網路評論串正負性對讀者感知之影響

Abstract

網路口碑一直是消費者在做出購買決策前的重要參考依據，因此過去文獻對於網路口碑的屬性已有部分探討，文獻證實在閱讀單則訊息下知覺助益性中出現的正面效應(positivity effect)，意即正面訊息會比負面訊息更感知有幫助，以及負面效應(negativity effect)，意即負面訊息會比正面訊息更感知有幫助，這兩種現象可用確認偏誤理論同時加以解釋，在高品牌聲望的先前態度下會出現正面效應，反之在低品牌聲望下則會出現負面效應。隨著網路口碑於各式電商平台越來越普及，潛在消費者在瀏覽商品資訊時，往往閱讀到的不只單則評論，而是由幾則評論所構成的評論串。因此，本研究欲探討，在閱讀評論串的過程中，是否同樣能用閱讀者對品牌的先前態度來解釋可能出現的正面/負面效應，因此我們以品牌先前態度為調節變項，來了解先前態度是否會調節評論串正負面和知覺助益性、感知資訊完整性及感知真實性的關係。另外，在閱讀之評論數多則的情形下，本研究同時欲探討，評論串中正負評論數量分布有變化時，是否能用以解釋知覺助益性、感知資訊完整性及感知真實性中可能出現的正面/負面效應，因此我們也以正負評論數量分布為調節變項，來了解其是否會調節評論串正負面和知覺助益性、感知資訊完整性及感知真實性的關係。經過實證研究後，結果支持本研究多數對確認偏誤為調節變項的假說，發現閱讀者在對品牌先前態度較佳下，會出現正面效應，認為正面評論串比負面評論串感知更為有幫助、完整、真實，反之亦然。另一方面，實證資料對於評論正負數量分布變化的調節效果則部分支持，在高度一致性之評論串下，感知真實性存在負面效應，閱讀者認為負面評論串(六負零正之評論串)會比正面評論串(六正零負之評論串)感知更真實，至於在低度一致性之評論串中知覺助益性及感知真實性則存在正面效應，閱讀者認為正面評論串(四正二負之評論串)會比負面評論串(四負二正之評論串)感知更有幫助、更真實。

Online Electronic word-of-mouth (eWOM) is one of the most important factors to take into consideration for customers before buying all along. Past studies have found that when reading a single review, the positivity effect (positive information is perceived to be more helpful than negative information) and the negativity effect (negative information is perceived to be more helpful than positive information) can both be explained by the theory of confirmation bias. There’s positivity effect in perceived helpfulness under favorable prior attitude toward brands and on the contrary, there’s negativity effect under unfavorable attitude. As eWOM becomes more and more widespread on the E-commerce platform, potential customers may read not only a single but a set of reviews when browsing products online. Therefore, we would like to know when reading a set of reviews, is it still appropriate to explain the possible positivity and negativity effect by prior attitudes of readers. Thus, we used readers’ confirmation bias as a moderator to understand is there moderating effect in the relationship between review balance and perceived helpfulness, perceived information completeness and perceived truthfulness. Besides, following the trend of reading multiple reviews at a time, we also want to know that is the positivity and negativity effect can also be explained by the extremity of the review set (amount of positive and negative reviews distribution). Hence, we used the review set extremity as the moderator to see if there’s moderating effect in the relationship between review balance and perceived helpfulness, perceived information completeness and perceived truthfulness. Our empirical results support most hypotheses which infer that confirmation bias is a moderator. We found that under favorable prior attitude, there’s positivity effect so that readers would perceive positive balanced review set as more helpful, complete and truthful and vice versa. On the other hand, the empirical results about the moderating effect of review extremity partially support our hypotheses. When reading high extremity of review set, there exists negativity effect so that readers perceive the negative balanced review set (six negative reviews without any positive reviews) to be more truthful than positive balanced review set (six positive reviews without any negative reviews). Whereas in low extremity of review set, there exists positivity effect in perceived helpfulness and perceived truthfulness so that readers perceive positive balance review set (four positive reviews plus two negative reviews) to be more helpful and truthful than negative balanced review set (four negative reviews plus two positive reviews).