A meta-analytic investigation of the role of valence in online reviews
Purnawirawan et al. (2015)
Online review valance: whether reviews in a review set are predominantly positve or negatve.
Consumers usually come across both positve and negatve reviews related to a product. The fndings
regarding the role of review valence on consumers’ evaluatons are not clear.
Literature review and hypothesis development
We consider two types of review sets:
A single review set;
A set of several reviews (the efect of valance can stem from a comparison between reviews).
We consider two types of positvity degree:
Positvity degree in positve reviews (Pppos);
Positvity degree in negatve reviews (Ppneg).
For both perceived usefulness and attuded the efect of valence should increase with increasing
positvity in the negatve reviews and decrease or remain stable with increasing positvity in the
positve reviews. A small amount of negatve informaton can increase persuasion of the (positve)
review because it indicates that the reviewer is telling the truthd which increases its efectveness. Sod
balanced reviews (that discuss both sides of the product) may be more efectve than extreme
reviews.
Hypothesis 1: usefulness and attude increase with increasing positvity in a negatve review set.
Compared with positve reviewsd negatve reviews gab more atenton and are more easily accessible
in memory. Negatve reviews allow customers to categorize the product as low quality because low-
quality product usually have negatve atributes. Positve reviews do not allow customers to
categorize the product as high quality because any product can have some positve atributes. Sod
increasing positvity in negatve reviews might lead to higher usabilityd but positve reviews are less
useful.
People are more likely to consider informaton that is consistent with their beliefs when forming an
evaluaton. Therefored it is likely that people give more weight to positve reviews (more consistent
with their decision). Positvity increases the efect of negatve reviews untl a high degree of positvity
is reached (ceiling efect). For attudesd positve reviews are more efectve.
Product types:
Search goods: products consumers can evaluate prior to purchase because they have
functonal atributes (electronics);
Experience goods: products with intangible atributes that consumers don’t know untl
purchase/use (hotels).
Because of the intangible character of experience goods and consumers lack product experience
before purchased perceived risk and uncertainty increase. Therefored they might seek additonal
informaton to reduce the risk and uncertainty.
Hypothesis 2: the efect of valence on its consequences is larger for experience than for search goods
Familiar brands have strong memory networks. The informatve efect of online reviews is stronger
for unknown than for well-known brands. Besidesd when customers are familiar with a brandd they
can make their own judgements without relying on online reviews.
Purnawirawan et al. (2015)
Online review valance: whether reviews in a review set are predominantly positve or negatve.
Consumers usually come across both positve and negatve reviews related to a product. The fndings
regarding the role of review valence on consumers’ evaluatons are not clear.
Literature review and hypothesis development
We consider two types of review sets:
A single review set;
A set of several reviews (the efect of valance can stem from a comparison between reviews).
We consider two types of positvity degree:
Positvity degree in positve reviews (Pppos);
Positvity degree in negatve reviews (Ppneg).
For both perceived usefulness and attuded the efect of valence should increase with increasing
positvity in the negatve reviews and decrease or remain stable with increasing positvity in the
positve reviews. A small amount of negatve informaton can increase persuasion of the (positve)
review because it indicates that the reviewer is telling the truthd which increases its efectveness. Sod
balanced reviews (that discuss both sides of the product) may be more efectve than extreme
reviews.
Hypothesis 1: usefulness and attude increase with increasing positvity in a negatve review set.
Compared with positve reviewsd negatve reviews gab more atenton and are more easily accessible
in memory. Negatve reviews allow customers to categorize the product as low quality because low-
quality product usually have negatve atributes. Positve reviews do not allow customers to
categorize the product as high quality because any product can have some positve atributes. Sod
increasing positvity in negatve reviews might lead to higher usabilityd but positve reviews are less
useful.
People are more likely to consider informaton that is consistent with their beliefs when forming an
evaluaton. Therefored it is likely that people give more weight to positve reviews (more consistent
with their decision). Positvity increases the efect of negatve reviews untl a high degree of positvity
is reached (ceiling efect). For attudesd positve reviews are more efectve.
Product types:
Search goods: products consumers can evaluate prior to purchase because they have
functonal atributes (electronics);
Experience goods: products with intangible atributes that consumers don’t know untl
purchase/use (hotels).
Because of the intangible character of experience goods and consumers lack product experience
before purchased perceived risk and uncertainty increase. Therefored they might seek additonal
informaton to reduce the risk and uncertainty.
Hypothesis 2: the efect of valence on its consequences is larger for experience than for search goods
Familiar brands have strong memory networks. The informatve efect of online reviews is stronger
for unknown than for well-known brands. Besidesd when customers are familiar with a brandd they
can make their own judgements without relying on online reviews.
