-attitude /brand familiarity in different level, effects of positive online WOM on consumer brand attitude and product judgment.

H1a: Faced with new brand, consumerrsquo; brand attitude change by accepting negative vague information is more significant than accepting negative clear information.

H1b: Faced with familiar brand, consumerrsquo; brand attitude change by accepting negative vague information is more significant than accepting negative clear information.

H1c: Familiar with the brand and has a good pre-brand attitude, consumerrsquo;s brand attitude change by accepting negative vague information is more significant than accepting negative clear information. The same situation when pre-brand attitude is bad.

H2a: Faced with new brand and high involvement product, consumerrsquo;s attitude change by accepting negative vague information is more significant than accepting negative clear information. The same situation faced with new brand and low involvement product.

H2b: Faced with familiar brand and high involvement product, consumerrsquo;sattitude change by accepting negative vague information is more significant than accepting negative clear information. The same situation when faced with familiar brand and low involvement product.

2. Method

Subjects were divided into 2 groups (high diagnosticity VS low diagnosticity), completed a material guided questionnaire, firstly, negative clear group was in the measurement of the pre-after overall attitude of a certain brand (the first page of the questionnaire), and then the subjects read some extreme negative WOM of this brand extracted from online shopping site. At last, try to re-evaluation the brand attitude again (Zemborain, 2007). The negative vague group read the vague negative WOM, the experimental procedure as above.

2.1 Experiment 1

62 students from a university(School of Business) in central region(including undergraduate and graduate students) are invited to participate in an experiment（the

experimental objects are sports shoes）, 32 boys and 30 girls , mean age is 22 years old. The experiment lasted 30 minutes. A 2 * 3 design experiments was taken, subjects

were randomly divided into 2 groups (negative clear group vs. negative vague group) (Between), respectively evaluate 3 sets of brand (Korten/Nike/ANTA) (Within), The experimental procedure is same as above.

Manipulation test. As shown in Table 1, brand familiarity of subjects on korten brand is low (Mk = 1.04, SD k = 0.51), but there is a high degree of familiarity with

the Nike brand (Mn= 5.73 , SDn = 0.48), the ANTA brand familiarity is also relatively high (MA = 5.30，SDA = 0.44),t test results is tn-k=52.6(plt;0.001) and tA-k = 49.7 (p lt;0.001) ,which shows that the manipulation on brand familiarity in this experiments is effective. The mean of pre-attitude about Nike brands group is Mnike=5.69(SD=0.56), and the pre-attitude means of ANTA brands group is Manta=3.09(SD=0.65)，t=23.49(plt;0.001), indicating that this experiment is in advance of brand pre-attitude manipulation.

Experiment Results

Results in Table 2 show that mean brand attitude change of kroten brand bring out by negative vague WOM is -0.36.After receiving a clear negative WOM, mean of attitude change is -1.53. (t=-10.05,pgt;0.05), H1a has not been verified. But when Familiar with the brand,the attitude change of negative vague group is more significant than negative clear group, the main affect th

剩余内容已隐藏，支付完成后下载完整资料

资料编号：[286973]，资料为PDF文档或Word文档，PDF文档可免费转换为Word