How can online reviews facilitate decision making: The impact of semantic and linguistic characteristics on the helpfulness of online consumer reviews

(1)

The impact of semantic and linguistic

characteristics on the helpfulness of online

consumer reviews

First supervisor: Dr. J. A. Voerman Second supervisor: MSc. J. A. Koch

University of Groningen

Faculty of Economics and Business MSc Marketing Management

Master Thesis Defense Dániel Hegedüs

(2)

Introduction

• A vast number of online consumer reviews on opinion platforms (e.g. Amazon.com)

• Readers can evaluate product attributes and benefits based on these reviews (Park

et al., 2007)

• Emerging need to determine the most helpful reviews

• Existing helpfulness voting systems do not distinguish between the different

attributes of reviews

(3)

Problem Statement and Research Questions

“How do the objectiveness, concreteness and linguistic style of online consumer

reviews influence the helpfulness of reviews written on search products?”

1. How does the increase in perceived content diagnosticity influence the helpfulness of online reviews?

2. How does the change in perceived argument quality effects review helpfulness?

3. How does the objectiveness of online reviews influence their helpfulness? 4. How does concreteness of online reviews affect their helpfulness?

(4)

Conceptual Model

• H₁-H₃: The effects of measured variables • H₄-H₉: Main effects of independent variables • H₁₀-H₁₅: Interaction effects of independent variables • H₁₆-H₂₁: The effects of control variables

(5)

Theoretical Framework

• Helpfulness

• Expresses the degree to which online reviews can aid in the decision making of readers (Li et al., 2013)

• Diagnosticity

• Shows the extent to which reviews can reduce readers’ uncertainty and equivocality, thus helping them distinguish between different assumptions regarding the focal product (Weathers et al., 2015; Lee & Koo, 2012; Herr et al., 1991)

• Argument Quality

(6)

Theoretical Framework

• Objectiveness (objective vs. subjective reviews)

• Reviews using factual, logical and reason-based information about the described products, in contrast to highly biased and subjective evaluations (Ghose & Ipeirotis, 2011; Park et al., 2007)

• Concreteness (concrete vs. abstract reviews)

• Reviews involving clear and specific depiction of the product experience, instead of vague and inexpressive statements (Li et al., 2013)

• Linguistic Style (reviews bearing high vs. low quality linguistic style)

• The particular set of words chosen to be included in the claims of a review, helping the

(7)

Theoretical Framework

• Control variables

• Frequency of online shopping

• Frequency of reading online consumer reviews

• Consumers who regularly shop online and read online reviews, become more

(8)

Research design

• Survey about an imagined online purchase situation

• 226 respondents

• 2x2x2 mixed experimental design

examining how helpful the respondents find the presented online consumer

reviews (Aronson et al., 1998)

• Full profile design (Malhotra & Birks, 2007)

Linguistic style

High quality Low quality

Objectiveness

Objective Subjective Objective Subjective

Concreteness Concrete Profile 1 Profile 3 Profile 5 Profile 7

Abstract Profile 2 Profile 4 Profile 6 Profile 8

Table 1 – Stimulus profiles

High quality linguistic style, objective and abstract review:

“The price of this camera is quite modest, and includes the full set. Its battery is enough for more than a day and takes really high quality pictures. It is also suitable for low-light settings.”

High quality linguistic style, subjective and concrete review:

“€350 for this camera is not too much. I don’t think you would need more than a 1150mAh battery, and 24MP is just stunning for such an entry device! And with an ISO of 6400, I knew immediately, that I have to take this one.”

(9)

Results

• Main effects are significant in all cases • Interaction effect only between

objectiveness and concreteness, only in the case of helpfulness

Table 2 –Condition means on Argument Quality, Diagnosticity and Helpfulness Figure 3 – Interaction effect between Objectiveness and Concreteness

2 3 4 5

Subjective Objective

Estimated Marginal Means of Helpfulness

(10)

Results – Regression Analysis on Argument Quality

• All main effects of the independent variables are positive and highly significant

• Interaction effects are insignificant • Slight positive effect of frequency

of purchasing

* : at least 90% of significance ** : at least 95% of significance ***: at least 99% of significance

Argument Quality Model 1 Model 2 Model 3 Model 4 Model 5

Constant 1.532 1.589 1.607 1.640 1.650 Objectiveness .400*** .370*** .360*** .360*** .360*** Concreteness .437*** .406*** .406*** .388*** .387*** Linguistic style .311*** .310*** .300*** .283*** .279*** Objectiveness*Concreteness .053 .053 .052 .052 Objectiveness*Linguistic style .017 .016 .016 Concreteness*Linguistic style .031 .031

Frequency of online purchases .048*

Frequency of reading online reviews -.023

R2 _.456 _.457 _.457 _.458 _.460

R2_Adjusted _.454 _.454 _.453 _.453 _.454

F 183.363 137.828 110.134 91.754 69.298

Significance .000 .000 .000 .000 .000

(11)

Results – Regression Analysis on Diagnosticity

• Interaction effects are insignificant • Slight effect of frequency of

purchasing

Diagnosticity Model 1 Model 2 Model 3 Model 4 Model 5

Constant 1.644 1.693 1.725 1.704 1.716 Objectiveness .411*** .383*** .365*** .365*** .365*** Concreteness .406*** .378*** .377*** .389*** .388*** Linguistic style .246*** .245*** .227*** .239*** .233*** Objectiveness*Concreteness .049 .049 .050 .049 Objectiveness*Linguistic style .031 .031 .031 Concreteness*Linguistic style -.020 -.020

Frequency of online purchases .062**

R2 _.402 _.403 _.403 _.404 _.408

R2_Adjusted _.400 _.399 _.399 _.398 _.400

F 146.926 110.397 88.296 73.509 55.935

Significance .000 .000 .000 .000 .000

(12)

Results – Regression Analysis on Helpfulness

• Only the interaction effect between objectiveness and concreteness is significant

• The effects of control variables are insignificant

Helpfulness Model 1 Model 2 Model 3 Model 4 Model 5

Constant 1.637 1.845 1.856 1.831 1.842 Objectiveness .348*** .248*** .243*** .243*** .243*** Concreteness .335*** .234*** .234*** .247*** .246*** Linguistic style .268*** .264*** .259*** .271*** .266*** Objectiveness*Concreteness .174*** .174*** .175*** .174*** Objectiveness*Linguistic style .009 .010 .009 Concreteness*Linguistic style -.022 -.021

Frequency of online purchases .047

R2 _.311 _.321 _.321 _.321 _.324

R2_Adjusted _.308 _.317 _.316 _.315 _.316

F 98.588 77.369 61.809 51.465 38.933

Significance .000 .000 .000 .000 .000

(13)

Results – Mediation Analysis on Objectiveness

• Indirect effect is

positive and highly

significant

• Small but positive and

significant direct effect

• Higher at high levels of linguistic style

• Strong partial

mediation in all cases

OB CO LI a b ab c (calculated) c’ Objectiveness - 0 0 1.3169*** .8092*** 1.0657 1.2553 .1896* - 0 1 1.1219 1.3765 .2546** - 1 0 1.2459 1.4505 .2046* - 1 1 1.3021 1.5717 .2696**

Figure 4 – Mediation analysis on Objectiveness

(14)

Results – Mediation Analysis on Concreteness

• Highly significant and positive indirect effect

• Positive direct effect is only significant at low levels of linguistic style

• Partial mediation at low levels of linguistic style

• Full mediation at high levels of linguistic style OB CO LI a b ab c (calculated) c’ Concreteness 0 - 0 1.4193*** .8102*** 1.1499 1.3387 .1888* 0 - 1 1.2557 1.2590 .0033 1 - 0 1.3285 1.5361 .2076* 1 - 1 1.4342 1.4564 .0222

Figure 5 – Mediation analysis on Concreteness

(15)

Results – Mediation Analysis on Linguistic style

• Positive indirect effect is highly significant

• Small negative direct effect is only significant in the case of subjective and concrete reviews

• Partial mediation for subjective and concrete reviews

• Full mediation for other three cases OB CO LI a b ab c (calculated) c’ Linguistic style 0 0 -1.0356*** .8102*** .8390 .8225 -.0165 0 1 - .9481 .7461 -.2020* 1 0 - .8956 .9452 .0496 1 1 - 1.0047 .8688 -.1359

Figure 6 – Mediation analysis on Linguistic style

(16)

Results – Mediation Analysis on Argument Quality

• Both the indirect and

direct effects are positive

and highly significant

• Strong partial mediation

• Relatively smaller direct

effect _a _b _ab _{c (calculated)} _c’

X = Argument quality M = Diagnosticity Y = Helpfulness

.8413*** .6658*** .5602 .9384 .3782***

Figure 7 – Mediation analysis on Argument Quality

(17)

Results – Measured and

independent variables

07.07.2016 | 17

argument quality on helpfulness.

H₂: Argument quality is mediating the effects of objectiveness,

concreteness and linguistic style on diagnosticity. Supported

H₃: An increase in the level of argument quality results in a higher level

of helpfulness. Supported

H₄: Higher online consumer review objectiveness leads to higher

argument quality. Supported

H₅: Higher online consumer review objectiveness induces higher

diagnosticity. Supported

H₆: An increase in online consumer review concreteness increases its

argument quality. Supported

H₇: An increase in online consumer review concreteness leads to an

increased perceived diagnosticity. Supported

H₈: Reviews possessing a high quality linguistic style tend to have

higher argument quality. Supported

H₉: Reviews possessing a high quality linguistic style lead to higher

diagnosticity. Supported

• All the hypothesized effects of the measured variables are supported • All the hypothesized main effects of

(18)

07.07.2016 | 18

Results – Interaction effects

the independent variables in the effect of

objectiveness on argument quality. supported H₁₁: There is a positive interaction effect between

the independent variables in the effect of objectiveness on diagnosticity.

Not supported H₁₂: There is a positive interaction effect between

the independent variables in the effect of concreteness on argument quality.

Not supported H₁₃: There is a positive interaction effect between

the independent variables in the effect of concreteness on diagnosticity.

Not supported H₁₄: There is a positive interaction effect between

the independent variables in the influence of linguistic style on argument quality.

Not supported H₁₅: There is a positive interaction effect between

the independent variables in the influence of linguistic style on diagnosticity.

Not supported

• None of the hypothesized

(19)

Results – Control variables

quality than those who shop online less frequently.

H₁₇: Consumers who read online reviews more often, will find online reviews to have higher argument quality than those who read these review less frequently.

Not supported

H₁₈: Consumers who shop online more often, will find online consumer reviews more diagnostic than those who shop online less frequently.

Supported

H₁₉: Consumers who read online reviews more often, will find online reviews more diagnostic than those who read these review less frequently.

Not supported

H₂₀: Consumers who shop online more often, will find online consumer reviews more helpful than those who shop online less frequently.

Not supported

H₂₁: Consumers who read online reviews more often, will find online reviews more helpful than those who read these review less frequently.

Not supported

(20)

Discussion

• In line with theory, the high levels of objectiveness, concreteness and linguistic style

all contribute to online consumer review helpfulness through argument quality and

perceived content diagnosticity

• There is only an interaction effect between objectiveness and concreteness on

helpfulness

• The effects of the independent variables on helpfulness are strongly mediated

through argument quality and diagnosticity

(21)

Managerial implications

• Opinion platforms and webstores could use a more detailed helpfulness voting

system

• Further adjustments for filtering a large number of reviews

• The operators of the sites themselves could filter which reviews would appear to

readers

• Filter, or reduce relevance of reviews with a high number of spelling or

grammar mistakes

(22)

Limitations and further research

• Relatively low number and uneven demographic distribution of respondents • Stimuli unsuitable or uninteresting for many participants

• More popular product categories (e.g. smartphones, tablets) • Only examining search goods

• Testing also for experience goods

• Further comparing the importance of review characteristics • In different product categories

(23)