Product Trials: The Effects of Direct Experience on Product Evaluation

Tilburg University

Product Trials

de Groot, I.M.

Publication date: 2003 Document Version

Publisher's PDF, also known as Version of record Link to publication in Tilburg University Research Portal

Citation for published version (APA):

de Groot, I. M. (2003). Product Trials: The Effects of Direct Experience on Product Evaluation. Optima Grafische Communicatie.

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal

Take down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

.--olo i

UNIVERSITEIT *12!f * VAN TILBURG |

n 'L O

i--RIBLIOTHEEK

L____!-ILBURG

Product Trials:

Groot, I.M. de

ProductTrials:TheEffectsofDirect Experience on ProductEvaluation

Cover Design: A.J. Tas

Printed_{by:Optima GrafischeCommunicatie,} Rotterdam

Product Trials:

The Effects of

Direct Experience on

Product

Evaluation

PROEFSCHRIFT

ter verkrijging van degraad van doctor aan de Universiteit van Tilburg, op gezag van de rector magnificus, prof.dr. F.A. van der Duyn Schouten,

in _{het openbaar te verdedigen ten overstaan van een door}

het college voor promoties aangewezen commissie in de aula van de Universiteit

op maandag 2 juni 2003 om 14.15 uur

door

InekeManon de Groot

Promotores:

Prof.dr. W.F.vanRaaij

Acknowledgments V

Acknowledgments

Tijdenshetafronden vaneen _proefschrift over_{product trials begin je je af}te vragen

of

hetwerken aaneen_{proefschrift eigenlijk niet} 66n_{grote ontdekking is van wat een} onderzoeker te wachten staat: het lezen van literatuur,bedenken van experimenten,

verzorgen van materiaal, vinden van proefpersonen en

_uiteindelijk

_{het rapporteren} vande_{bevindingen. Promoveren is in duseigenlijk}66nlange

trial.

Zonder_{proefpersonen hadden alle experimenten} niet_uitgevoerd kunnen worden en daarom wil

ik

iedereenbedanken die_{heeft meegedaan aan 66n}ofmeerdere van de

experimenten die ikde_{afgelopenjarenheb uitgevoerd:} de_{(destijds) HAVO en VWO}

scholieren van de _{Vlaardingse Openbare Scholengemeenschap (te Vlaardingen) en}

het Libanon _{Lyceum (te} Rotterdam). De studenten van verschillende universiteiten

dieallenbereidwareneenverscheidenheidaan_{goederen uit te proberen.}Verder ben

ik

dank _verschuldigd aan _{bedrijven die gratis producten verschaft} hebben die gebruikt werden inhet onderzoek: BMGHilversum voordemuziek cd's enProcter &

Gamble voorde_Pringles.

Verder wil ik mijn

begeleiders bedanken die mij gedurende de _afgelopen

_jaren

hebben geholpen: Daniel Read, Fred van Raaij en Gerrit Antonides. Ook ben ik Jack

Knetschdankbaar voordediscussies die we_tussentijdshadden.

De enorme steun vanvriendenheeft_{enorm geholpen in alle fasen van}DeTrial. Ik wil danook_{Rapail Akamavibedanken voor de tijd die we op}

de'AiO

zolder'

in

Leeds

doorbrachten en hij voor support zorgde. Ook wil

ik

Young-A Son en _{Maggie Kaska} bedanken. Dankzij hun gezelligheid in huis werden de _{avonden langer en} de _dagen

korter. Natuurlijk mag ikde_volgendevrienden niet_{vergeten. Metzijn} drieSnhebben

we inhetzelfde schuitje gezeten: Marina "Flambina" Pool die bijhet woord'grasveld'

.,

altijd wist hoe laat het was. Met Stdphanie "Pinga" van den Berg enAnnelise "Lies Notenboomwaren er

altijd

wel redenen te bedenken om portte drinken waarbij de

lange_{treinreizen om dit doel} tebereiken nietgeschuwd werden. Verder wil

ik

Joost Loef bedanken voor zijn tips omtrent achtergrondmuziek. Bij het schrijven van dit proefschrift heb ikdaar zekergebruikvan gemaakt.

Tenslotte wil ik

mijn

familie bedanken: Bram, WilenZahuradeGroot._Zijhebben

VI Contents

Contents

1. Introduction 11

Introduction 12

AdvantagesandDisadvantages of DistantSelling 12

Definition of Risk 13

Taxonomy ofLosses 14

Benefits of SalesStrategies 15

ResearchonSalesStrategies 16

Objective of theThesis 17

Managerial Relevance 17

ScientificRelevance 18

Overview of theThesis 18

2.

Product

Trials and

Product

Evaluation 21

Introduction 22

Waysto ReducePerceived Risk 22

Rationale_{for Studying Product}Trials 23

Direct and Indirect ProductExperience 23

Imperfect Informationin_{Consumer Judgment 25}

A Three-StageModelofTrial Behaviour 27

AttitudeandAttitude Formation 29

The Elaboration Likelihood ModelofPersuasion 31

Implications oftheEndowmentEffectfor ProductTrials 33

Attribute Information Typesand Related ConsumptionBehaviour 34

Attitude Formationin Product Trials 34

Factors inthe Trial_{Stage 36}

Cognition andAffect 36

ExpectationsandDisconfirmation 39

Overcoming PoorPerformance 40

FomardandBackward Effects of ExpectationsonSatisfaction 41

Personal Involvementand InformationAcceptance 43

Involvement 43

Factors inthe Post-TrialStage 45

SatisfactionandDissatisfaction 45

Contents VII

Summary 47

3. Perspectives

on

Object Valuation and the Endowment

Effect 49

Introduction 50

Expected Utility Theory asaNormative Theory of Choice Behaviour 50

FromExpectedUtility Theory to Prospect Theory 52

The Implications of Prospect Theory forthe_{Theory of Choice Behaviour 55}

ReferencePoints 55

Reference Pointsand Framing Effects 55

Labelling Changes from the ReferencePoint 58

Value

_{Function 60}

Decision_{Weights 61}

The_{Asymmetrical Treatment of}Gains andLosses in aRiskless Context 62

The EndowmentEffect 62

Causes oftheEndowmentEffect 65

Cognitive Explanations of the EndowmentEffect 66

Cognitive DissonanceandConsolidation Processes 66

Affective Explanations of the EndowmentEffect 67

Loss Aversion 67

Mere Exposure 69

Mere Possession 69

Mere Ownership 70

Attachment 70

Conditions Affecting the Strength of the EndowmentEffect 70

Transaction Demand 70

Time 71

Product Related

Factors 71

The Transaction between MoneyandGoods:QuadrantsandMeasures 73

Implications of the EndowmentEffect 75

Summary 77

4 The Role of the Endowment Effect

in

Sales Strategies 79

TheoryandHypothesis 80

Overview of theExperiments 82

Experiment 1 83

Introduction 83

VIll _Contents

ResultsandDiscussion 85

Experiment 2 86

Introduction 86

Method 86

Resultsand _{Discussion 88}

Experiment 3 90

Introduction 90

Method 90

Resultsand _{Discussion 91}

Experiment 4 95

Introduction 95

Method 95

ResultsandDiscussion 97

Experiment5 100

Introduction 100

Method 100

Resultsand Discussion 101

Limitations 110

Generaldiscussion 111

Implications 113

5 Measuring the

Value

of Product

Trial

115

Introduction 116

Experiment6 117

Method 117

Resultsand Discussion 117

Experiment7 119

Method 119

ResultsandDiscussion 119

General Discussion 120

6 Expectation

and

Disconfirmation Effects for Trial Products

121

Introduction 122

Theoryand_Hypothesis 122

When Expectations Dominate Disconfirmation 123

Contents IX

Which is More LikelytoDominatein _{Product Trials: Expectations or} 126

Disconfirmation?

Experiment8 128

Introduction 128

Method 128

ResultsandDiscussion 130

GeneralDiscussion 137

7

General

Conclusions

and Final Remarks

141

General Conclusions 142

The Relationship Between Product Trials and the Endowment Effect 142

Explaining Product Devaluation after Product Trial 143

The Role of Expectations 145

FinalRemarks 145

Limitations 147

Implications for the Marketplace 147

Future _{Research Suggestions} 148

References

Chapter 1

12 Chapter 1

Introduction

On 15 March

1962, President JF Kennedy

declared to the

US Congress that

consumers havefourbasic_rights.Theseinclude:'

The right

to

safety. To

be protected against products, _{production methods and} serviceswhicharehazardousto_{health or life;}

The right to

be

informed. To

begiven the facts needed to make informed_choice,

to be protected against dishonestor misleading advertisingandlabelling;

The right

to choose. To be able to select from a range

of

products and _services, offered at_{competitive prices with}anassurance

of

_{satisfactoryquality;}

The right to

be

heard. To

haveconsumerinterests represented in themaking and

execution

of

_{government policy, and in}the_development

of

_{products and}services.

Of these

four

basic rights, the focus here will be on consumers'

right

to information.

Namely, despitethis right,the

introduction of

new selling methodsbrings along that

consumers have

little

information _{available. For example,} mail order _companies

started to offer an easy shopping method: choose what you want from a catalogue,

send_{your order by direct mail or over the phone, anddelivery}

_will

follow some days later. Likewise, with the _{introduction (and} the

popularity) of

the Internet as a new medium, buyingproductsbecameeasier. Witha_{single click on the}PCconsumers can nowadays buy all sorts

of

productsonline. Most popularcategories tobuyonline from arebooks, computer_{hardware/software, and music/video(Rasch& Lintner, 2001).}

Thus, increasingly, retailers do not

only

offer their goods in-store. Additionally, they have extended their services by selling

their

goods via methods that require a

less_{direct relationship}betweenretailer and consumer. Werefertothesemethods as

distant

selling methods.2

Advantages and Disadvantages of Distant Selling

Distant selling methods are _{becoming more popular among consumers.} Online

selling,

for

example, offers many advantages. In Europe, the top five reasons for buying onlinecited by consumersare(Rasch& Lintner,2001):

• The ability_{to browse and purchase}online;

• The ability to save time;

• The ease of theonlinepurchase process;

'

TheInternationalOrganisationofConsumerUnionshas added thefollowing four rights: the rightto satisfactionofbasic needs,therighttoredress,theright_{to education, and the}right to

a_{healthy environment.} 2

Introduction 13

• The access to products not availableoffline;

• The ease

of

comparingprices and features.

There are also disadvantages _{of Internet (or} catalogue) purchases. For example, people havebeenfoundto_{dislike (Smith &}Swinyard, 2001):

• Theadditionalcosts_{(shippingcharges}and_handling);

• The difficulty of judging the quality of products as physical inspection is _highly restricted;

• Thetransactioncostsof returningunwanted items.

Fromaretailer's point of view, the advantageof distant selling is that it cuts down

thetotalcosts.Salespersons arenot_{needed and high rent rates (as}inshoppingmalls)

are _{avoided. Instead, goods} are storedin large warehouses rather than in expensive displays.3 On the other hand, distant selling couldbe disadvantageous

for

retailers, because interaction between a salesperson and

customer does not

take place.

Interaction effects, such as persuasion of consumers _{(Cialdini, 200la, 2O01b), or} impulsebuying(e.g., _{Dittmar, Beattie,} &Friese, 1996),which usually work

in

favour

of

theretailer,areabsent.

Thus, although the development of new selling methods is

attractive for both

consumers and retailers, there are also disadvantages such as lack

of

interaction

effects (retailers), and lack

of

important information (consumers).

In

general, if the consumer has no _{product information and, consequently,} is unable to

_{judge the}

product, he or she

will

perceive the purchaseasriskier._{The concept}

of

consumer risk

is discussed next. Definition of Risk

Risk and attempts to reduceits impactispervasiveinsociety. Forexample, youmight

not want to risk arriving late at work, so you go bycar.Alternatively, an investor on

the stock market may

intentionally take the risk

of losing money when he buys particularstocks. Nowadays, cigarette packageswarn (potential) smokers of the bad

consequences of smoking withtextsloganslike: "Smokingenhances the riskof heart

andvascular diseases".

A

music chain inThe Netherlands advertises with a 'no-risk disk': when thepurchased CD is unsatisfactory, consumers mayclaim the purchase price back. These

examples show the use of the word risk

in different contexts

(varying from financial riskto_{health risk). What}theseexamples have in common is that they all involve a negative feature ofa _{decision, or}the chance of some loss (of

3 This is how Dell Computers explains whytheir pricesarelowerthan those ofcompetitors:

14 Chapter 1

money, health) (Yates & Stone, 1992). Despite many

definitions of

risk, Yates and Stone (1992) propose that theriskconstruct always contains: (1) potentiallosses, (2) the significanceofthose losses, and (3)the uncertainty

of

thoselosses.

The important distinction between riskand uncertainty is that the

probability of

occurrenceis unknownfor uncertainty,

but

known for risk(seeAntonides, 1996). In addition to this statisticaldifference, Slovic, Fischhoff, andLichtenstein(1984)found

that the statistical difference between risk and uncertainty is reflected

in

people's perception

of

events._They found that the risk construct consists of twodimensions:

the 'dread risk' factor

(first

dimension), and the

'unknown risk'

factor (second

dimension). Especially the second dimension distinguishes

between risk and

uncertainty. The endpoints of this dimension are defined by, among others, 'not

observable'vs.'observable'. Associated

with

these_endpoints areothercharacteristics such as 'unknown to _{those exposed' (vs.} 'known to those exposed'), 'effect delayed' (vs.'effectimmediate'), 'new risk' (vs. 'old risk'), etc.Clearly, theendpointsrepresent the difference between

events with

an uncertain probability of occurrence

(uncertainty), and events with a known probability

of

_{occurrence (risk). Thus,} uncertainty and risk are two endpoints ofacontinuum,and peopleassociatedifferent

characteristics ofa_{situation with uncertainty and risk.}

Relating this definition of risk (vs. uncertainty) to the application of sales

strategies, we may see that

most techniques are

aimed to

move consumers

perception ofa_purchasesituation from uncertainty to risk, or even from uncertainty to certainty. This research focuses on the

application of one of

such techniques designedtoestablish this,the product

trial.

Taxonomy of Losses

Assaid_above,riskinvolves some kind of loss. Whenwethink of a lossweoften

think

of something we used to have but do not

have anymore. In other words, when

experiencing a loss we may feel deprived of something we had in the past. Another

loss experience may occur when we end up with an outcome that is perceived as worse than before, e.g., the amount of free time we experience when we work (vs. when we do not). In order toassesswhether_{something is perceived as a loss we need}

areferencepoint.Thenotion ofareferencepoint willbediscussedinChapter 3. Losses may occur in different domains. Jacoby and Kaplan (1972) constructed a

taxonomy

of

losses. Their classification mainly applies to consumer behaviour and

contains financial _{losses, performance} loss, physical loss, psychological loss, social loss, and time loss. The taxonomymainlyapplies to consumer behaviour. We speak

offinancial

losseswhen consumerslosemoneybecausetheproduct does not work at

all, because it costs a lot to maintain the _{product, or when} an equal or better

Introduction 15

Notethat _performancelosses andfinanciallosses go hand in hand. For example, we may spend additional money to replace or fix a deficit product. An example of a financial loss

without

a _{performance loss is the purchase of} a _non-winning

_lottery

ticket. Physical losses are losses resulting from unsafe products (chemical or

toxic

liquids), orproducts that threaten our health (e.g., cigarettes). Psychological losses

follow from

a _{discrepancy between} a

_{product and}

a consumer's _{self-image. For} example, someone who aspires to be a vegetarian, but buys fast-food hamburgers

twice a week

will

experience a psychological loss when

confronted with his food

intake. Social lossesare losses _{resulting from}a _{negative way of how}people

think of

you.Thisnegativeperception mayfollowwhen people associate you witha

_particular

(purchase of)aproduct.For example,buying marihuanamay negativelyinfluence the way people

think

about you. Last,there is time (and e#ort) loss. It is the loss of time

when consumers need to return, replace, or repair, _etc., products. The focus in this

research is on avoiding the financial loss,i.e., offeringsales_{promotions to}consumers as a waytoreduce

their

perceived risk ofa_purchase.

Benefits ofSales_Strategies

Sales _{promotions, such} as _{price discounts, help both} consumers and retailers. Irrespective ofthe appliedpromotion technique theygive consumers the

opportunity

to learnabout a product. As with many otherfactors, uncertaintyabouta product's

possible benefits in the future may strongly

inhibit

consumers' willingness to buy.

Together with otherfactors (e.g.,budgetrestrictions) this uncertainty contributes to the perceived risk ofaparticularpurchase. In thedomainofconsumerbehaviour the perceived risk ofa _{purchase often involveslosingmoney, or}afinancialloss_{(Yates &} Stone, 1992).By making use ofsales strategies consumerscan reduce the risk of the negativefinancialconsequences

of their

purchase decisions(Roselius, 1971).

Given the rapid development of new products, extended products, or improved

products, consumers mayhave_great

difficulty

indeciding what to buy.Asmentioned earlier,retailersare_{obliged toprovideproduct information in order for}consumers to makebetter decisions. The_application

of

salesstrategies, especiallythose techniques

which provide product information before the purchase decision, is a good way for retailers to provide enough product information. Especially when the promotion techniqueisperceived as a'lowrisk'strategy, consumers may become more

willing to

try (e.g., product sampling, product

trials) or to buy

(e.g., in-store coupons,

money-back guarantee). Thus, fromaconsumer's

_{point of}

view, sales_{strategiesmayprovide}

product information while froma retailer's point of view,

it

_{provides consumers'legal}

right

toinformation. Table 1.1. summarises common sales strategies as well as

their

objectives

(trial and/or

repeatpurchase)ofthese techniques. Generally,

trial

isaimed

16 Chapter 1

Table 1.1. Examples

of

salesstrategies and objectives

(adapted from: Williams, 1996).

Sales_promotion Trial _{Repeat purchase} technique

Price

_{discount 4}

Couponing 4 4

Free_{product packs; 9/ 14} Bonus packs In-store 4 demonstrations Sampling 4

Cash refunds 4

Buy backs Mail-in premiums 4 4

Research on Sales_Strategies

Despite the existence and _{application of many}sales strategies (for a more complete overview see Williams, 1996) these techniques have rarely been studied

in

terms of

how _they affect _product choice _{or product evaluation. Searching the}

literature it

appears that promotion

techniques are not the

main focus of many researchers. When studies report on retailing techniques they mainly focus on modelling the

technique and hardly on possible effects of the techniques on actual consumer behaviour.

The most researched techniques found are product trials (Dubas, Dubas, &

Atwong, _1999; McGuiness, Gendall, & Mathew, 1992; Goering, 1985; Dickinson &

Wilby, 1997; Scott _{& Yalch, 1980; Simonson, Carmon, &} O'Curry, 1994) and to a

lesser extent money-back guarantees (Heiman, MeWilliams, & Zilberman, 2001; Mann & Wissking, 1988; Moorty & Srinivasan, 1995; Davis, Gerstner, & Hagerty, 1995)· Closely related to product-trial research is research on direct vs. indirect productexperience(e.g., Fazio_{& Zanna, 1978,1981).}

Most studies on salesstrategieshave focusedon building models to describe how

the techniques will work, or have focused on consumers' covert behaviour, such as attitudes and beliefs. When the

effects of,

for example, product

trials were

Introduction 17

Actual buyingdecisionshave_rarelybeenstudied.

Thus, the gap in theliterature is the lack ofexperimental research inthedomain of how sales strategies, such as product trials and money-back guarantees,

will

affect consumers' actual purchase decisions. Although money-back guarantees are

considered in the

first

twoexperiments of_{the thesis, the}mainfocusshiftstoproduct trialsand

their

effectsonconsumers'productevaluation

in

Experiments 3-7.

Objective of

the

Thesis

The sales strategies presented here were introduced as ways (for consumers) to reduce theperceived risk in buying decisions. Claims in the psychological literature

(PloUS, 1993) as well as in the management literature (Smith &Nagle, 1995) bring

forth another perspective: not only are these techniques useful ways to reduce perceived risk

for

consumers, they are alsosubject to psychological effects, which are

thought to be advantageous

for

retailers. Namely, it is expected that product trials

will

give consumers the sensation of ownership, thereby shifting

their

reference

point. This shift

in

reference

point will

then cause an endowment effect and,

consequently,

result in

an increased

liking for

the

trial

goods.

Without

going into detail right now, the endowment effect is the observation that people value goods, once

in

possession, more highly than goods not in possession _{(Thaler, 1980).}

Additionally, given the lack

of

experimental studies on the effectiveness

of

product trials (orsalesstrategies

in

general), the aim of thisthesis is tostudy experimentally

the effectiveness

of

product

trials and, to

a lesser extent, money-back guarantees. This thesis reports aseries

of

experiments to testthe effects of temporalendowment

in

the product

trial

technique. In areal-purchase context, the effectof product trials

is studied.

Managerial Relevance

The applicationof risk-reducingtechniques canbeadvantageous forbothconsumers and retailers. Consumers may reduce the perceived risk and retailers can make

products more

familiar

toconsumers. Giventhe application

of

product

trials it

is

still

not known to what extent product trials are successful techniques. For example, are

product trials by definition successful, or is a successful application dependent on otherfactors?What determinesthe successful_application

of

product trials? How do product trialsaffect consumers'productevaluation(attitudes,monetaryevaluation).

This researchhasseveral managerialimplications. A

first

implication follows from our study onthe effects

of

product trialsonattitudes and economicvaluation. These I are two different aspects

of

product evaluation: attitudes are more indicative of an

18 Chapter 1

behaviour_{(e.g.,purchase}likelihood). Asecond_{implication follows from ourstudy on}

conditions under which product

trials may be more (or

_{less) successful. First, we}

focus on the type

of

trial

product. Knowledge about what typeof product will be more

suitable to use as a

trial

product may help retailers to decide on whetherto offer a product

trial or

not. _{Second, we}study the roleofexpectations. Expectationsareoften created by television commercials or

printed ads, and

may influence people's attitudes. For example, a TV commercial for shampoo showing happy _{women with}

shiny hair

may create high expectations (and a favourable

_{attitude) of the}

effectiveness of the shampoo. Consumers may then use aproduct

_{trial to find out if,}

and to what extent, their attitudes are justified. Thus, knowledge about the role of

expectations in product trials givesretailers insight into whetherexpectations affect

productevaluation,whether_{direct experience affectsproduct evaluation, or that both} contribute to productevaluation.

ScientificRelevance

The effectiveness

of

producttrials is aless_{studied (if}not_{unstudied at all) topic in the}

marketing andthe psychological literature. In the marketing, orconsumerbehaviour

literature we

find hypotheses of how product

trials may work (e.g., Hoch &

Loewenstein, 1991). However, these

hypotheses have not yet

been tested

experimentally.

This research helps to complete the literatureon product trials. First,

it

provides

experimental evidence of the working

of

product trials. Second,

it

contributes to a

better understanding of the role

of

endowment in promotion techniques. It makes use ofan important phenomenon inthedecision-making literature, the endowment

effect, to test theeffectivenessof product trials.

Until now it

has oftenbeen thought that product trials benefit from the endowment effect (Plous, 1992; Smith & Nagle, 1995) andthisseriesofexperiments teststhis assumption. To conclude,thisresearch

has implications for

the marketing literature (adding experimental research to hypothetical research), and the social and economic psychological literature (testing an application of theendowment effect ondirect productexperience).

Overview of

the

Thesis

Thisdissertationconsists ofsevenchapters. Chapter2 gives atheoretical description of product trials as one ofseveral retailing techniques. It starts with a comparison

between different retailing techniques and gives a rationale for a study on product trials. The concept of attitude is introduced and product trials are discussed in

relation to

a theory of attitude formation, the elaboration likelihood

model of

Introduction 19

for

_{product trials this chapter} continues with a discussion

of

different aspects of attitude formation

in

(different stages of) a product

trial

and

their role

in product trials.

Chapter3 gives possible explanations for the endowment effect, which serves as

the basis for the assumed effectiveness of product trials. It discusses terms such as prospect theory, value function, loss aversion, reference point, framing effects, and explains how prospect theory may affect product evaluation. The term endowment effectisintroduced andanoverviewof_{experiments on the endowment effect, as well} as implications oftheendowment effect

for

differentdomainsisgiven.

Chapter 4 reports five experiments designed to test the idea that theendowment

effect would make the evaluation

of trial

products more favourable.4Experiments 1-3

measure consumer willingness to pay (WTP)

for

different consumergoods (CDs in Experiment i and2;variety ofgoods

in

Experiment 3) as afunctionofsalesstrategy

(product

trial,

money-back guarantee, normal sale). Product trials performed not

betterthan money-back guarantees; theyeven performed worse than whennosales promotion techniquewasoffered. Experiment₄tests the role ofactual ownership in

trial-based product evaluation. To complete our research design in line with a more

valid test for

the endowment effect, Experiment 5 was

conducted to test the

relationship between product

trial

and endowment, and how this affects economic

productevaluation. Evidence that the endowment effect had occurredduring product

trial,

therebyincreasing the value of the good, wasnot found.Giventheresults of the analyses, additional structural equation modelling (SEM) analyses were performed

where people'sattitudes_{were assigned the role}of mediatorvariable.

A

two-foldeffect

of

_product

trial

oneconomicproduct valuationwasfound.

In Chapter 5 two experiments are reported that seek explanations for the poor

performanceof product trials. In Experiment 6 itwas argued that the status ofatrial

good ('second-hand' after

trial)

could explain this finding. Indeed, when subjects

were asked

for

their WTP for

a brand new

radio and a

test

_{radio, WTP was}

significantly lower for thetest radio.Thiseffect

of trial on

WTPimplies thatthe

trial

itself canbe represented bysomepositivevalue. Experiment₇testedthe value of the

trial

itself, and showed that

trial

can indeed be represented by a positive economic

valuation.

In Chapter 6 the role

of

expectations (Oliver, 1980) prior toaproduct

trial

istaken into account. Two alternative hypotheses are tested: the perceived performance hypothesis and thedisconfirmation ofexpectations hypothesis. Theformerstates that consumersassigngreaterweight tothedisconfirmation of expectations in satisfaction judgments thanto actual performance,while the latterassumes the opposite: actual performance receives greater weight in satisfactionjudgment than disconfirmation. Neither _hypothesis was confirmed. _{Both expectations and} disconfirmation were

20 Chapter 1

found to contribute to attitude

_{formation, with}

a two-fold effect

of

expectations on

attitude formation.

To conclude the thesis, Chapter7gives general conclusions andadiscussion of all

Chapter 2

22 Chapter 2

Introduction

Consumers want to spendtheir money in the best possibleway, given

their

budget

constraint. Many purchasesareroutine, and apart fromthefirstpurchase, consumers

buy them without thinking much. Routine purchases are common low-priced,

frequently purchased products, or

fast-moving

consumer goods (e.g., toothpaste,

washing-up liquid).5 When these types

of products do not

meet expectations the

financial consequencesare minimal.However,whenproducts areunknown or more expensive, consumers become more reluctant to buy, or they want more product information_{before they make} a decision. Lack

of

_{product information, especially for}

expensive goods, generates perceived risk on the consumer's side, and the retailer's task is to reduce this perceived risk by providing

information, or

by reducing the potential financial loss. There are different retailing tactics, each of which reduces perceived risk in its own way. Although the strategies are closely related, there are someessentialdifferences that willbediscussed inthischapter. Further, thischapter

will

also

_justify

a_studyon_producttrials as well asthefactors in two ofthe

trial

_stages

ofconsideration. Finally, this chapter ends witha_{discussion concerningwhy product}

trials might beaneffectivetactic froma_{decision-makingperspective.}

Waysto ReducePerceived Risk

Several strategies may be _{employed to reduce risk,} including _rebates, product

samples, _{product trials,} and money-back guarantees. When offered a _rebate,

consumers arepromised to receive (part of) the purchase price backafter purchase. Theonly requirement is to send in acoupon that goes withtheproduct.The product

mustbeboughtbeforethe coupon can be sent in,solimited product information may

be available at themoment

of

_purchase.

_{In product}

_sampling,oftenapplied with food or drinks, consumers receive a trial-size

portion of

a product. The aim

of

product sampling is togenerate product trial so

that

consumerswouldbecome

familiar with

and, finally, committed tothe

trial

product (MeGuiness, Gendall, & Mathew, 1992).

Information

about a product is not available at the moment ofsampling (assuming

the sampling of unknownproducts). Examples

of

productsampling are the provision offreeCD-ROMs_{with computer}magazines, andprovidingsnacks ina_{shopping mall.}

After product sampling the actual

product

trial

takes place and consumers can

interact with the _{product (trying out the software, or tasting} the snack). The trial

provides information about, and direct experience with the _product and consumers are informed about its properties. Sampling and

trial

are closely related. Sampling

involves only the distribution

of

products _{among consumers, and the}

trial

allows consumers to assess a _{product's properties.} Thus, samplingprecedes trial. The last

Product Trialsand ProductEvaluation 23

strategy is the money-back guarantee. Consumers areofferedthe guarantee that the

full

retail

_{price will}

be

refunded if

the _{product is not satisfactory. Like} free trials,

consumers whoareofferedamoney-back guarantee are abletolearn abouta product (information acquisition), but unlikefreetrials_{the money is paid up}front. The main

difference between free trials and money-back guarantees is thus the

moment of

paymentrelative tothe moment

of

information acquisition. It is only withfree

trials

that the moment of payment takes place

_after

_{information acquisition. Thus,} consumers will form

their

attitudestowardaproduct_{beforepurchase with}a _product trial, and

_{after purchase with a money-back guarantee. Finally, there is the}

no-tactics-at-all wayofsellingproducts. Itisreferred to asa

normal

sale. Consumers are

neither promised a

trial,

guarantee, or refund. They buy a

product. Of all tactics

discussed here, the risk of a bad purchase is highest with a normal sale. Table 2.1 summarises the different retail tactics and how different

factors play a role in the

purchaseprocess.

Rationale for Studying Product Trials

A product

trial

gives the consumer the opportunity to learn about, or receive

information about a product before the actual purchase. Product trials are often

offered

_{in-store, as well as at}

home.

Apart from

the transaction

costs for the

consumer, namely, the consumers must repackagetheproductand return_{it, product}

trialsare_highlyadvantageous

for

consumersbecause

it

givesthem plenty of time and

opportunity to

interact with

the product before deciding on a purchase. This opportunity reduces, among others, the risk ofa financial

loss:

In addition, product trials_{provide reliable and}direct information totheconsumer about theproduct.

From a retailer's

point of

view product trials could serve as a _{signal to the} consumer that the

product is of

a certain level of quality (a

retailer will

only offer

product

trials for

good products). This should create confidence at the consumer's side. The down side ofthis tactic is that theretailer runs the risk

of

ending up with a pile of_used, unsoldproduct that did not meet expectations. Reviewingthe research onproducttrialsthere are two reasonsthatsupport the choice fora study onproduct trials.Thefirst isthe differences between directandindirectexperience on consumer

judgments. The second is the typeofjudgment that is used inproduct trials.

Direct and Indirect Product Experience

Information is better accepted when it is obtained by first-hand experience (direct experience) than when it is obtained by other sources (indirect experience). Also, direct _{experience is considered more reliable} than _{indirect experience. Hence, it is} 6

Othertypesoflossesare discussed in Chapter 1 (cf., Yates, 1992; Roselius, 1971; Jacoby &

24 Chapter 2

Table2.1. Retail tactics used to reduce perceived risk and the role

of

differentfactors

in_{the purchaseprocess}

Coupon/ Money-back

Factor Rebate Product

trial

_guarantee Normal sale

Direct

After

purchase _Before

_After

purchase _Afterpurchase

product purchase through through

experience decision interaction interaction through

trial

Accessto

After

purchase Before

After

purchase

After

purchase

direct _purchase

product decision

experience

Way of risk Throughprice Through Purchase can be No

reduction reduction disclosureof undone

product information

Potentialrisk _{Coupon is}not Norisk _{Money is}not Product does

for

consumer sent

in

because claimedback _{not satisfy, i.e.,}

of

transaction because

_ofhigh

financial loss.

Costs. transaction

Costs

Commitment Yes No No Yes

to purchase

expected that product trials are perceived as more reliable sources of information

than othersources (e.g.,printed _ads).Indeed,information isbetteracceptedthrough

trial

than through indirectsources(Smith & Swinyard, 1982,Wright &Lynch, 1995). Moreover, there is evidence that direct experience yields different results than

advertising. Attitudes toward a food snack were more extreme (e.g., theextremity of

affect) after

trial of

the snack than after

exposure to an ad of it.

_{In addition,}

confidence with one's own attitude was also _{higher after}

trial

than after_{exposure to}

ads (Smith &_{Swinyard, 1982,1983;} Marks & Kamins, 1988).7Moststudies ondirect

vs. indirect _experience have focused on the contribution

of

_{cognitions and affect}

(Fazio & Zanna, 1978, 1981;

Millar

& Millar 1996) in consumerjudgments. Results

Product Trialsand ProductEvaluation 25

fromtheseexperiments showed, among others, that the role

of

cognitions and affect

depends onwhether direct or indirectjudgments were made. Also,directandindirect experience predicteddifferent types ofconsumerbehaviour

_(Millar

&Millar, 1996). What these studies have not investigated is whether and how the type of product experience affects subsequent purchase behaviour (e.g., WTP). Should retailers provide freeproduct trials at all?

Imperfect Information in Consumer Judgment

By studying product trials we might gain

insight into

how missing information

influences consumers' evaluation of an item when it is the only option (non-comparative

judgment) compared to when it is one of

several in a set of goods

(comparativejudgment)(Kivetz &Simonson, 2000; Moon &Tikoo, 1997).

Hsee (1996)differentiates between_{judgments following from}

_joint

evaluation and

judgment following separate evaluation. In

joint

evaluation (JE) there are several options available and onecan compare (attributes of) one_{option with (attributes of)} another option. In separate evaluation (SE) there is only one option available and

judgment is made

without

any other available alternatives to compare with. The

evaluation mode has an _{effect on object evaluation. For example, Hsee (1998)} asked subjects

in

different conditionstoindicate how much theywerewilling to pay for ice

cream. In the JE condition subjects werepresented

with

drawings of two servings of ice cream: 8 oz. oficecream in a 10 oz. cup(underfilled), and 7 oz. oficecream in a 5

oz. cup_{(overfilled). In each of the} SEconditionssubjects werepresented with one of

the two servings and asked how much they were

willing to pay for the

ice cream. It was found that in JE people werewilling to pay more for the 8 oz. ice _{crearn/10 oz.}

cup, while inSEpeople werewilling to pay more for the 5 oz./7 oz. cup. Obviously, in

JE people had focused more on the amount of ice cream,

while in

SE people had

focused more on the

amount of

ice cream relative to the size of the cup. In other

words, whenmakingSE_people alsofocus onirrelevant attributes (such as the size of the ice cup),while in JEpeople arelesslikely topayattentiontoirrelevant attributes.

In addition

to above example, the following example also

illustrates the two

judgments types. When deciding on the purchase of a cell phone we

might find

ourselves _{contemplating} one _{particular mobile} phone (A). Suppose we know its

stand-bytime,storagecapacity

for

phonenumbers, but not itsreceptionquality. Our evaluation of A depends on the weight assigned to the separate _{attributes, and the} decision is made onthebasis _{of comparison of}theattribute values ofone single cell phone (separate evaluation). In another situationwe might

think of

the purchase of

oneof_{three cell phones (including}A). Again, there are thethree attributes stand-by

time, storage capacity, and reception quality. Assume that

for

neither ofthe phones

we haveinformation of all three attributes. Forphone A, we knowthe stand-by time

26 Chapter 2

Table 2.2 Example of choices

with

_{missing information (Kivetz &} Simonson, 2000)

PortablePC PortablePC Portable PC

A B C

Speed (Information 166

MHz

100 MHz

(range: 85 -200 _unavailable)

MHz)

Memory 12 MBRAM

_{(Information 24 MB RAM}

(range: 4 - 32 MB _unavailable)

Ram)

Battery life 8hours 3 hours

(Information

(range: 1 - 11_{hours) unavailable)}

and

for

_{phone C the reception quality and storage capacity} is

known. Now our

evaluation ofthe attributes

of

phoneAinvolvesa comparison between theattributes

of

three phones

(joint

_evaluation),

rather than

a trade-off between the attribute

values

of

_{phone A only.}

In general, the evaluation of attribute performance ofoneoption (and hence the overall _{evaluation of that option) depends} on whether the evaluation is made in isolation, or simultaneously. When making separate evaluations, people focus more

on attributes for which it is easyto determine independently how gooda particular

value is _{(easy-to-evaluate}

_attribute).

When _making

_joint

_{evaluations, they focus} more onattributes

for

which it is hardto determine_{independently how good}avalue is_{(hard-to-evaluate}attribute) (e.g., Hsee, 1996; Hsee &_{LeClerc, 1998). Examples of}

easy-to-evaluateattributes are form_{and weight.Qualitative properties, like reception} quality, ornumber

of

entries ina _dictionaryarehard-to-evaluate attributes.Theshift in the relative impact of the two type

of

attributes in judgment is referred to as the

evaluability hypothesis (Hsee, 1996).

The type

of

judgmentbeing made not onlyaffects theevaluation ofan option, it also affectstheweightassignedtodifferenttypes

of

attributes. Kivetz and Simonson

(2000) studied

the effects of missing information on consumer choice. They

presenteddifferentgroups

of

subjects

with

differentpairs

of

options andthe subjects'

task was tomake binarychoices.An example ofastimulus (portable PC)isshown in

Table 2.2. For example, one group chose _{between PC A and B,} a second group between B and C, anda

third

group between PC A and C. Using a variety of stimuli

they found

that common attributes (values

available for

all

options) were 2

significantly more often used than _unique attributes (values not available for all

Product Trialsand ProductEvaluation 27

attribute of

an _option on product judgment was dependent on whether that option was superioror inferior onthe common attribute. When an option was _{superior on} thecommon attribute (e.g., Aissuperior to B on thecommon attribute 'battery life'),

missing

information on

the unique attribute (Speed) was not a _significant reason

againstthat option. But whenanoptionwas inferior on thecommonattribute (e.g., A is inferior to C onthecommon attribute 'memory')missinginformation (Speed) was

found a _significantreason against that option. Thus, theattractiveness ofan option

may change when anoptionbecomes

inferior

(or superior) onthe commonattribute.

Also, missing information was found to result in intransitive preferences. For

example, 62%preferred A to B, 64%preferred B to C, but79%preferred C to A in the PC example.

In _{another study on missing}information and _{product valuation, Moon}andTikoo

(1997) found that

when subjects perceived attributes as correlated (a camera's

durability and

lens _{sharpness), they used} a correlated

attribute from the same

product (e.g., lens sharpness) to infer the

value of

a missing

attribute (e.g.,

durability). Information from other products' attributes (e.g.,lens sharpness of other

cameras) was only used when the variance of that attribute was lowacross products (i.e., low inter-brand variance). In the camera example, people only used other

cameras' lens sharpness valuesto inferthe

durability of

a camera when thevalues of

lens sharpness didnot differ too much.

In sum, despite the large amount

of

research on the effects of directand indirect

product experience on consumer choice, research that relates the role of direct and

indirect experienceto actual purchasebehaviouris

still

lacking. Last, much research

has investigated the impact of missing information on product evaluation when multiple options were available. Using only one _option, a study on _product

trials

wouldgive

insight into

how _missing information

within

a _{single option may affects}

consumers'_judgmentsand purchase decisions.

A_{Three-Stage Model}of Trial Behaviour

In this section we

will

describe a model of

trial

behaviour. We conceptualise it as a three-stage model as shown in Figure 2.1. In each stage factors

related to both

cognitionandemotions arepresent. First, consumers must havea minimum level of

interest in

a product

before they will try it. One way

to induce

trial

is through

sampling. Presumably the morecostly it istoundertake the trial, themoreinterest is

required. However, the prior interest does not have to be very high to get people to try something new. For instance, product samplinghasproven to beaneffective way to persuade peopleinto

trial

(McGuiness et al., 1992). The focus is therefore not on how people canbeinduced to try, but on theprocesses oncethe

trial

hascommenced.

A product

trial

provides newinformation and theconsumer's task is to process the

28 Chapter 2

(IPre-#iallitage l Tdallstage l cpost-Irialstage

No or indirect product Direct product experience Evaluation of direct product

experience experience

Cognitions Cognitions Cognitions

• Perception of a • Processingincoming • Finalobject evaluation need information • Purchasedecision

• Initialopinion • Integrating new information • Updating initial

opinion • Disconfirmation

judgment

Emotions Emotions Emotions

• _{Curiosity • Attachment • Satisfaction/} dissatisfaction

Effects resulting in Effects resulting inparticular particular emotions emotions

• Mere-exposureeffect • Endowment effect • Mere-possession

effect

Figure 2.1.Athree-stage model

of trial

behaviour

flow of informationrequiresthatconsumers update

their

attitudes in_{each stages i.e.,}

each time a newpiece

of

information is encountered. Like purchase behaviour,

trial

behaviour canbecategorisedintothreestages: apre-trialstage,theactual

trial

stage and apost-trial stage: Inthe

_pre-trial

_stagethe consumer_mightconsider whether he or she is interested enough in an object to engage in a product

trial. It is

the stage where _{interest, need, or}

_{curiosity for}

a _product is induced _{through advertising,}

product sampling, etc. Sufficiently interested consumers move to the

trial

_stage where _{they actively interact with the object in} the store or at home. The consumer processes incoming

information that will

be _{integratedinto existingknowledge. The}

integration of the new information is _cognitive

in

nature and is a _{cyclical, step-wise}

process: new

information-integration-attitude

updating-new

information, etc.

Emotions in thisstage arethoseproduced_{by, among others, mere-exposure (Zajonc,}

1968), mere-possession (e.g., Sen & Johnson, 1997) andproduct attachment(Schultz Kleine,

Kleine III,

&_{Allen, 1995).} After the

trial

stagethe consumermoves into the

post-trial stage in which the cognitive

factors represent object

valuation and the

purchase decision. _{Emotions are satisfaction/dissatisfaction (Oliver, 1980) and, as} proposed here, the endowment effect (Thaler, 1980) is assumed aconsequence of the

8 At a higherlevel thesetrial stages are part ofthe pre-purchasestage, becausetheyprecede

Product Trialsand ProductEvaluation 29

emotions and effects in the

trial

_stage.

The research reported here focuses on the second

(trial)

and

third

stage (post

trial).

The reason for not including the

first

stage is that it is relatively easy to get

people to try

a product, and this stage has previously received ample research. Rather,thefocus is on how the newproduct information affects attitudes, howthese

attitudes yield some _{final product evaluation and how this final judgment} affects

buying behaviour (willingness to pay). To summarise, the

focus is on how the

disclosure of missing _{information, through}

_trial,

affects subsequent product evaluation and purchase decisions. The next section

will

describe to process of

evaluation, especially howattitudescontribute to productevaluation.

Attitude

and

Attitude Formation

When evaluating objects (ideas, issues, events) people makejudgments about them.

The objects being evaluated are

attitude

objects and the evaluationsofthose objects areattitudes._{There are many}definitions of attitudes, butmost refer to anattitude as

some evaluative _{judgment or} feeling toward an _{object. For example, Petty and} Cacioppo (1986) define attitudes as"(...) general evaluations people hold

in

regard to

themselves, otherpeople, objects, and issues" (p. 129). FazioandZanna (1981)define

an attitude as "(...) the evaluative feeling that is evoked by a given object" (p. 162).

Thus, an attitude is an evaluativejudgmentabout an object and canbe expressed in threedifferentways: cognitive, affective, andbehavioural (foran overview see Eagly

& Chaiken,1,993).Cognitiveresponses arerepresentedby thoughtspeople haveabout

an attitude object ("Dancing is a nice way to spend my free time"). The affective category consists offeelings oremotionstowardtheattitudeobject ("Dancingmakes me feel_happy"). The_{behavioural response contains the overt actions with respect to} the attitude objects ("I often go to discotheques to dance"). Figure 2.2 depicts the

three classes

of

evaluative responses

of

attitudes. This model is called the

three-component modelof attitude.

In line with the idea

that attitudes can be _categorised into different classes of

responses, is the assumption that attitudes have different antecedents as shown in

Figure 2.3. These antecedents are cognitive, affective, and behavioural in nature.

Note that from

this perspective, attitudes are considered as a product

of

cognitive,

affective, and behavioural components (see Eagly & Chaiken, 1993). The cognitive

component

of

attitudesrefers tocognitive learningprocessesthatoccur when people

form

their

attitudes. _{These processes occur when people gain} information about an object to form beliefs. The information can be obtained by direct experience (e.g.,

30 Chapter 2

Cognitive Processes

Attitude_Objects Attitude Affective

Processes

Behavioural Processes

Observable Inferred Observable

Figure2.2.Three-component modelof attitude(Eagly & Chaiken).

Cognitive

Processes

Affective

_Attitude

Processes

Behavioural

Processes

Figure_2.3·Attitudesasconsequences

of

responses (Eagly&Chaiken, 1993).

they learn from it) to form their beliefs and attitudes. Second,

_{the affective} antecedent refers to the pairing ofan attitude object (conditioned

stimulus) with a

stimulus that elicitsa particularresponse (unconditioned stimulus). For example, an

advertisement for

soft

drink

(unconditioned stimulus) may

repeatedly be

accompaniedwithfeelings

of

freshness, energy, (conditioned stimulus) etc., such that

feelings

of

freshness become associated with the drink. Thus, affective antecedents

Product Trialsand ProductEvaluation 31

emotions.9 Last, behavioural components of attitudes are included in theories of cognitive dissonance (Festinger, 1957) and self-perception theory (Bem, 1972). For example, Bem's (1972) self-perception theory poses that people's attitudes are inferred from

their

own

behaviour ("I

often visit

_{discotheques, so I must like}

dancing"). Likewise, cognitive dissonance poses that people's need

for

dissonance

reduction results

in

attitudes that display the least discrepancy between

their

cognitions and

their

overtbehaviour. For example, peoplewhosmoke

light

cigarettes

will reduce dissonance_{by claiming that} the_{light cigarettesbeing advertised are less}

harmfulto

their

health than regularcigarettes.Anotherfamous model inthedomain of attitudes willbediscussed next.

TheElaboration Likelihood Modelof Persuasion

According to elaboration likelihood model

of

_persuasion (ELM) (Petty& Cacioppo, 1986b) attitudes are

formed by

the elaboration on the information (arguments) relevant to the objectbeing_{evaluated (issue-relevant}thinking). In the ELMthere are

two variables that determine whether people engage

in

issue-relevant

thinking

(elaboration): motivation and

ability

(Petty & Cacioppo, 1986b; Petty & Cacioppo,

19868).10

Motivation

refers to the extent towhichpeople aremotivatedto consciously

processtheinformation (ormessage). People'smotivation is influencedby variables

in a communication. For_{example, is the}attitude_objectofpersonal relevance? Is the person motivatedtolearn (see "needfor cognition",Cacioppo& Petty, 1982).

Abuity

refers to the extent to which people are able to process the

information. Both

personal and environmental factor may affect people's ability. For example, are

people distracted byexternal sources? Do they understandthe arguments? Do they

have_{topic-relevantknowledge?Elaboration likelihood is high}when peopleare_highly

motivated and able to engage

in

issue-relevant

thinking. If, on

the other hand, conditions

inhibit

people's motivation and

ability

elaboration is

likely to be low.

Depending on the degree

of

elaboration people either follow the

'central' or

'peripheral' route of persuasion to form their attitudes (e.g., Petty & Cacioppo, 19868).Central-route attitudesareformedwhenelaboration is high.Theinformation provided is _{carefully scrutinised} and information _processing is extensive (much

cognitive effort). Alternatively, peripheral-route attitudes are

formed when

elaboration is low.

The amount

of

_cognitive

effort that has

been

_{spent on the}

9 The ideathat_{cognitiveresponses,}like_{learning,change and}formattitudesisincorporated in the cognitiveresponsemodelofattitude formation(e.g.,_{Hovland, Janis, &}Kelly, 1953)

'° In addition to ELM, Batra & Ray (1986) discuss a thirdelement in relation to cognitive processing:opportunity. In their MAO model (motivation, ability, opportunity) they use the

32 Chapter 2

PersuasiveCommunication ,

Peripheral AttitudeShift

,C Attitude is relatively temporary.

; susceptible. and unpredictive of ; ; behaviour. ;

Motivated to Process? .

(personalrelevance: need for cognition: NO Yes I

personal responsibility: etc.)

\ * PeripheralCuePresent?

Yes Positive/negativeaffect; attractive

V /expertsources:number of

Ability toProcess? / arguments: etc.

Idistraction;repetition:priorknowledge: NO A

message comprehensibility) _No

Yes

V

Nature of CognitiveProcessing

i Retain or

Favourable Favourable Neither or

Regain Initial

thoughts thoughts neutral

Attitude

predominate predominate predominate

V V

CognitiveStructure_Change:

No

Arenew cognitions adopted and stored

in memory@ Aredifferentresponses

mode salientthanpreviouslyg

1

Yes 1 Yes

(favourobie) unfavourable) ' _Central i Central ; ; positive ; _{negative ;} 1 attitude ; attitude i ; change t change Z ; Attitudeisrelatively 1

endunng. resistant, and

' _predictiveofbehaviour.

Figure 2.4. The elaborationlikelihoodmodel

of

persuasion

(Petty &Cacioppo, 1986b).

information islow. People's attitudes are more orless formed on the basis ofsimple cues in the environment or on existing schemata. It is

further

argued that central-route attitudesare_{persistent, resistant, consistent and more predictive of}behaviour,

while peripheral-route attitudes are more orlesstemporary, superficial, susceptible, and less predictive of behaviour. For example, someone who has extensively processed the relevant information (arguments) andanalysed the pros and cons of

each piece

of information is

more

likely to form a

more persistent attitude than

someonewhotakesworth-to-mouth information asa_principalsourceof information.

Product TrialsandProductEvaluation 33

Implications of the Elaboration Likelihood Model for Product Trials

TheELMdescribesconditions under whichpeople

form

particulartypes

of

attitudes.

Depending on the elaboration

likelihood of

an information source they

will

adopt eithertemporary attitudes (peripheral route) or persistent attitudes (central route).

Especially

in

highly competitive environments,retailers wishto persuade consumers

of

the benefits

of their

_{products over all other competitor} brands, and persistent attitude _{changes through central-route persuasion is what}

retailers may wish to

reach.

Let us consider the two

factors in the

ELM, motivation and ability, which

determine the amount ofelaboration.

In

order for a _product

trial to be

an effective tactic, consumers must be motivated to engage with the product

trial. When

consumers do not expect to learn anything new

from trial, they are

less

likely to

appreciate the

trial and

this results

in

low-order beliefs. Once a product has been

selected as appropriate

for trial,

product sampling can increase motivation. The

second

factor in the

ELM, ability, asserts that consumer's

prior

knowledge,

environmental factors, or clarity of

information may act on

the

amount of

elaboration. It will not make sense to ask anon- mechanic to assessthequality of a

car engine. Likewise, the true advantagesof walking shoes cannotbe assessed when they areworn during acycle tripor duringa_{5-minute walk to the railway station. In}

short, when the credibility and representativeness of a _{product's true performance}

cannotbe assessed_{during the}

trial

period it is said that the

trial

validity is low. When the credibilityand representativeness of true performance canbejudged inthe given

trial

period the

trial validity is high

(e.g., Kempf & Smith, 1998). Related to

trial

validity is the trial diagnosticity and is defined as the degree to which consumers believethe

trial

isusefulin evaluatingaproduct's attributes (Kempf &Smith, 1998). The degreeof diagnosticityisinfluenced bywhetheraproduct'ssalientattributes can

be evaluated by

trial or not.

A

trial

is highly diagnostic when, through

_{trial, the}

product's attributes can directlybe evaluatedwithout ambiguity in the information, or

without

problems due to lack

of

expertise. On theother hand, a

trial

ishighly

non-diagnostic when a product's salient attributes cannotbe evaluated bytrial, or when

expertise is missing. An example of a highly diagnostic

trial

is eating an entire

chocolate bartoassessitssweetness.

34 Chapter 2

Attribute Information Typesand Related_{Consumption Behaviour}

Product attributes can roughly be classified in three types: search attributes,

experience _{attributes, and credence attributes} (e.g., _{Nelson, 1970;} Wright & Lynch,

1995; Kempf, _{& Smith, 1998). Search} attributes can be evaluated

without

direct experience, and can be obtained _{indirectly (e.g., through advertisement claims,}

product reviews). Examples are form, shape, and colour. Experience attributes can

onlybe evaluatedthrough directexperience. Examples aretaste, userfriendliness of appliances, speed, or, more generally, true performance of a good. Products that

consist

of

mainly search attributes are labelled search goods and products that

consist

of

_{mainly experience}attributes are labelled experience goods. In addition to

these goodsthere is also the type of good

of

whichanattribute cannotbe evaluated

but has tobebelieved.These goods are_{called credence goods. Examples}

of

credence goods are green electricity, free-range eggs, and biological products.

In

short, there aresearch_{goods,experiencegoods,} and credence_goods.

Searchand experiencegoods fit more or less into two perspectives

of

consumption

behaviour: theinformation processing view and the experiential view _(see Holbrook & Hirschman, 1982). According to the _{information processing view the}consumer's task is to evaluate andjudge through'secondary process' thinking. Thesesecondary processes include search

for

information, retrieval from memory and _{weighting of} evidence._{Alternatively,} the _{experiential view emphasises 'primaryprocess'thinking} andhedonic responses. This typeof consumptionis pleasureoriented andresults in pleasure, fun, and _{gratification.}

_{Attention is paid to}

_{the hedonic components of}

consumption (e.g., a

visit to

a theme park).11 _{In product trials, consumers are}

searching

for

information about search _attributes, so product

trials fit into the

information processing view.12

Attitude Formation

in

Product Trials

The purpose of a product

trial is

for

consumers to evaluate products, and assess

product quality while interacting with them. Productinformation isobtained_directly and continuously. Because of the interactive learning process (new

information is

encountered and integrated

into

existing knowledge), attitudes toward the product (or productfeatures) are_{likely to} be _{'updated' instantly. For example, once in a PC}

11

It should bementioned thatthere is no one-to-one relationshipbetween the information processing view and search _{attributes, and} the experiential viewand experience attributes. For example, one maysearch_{for information (information}processing view) about the taste of

a newsnack_{(experienceattribute).}

12