A Framework forConceptualizing and Measuring the Involvement Constructin Advertising Research

Therefore, the purpose of this paper is to propose a framework that closely scrutinizes the involvement construct's antecedents, state properties, measures, related constructs, and conse

Trang 1

A Framework for

Conceptualizing and Measuring the

Involvement Construct

in Advertising Research

Srinivas Durvasula

Syed H Akhter

J.Craig Andrews (Ph.D., University of South

Carolina) is assistant professor of marketing,

Marquette University.

Srinivas Durvasula (Ph.D., University of

South Carolina) is assistant professor of

marketing, Marquette University.

Syed H Akhter (Ph.D., University of

Oklahoma) is assistant professor of marketing,

Marquette University.

The authors gratefully acknowledge helpful

comments and suggestions by Terence A.

Shimp on an earlier version of the manuscript.

Adequately conceptualizing and measuring the involvement construct has been one of the most controoersial topics in advertising research This task is especially important given the many involvement-driven frameworks (e.g., the elaboration likelihood model) now being advanced in advertising The present study proposes a framework that closely scrutinizes the inoolvement constructs antecedents, state properties, measures, potential confounds, and consequences Four emerging research streams in involvement are presented in the context of the framework Im-plications for researchers attempting to manipulate and measure inoolvement in experimental advertising research areprovided

The conceptualization and measurement of involvement has long been an im-portant and controversial topic for advertising researchers It is especially

im-portant because of the recent advancement and testing of many involvement-driven models in advertising, such as the elaboration likelihood model (Petty and Cacioppo 1981a; 1983; 1986), the attitude-toward-the-ad model (Lutz 1985; Mitch-ell, and Olson 1981; Shimp 1981), and the integrated information response model (Smith and Swinyard 1982; 1983) However, involvement is also controversial due

to the many different proposals and ideas for conceptualizing (cf., Andrews 19BB;

Antil 1984; Cohen 1983; Gardner, Mitchell, and Russo 1978, 19B5; Greenwald and Leavitt 1984; Houston and Rothschild 1978; Krugman 1966-1967; Lastovicka and Gardner 1979; Mitchell 1979, 1981; Park and Young 1986; Wright 1973, 1974; Zaichkowsky 1986) and measuring the involvement construct (cf., Bloch 1981; Celsi and Olson 19BB; Laczniak, Muehling, and Grossbart 1989; Laurent and Kapferer 1985; Wells 19B6; Zaichkowsky 1985) There have been noteworthy attempts to reconcile these differences by separately examining: involvement models (e.g., Mitchell19B1;Zaichkowsky 19B6),involvement state properties (e.g., Cohen 1983; Mitchell 1981), empirical examples of involvement manipulations (e.g., La-czniak, Muehling, and Grossbart 1989), and illustrations of the operationalization

of advertising involvement (e.g., Andrews 1988) However, what is really needed

is a comprehensive, detailed, and operational framework that helps integrate these separate involvement issues by clearly examining the underlying properties of the involvement construct, and that separates the construct from its many antecedents, consequences, and potential confounding or related constructs A clarification of this nature would be of value to advertising researchers seeking to conceptualize, manipulate, and measure involvement in their own research

Therefore, the purpose of this paper is to propose a framework that closely scrutinizes the involvement construct's antecedents, state properties, measures, related constructs, and consequences The framework's relationship to four emerg-ing streams of research in involvement is then presented to help reconcile the many proposals and ideas for conceptualizing and measuring involvement Im-plications for those attempting to manipulate and/or measure involvement in

©}oumal of Advertising

Trang 2

experimental advertising research are

also provided

The Framework

A necessary condition for adequately

measuring a construct is to first

pre-cisely specify the domain of the

con-struct As suggested by Churchill

(1979), "The researcher must be

exact-ing in delineatexact-ing what is included in

the definition and what is excluded" (p

67) Unfortunately, in the case of

in-volvement, many researchers have not

defined the domain of the construct,

nor provided manipulation checks in

experimentation with involvement

The position taken in this paper is

that involvement is an individual,

in-ternal state of arousal with intensity,

direction, and persistence properties

(Note: For those interested in the

con-ceptualization of involvement as a

pro-cess, see Greenwald and Leavitt 1984.)

The focus of involvement is on the

dividual consumer That is, it is the

in-dividual consumer who is involved, not

products, or advertising content,

me-dia, objects, or situations The

consum-er's internal state of arousal determines

how he/she responds to stimuli, such

as advertisements or products (Note:

While some authors[ef.,Mitchell 1981]

have defined involvement in the

con-text of motivational properties evoked

by a particular stimulus or situation,

our definition focuses on individual

arousal levels with respect to stimuli.)

This internal state of arousal,

acti-vation, or preparedness can be

sepa-rated from its numerous antecedents

and consequences as depicted in our

framework in Figure 1 The

frame-work provides a nomological netframe-work

of relationships among involvement

antecedents, state properties, related

constructs, and consequences (A

"nomological network" refers to the

predicted pattern of theoretical

rela-tionships that helps define a construct;

Cook and Campbell 1979, p 70.) It is

apparent that the primary problem in

defining the domain of the involvement

construct is that numerous

anteced-ents and consequences of involvement

have been confused with involvement

per se For example, it would be a pre-carious assumption to treat cognitive-response activity as an indicant (v con-sequence) of the state of involvement

The reasoning is that cognitive-re-sponse generation can be the result of

a multitude of factors beyond one's

lev-el of involvement, including one's ac-cessibility of thoughts, response oppor-tunity, general verbosity, and person-ality traits (Wright 1980) Furthermore,

as suggested by Cohen (1983), if cog-nitive responses were synonymous with the involvement construct, there would

be no need for such a construct Care should also be taken not to define in-volvement as risk or familiarity (ante-cedents/related constructs), or to label involvement as "cognitive involve-ment," "emotional involveinvolve-ment," or

"behavioral involvement" (conse-quences) In summary, it would be in-appropriate to conclude that the state

of involvement can be inferred by the

virtue of measuring involvement an-tecedents or consequences Rather, measures attempting to tap the state (v

antecedents or consequences) of in-volvement are preferable For example,

in an experiment manipulating in-volvement in a particular advertise-ment, manipulation checks can be included to measure the intensity or degree to which an individual felt he/

she was engaged in, engrossed in, ab-sorbed by, paid attention to, concen-trated on, carefully examined, thought about, focused on, or was involved in the particular advertisement in ques-tion

A closer examination of Figure 1 re-veals numerous antecedents to involve-ment grouped into personal and situational/decision factors For exam-ple, studies by Petty and Cacioppo (1986) have manipulated the personal relevance of advertised products and measured differing levels of a person's need for cognition These factors can serve to drive one's involvement to-ward an advertised product or issue at hand Furthermore, factors such as one'sopportunity toprocess (e.g., due to

distraction, media type, etc.) andability

to process (e.g., due to product

knowl-edge, familiarity, etc.) can serve to limit

or constrain the impact of these in-volvement antecedents on the involve-ment state (cf., Andrews 1988; Batra and Ray 1986) Numerous conse-quences of manipulated involvement

levels have also been determined, in-cluding a greater number of total and directed cognitive responses (Park and Young 1986; Petty and Cacioppo 1979), more elaborate and complex encoding strategies (Celsi and Olson 1988), sig-nificant effects on recall and recogni-tion measures (Leigh and Menon 1987), and persuasion that is more enduring, predictive of behavior, and resistant to counter-propaganda (Petty and Ca-cioppo 1986)

The position taken in this paper is that

involvement is an individual, internal state of arousal with intensity, direction, and persistence properties.

As an internal state of arousal, in-volvement has three major properties: intensity, direction, and persistence(cf., Bettman's 1979 motivational mecha-nisms; Shimp 1982) While these in-volvement properties are grounded in motivation, we feel that motivation represents a broader construct than in-volvement As such, motivation refers

to forces/drives ("motivational prop-erties"-Mitchell 1981; "motivational aspects"-Cohen 1983) that move one from an initial state to a desired state (Bettman 1979) These forces/drives help direct arousal or activation levels toward stimuli, and play an important role in facilitating involvement-related outcomes or consequences (e.g., mes-sage-oriented thoughts; message-argu-ment recall; Petty, Cacioppo, and Schumann 1983; Wright 1973)

Trang 3

FIGURE 1

A Framework for the Conceptualization and Measurement of the Involvement Construct

a Personal needs, goals and char- a Search behavior

shop-• personal goals, values, and con- Related Constructs: ping behavior

sequences • opportunity to process (e.g., dis- • increased complexity of

de-• cultural values j- traction, media type) cision process

• degree to which the object has

I • ability to process (e.g., knowledge, • greater time spent

examin-ego-related significance

I famUiarity) ing alternatives

• personal relevance of the object I • greater perceived product

• importance of the object J attribute differences

• personality factors (e.g., need for I b Information PTocessing

cognition) I • increased total and directed

b Situational and decision factors J cognitive-response activity

• purchase occasion I Involvement: • greater number of personal

• perceived risk of decision • intensity • more elaborate encoding

• direction

• size of decision consequences

• persistence strategies

• imminence of decision • increased recall and

compre-• degree of decision irrevocabUity hension

• degree of personal responsibil- c. Persuasion

ity associated with decision • if present with cogent

ar-guments, greater central (v peripheral) attitude change

• attitude change that is more enduring, predictive of be-havior, and resistant to counter persuasion

NOTE: Solid arrows represent direct influences, while the double-ruled arrow represents a mediating influence

By involvement intensity, we mean

the degree of arousal or the

prepared-ness of the involved consumer with

re-spect to the goal-related object That

is, the involved consumer has a certain

degree of arousal to engage in specific

information-processing or goal-related

behaviors However, intensity refers to

the arousal level and not the actual

pro-cessing or behaviors that remain as

consequences of involvement This

in-tensity level should also be thought of

as operating on a continuum, as

op-posed to "high" or "low" levels of

in-volvement This is an important

distinction since "moderate" levels of

involvement have all but been ignored

in experimental research A notable

exception, however, is based on a stream

of research on source effects in

per-suasion (Heesacker, Petty, and

Caciop-po 1983; Puckett, Petty, CaciopCaciop-po, and Fisher 1983).As suggested by Petty and Cacioppo (1986), source (and periph-eral) factors can influence the extent of message processing when the personal relevance of a message is moderateor ambiguous

The direction of involvement refers

to the target of the involvement inten-sity level That is, direction refers to the stimulus (e.g., issue, product, ad-vertisement) toward which the arousal

is channeled (cf., Mitchell 1981) This definition differs from the previous use

of direction to refer to the amount of cognitive and behavioral activity

relat-ed to the stimulus or goal-object (ef.,

Bettman 1979; Gardner, Mitchell, and Russo 1978; Park and Young 1986) The rationale is that cognitive and behav-ioral activity remains a consequence (as

opposed to an indicator) of the state of involvement The antecedent condi-tions of involvement (e.g., purchase oc-casions, personal relevance, risk, etc.) will first help determine the consum-er's involvement direction or goal-ob-ject selected For example, in an experiment heightening subjects' per-ceived relevance of an ad's message content, the direction of their involve-ment should be toward the content of the ad and not toward background cues found in the ad This can be checked

by a measure of their relative concen-tration on ad content versus back-ground cues (e.g., endorsers, music, colors) in the ad (Wright 1973, p 57) Involvementpersistencerefers to the duration of the involvement intensity Depending upon the different cate-gories of involvement, there will no

Trang 4

doubt be varying persistence levels For

example, long-time car enthusiasts,

health "nuts:' wine connoisseurs, or ski

"bums" would be expected to be

high-er in involvement phigh-ersistence than

those under situational involvement

states (cf., Bloch 1981; Celsi and Olson

1988) The recognition ofenduring

in-volvement is important because the

fo-cus of work on involvement

conceptu-alizations may be limited to

involve-ment definitions that are situational or

transitory in nature While the

in-volvement levels in the preceding

ex-amples endure over time, the level of

persistence forsituationalinvolvement

is expected to decline when the

ulti-mate goal is achieved, or when the

sit-uation changes (Celsi and Olson 1988)

With situational involvement, this level

of persistence is also expected to be

greater for high- versus

low-involve-ment subjects In particular, persuasive

effects (e.g., attitude changes,

attitude-behavioral intention correlations) are

predicted to be more persistent under

high versus low involvement (Petty and

Cacioppo 1986, pp, 175-178) For

ex-ample, brand beliefs were found to

in-fluence brand attitudes for

high-involvement subjects on both

imme-diate and delayed measures (Muehling

and Lacmiak 1988) Under low

in-volvement, however, these effects did

not occur Given these properties,

an-tecedents, and consequences of

in-volvement, we now turn our attention

to the relationship of our framework

to previous involvement

conceptuali-zations and measures grouped on the

basis of four emerging streams of

re-search in involvement

Conceptual and

Measurement Issues: Four

Research Streams

Conceptual Issues Given the

impor-tance of involvement as an integral

moderating condition in advertising

re-search, it is somewhat disconcerting to

see numerous, conflicting conceptual

and operational definitions of the

in-volvement construct However, this

frustrating situation is not as hopeless

as it first appears, in that some under-lying commonalities can be observed

For example, we have identified four major groupings of involvement con-ceptualizations and have labeled these as: (1) attention/processing strategies, (2) personal/situational involvement, (3) audience/process involvement, and (4) enduring/product involvement These four research streams are directly re-lated to our framework in that the at-tention/processing strategies field has contributed to the examination of the state of involvement (e.g., involvement intensity and direction), the personal/

situational and enduring/product in-volvement streams have contributed to the manipulation and/or measurement

of involvement antecedents, while the audience/process involvement stream has advanced our knowledge of the consequences of involvement (e.g., cog-nitive responses, message-argument re-call) that require varying levels of attentional capacity

Table 1 provides a chronological sample of conceptual definitions of in-volvement categorized and grouped on the basis of the four research streams

The rationale for the involvement def-inition groupings is as follows Studies included under attention/processing strategies have proposed involvement conceptualizations accounting for both the level of attention and the direction

of processing (e.g., brandv.non-brand processing; cf., Gardner, Mitchell, and Russo 1978) That is, attention and processing are viewed as critical stages

in information acquisition, influenced

by the level and direction of involve-ment (Mitchell 1981) For example, un-der high involvement, individuals are expected to devote all attention to the advertisement and execute a brand-processing strategy Under low involve-ment, individuals either do not allocate attention to the advertisement or they invoke a non-brand-processing

strate-gy Research definitions examining Mitchell's (1979; 1981) extension of this field, expressing involvement as astate

(v process) construct, are also included

in this category (e.g., Andrews 1988;

Cohen 1983; Lacmiak, Muehling, and Grossbart 1989) An important

dis-tinction should be made, however, be-tween attention/processing strategies and audience/process involvement (e.g., Greenwald and Leavitt 1984; Krugman 1966-1967) As Mitchell (1981, p 27) indicates:

Krugman defines involvementasone

of the dimensions of the' type of pro-cessing that occurs during exposure

to the advertisement In contrast, I define involvement as a particular state of the individual at a point in time Our state variable concep-tualization of involvementaffectsthe type of processing that occurs dur-ing exposure

Similarly, Greenwald and Leavitt (1984,

p 590) indicate that: "Mitchell's (1979) interpretation of involvement in terms

of varying arousal appears to fit least well with our analysis." This is because Greenwald and Leavitt (1984) view in-volvementasa process represented by four process-involvement levels (i.e.,

preattention, focal attention, compre-hension, and elaboration), each requir-ing greater attentional capacity In contrast, Mitchell's (1979; 1981) in-volvement state serves to influence(as opposed to represent) these process stages or levels

The second research stream, person-al/situational involvement, represents a collection of involvement definitions based on the idea that issues, situations,

or messages can have significant conse-quences on,orbe personally relevant to, one's own life (Apsler and Sears 1968) Recent research regarding the person-al/situational-involvement field is exten-sive (cf., Petty and Cacioppo 1981b; 1986; Petty, Cacioppo, and Schumann 1983) and makes use of personal rele-vance (and other) manipulations to in-duce varying levels of involvement and subsequent persuasion While the at-tention/processing-strategies field has contributed to our understanding of in-volvement-state properties, personal/ situational-involvement research has ad-vanced our knowledge regarding the role of antecedents (e.g., personal rele-vance, need for cognition, personal re-sponsibility, etc.) in involvement-induced persuasion

Trang 5

Study Focus

TABLE 1 Conceptual Definitions of Involvement

Definition

A AttentionfProcessing Strategies

Gardner, Mitchell, and advertising

Russo (1978)

Laczniak, Muehling, and advertising

Grossbart (1989)

B.Personal/Situational Involvement

Sherif and Cantril (1947) object

C Audience/Process Involvement

Krugman (1966-1967) advertising

Mitchell (1979, p 194;

1981, p 25)

Cohen (1983)

Andrews (1988)

Zimbardo (1960)

Apsler and Sears (1968)

Wright (1973)

Houston and Rothschild

(1978)

Petty and Cacioppo

(1981b)

Burnkrant and Sawyer

(1983)

Celsi and Olson (1988)

(1978)

stimulus

message

response

issue

advertising

situational

message, issue

message

situation

response

1.high-inllolliementlearning (brand set): the interest in the product category

is high and the consumer is actively processing the information in the advertisement to reach an overall evaluationofthe advertised brand (p 584)

2 low-inllOlliement learning (nonbrand set):

a)strategy-limited: the person processes the advertisement with other than

a (brand) evaluation strategy A trace ofthe advertisement is stored in episodic memory (p 584)

b) attention-limited:the advertisement does not receive enough attention for it tobefully perceived or evaluated (p 584)

Anindividual level, internalstatevariable that indicates the amountofarousal, interest, or drive evoked by a particular stimulus or situation Involvement, therefore, has two dimensions, intensity and direction (p 194)

A state of activation directed toward some portion of the person's psycho-logical field (p 326)

An individual, internal state of arousal with intensity and direction prop-erties (p 24)

The motivational state of an individual induced by a particular stimulus or situation (p 30)

ego-inllOlliement: when any stimulus or situation is consciously or uncon-sciously related to [the contents of the ego] by the individual (p 117)

responseinllOlliement:the individual's concern with the consequences of his/ her response or with the instrumental meaning of his/her opinion (p 87)

personal inllolliement: the expectation of an issue to have significant conse-quences for an individual's own life (p 162)

content-processinginllolliement:when a person is confronted by an advertise-ment which he/she perceives as particularly relevant to an impending de-cision (p 55)

situational inllOlliement: the ability of a situation to elicit from individuals concern for their behavior in that situation (p 184)

In high-inllOlliement situations, the persuasive message under consideration has a high degree of personal relevance to the recipient, whereas in

low-inllolliementsituations, the personal relevance of the message is rather trivial (p.20)

A motivational state based on the message receiver's need for information and the meaningfulness of the message content (pp 57-58)

situational sources of personal releeance (SSPR): a wide variety of specific

stimuli, cues, and contingencies in a consumer's immediate environment that activate or are closely associated with self-relevant consequences, goals and values (pp 211-212)

adlieTtising inllOlliement: the number of "connections," conscious bridging experiences, or personal references per minute, that the subject makes be-tween the content of the persuasive stimulus and the content of his/her own life (p 584)

response inllOlliement: the complexity or extensiveness of cognitive and

be-havioral processes characterizing the overall consumer decision process.Itis the result of situational and enduring involvement (p 185)

Table continued

Trang 6

Study Focus

TABLE 1 (Continued) Conceptual Definitions of Involvement

Definition

D Enduring/Product Involvement

Day (1970) object

Leavitt, Greenwald, and

Obermiller (1981)

Batra and Ray (1983)

Greenwald and Leavitt

(1984)

Park and Young (1986)

Baker and Lutz (1987)

Leigh and Menon (1987)

Bowen and Chaffee (1974)

(1978)

Lastovicka and Gardner

(1979)

Bloch (1981)

Zaichkowsky (1985)

processing

message

audience

information processing

advertising

audience

product class

individual and situation

product class

object (i.e., product class, advertise-ments, purchase decisions) individual knowledge and experience

1 highinlloillement: The interpretations of high involvement as cognitive responding or as establishing personal connections are grouped together

as variations on the theme of encoding elaboration (p 17)

2 low inllOlliement: "we are inclined to identify low involvement with the achievement of focal attention, accompanied by minimal encoding"

(p. 17)

message-response inllOlliement: a situational state measured or calibrated by the depth and quality of message-evoked cognitive responses (p 309)

audienceinllOlliement:the allocation of attentional capacity to a message source,

as needed to analyze the message at one of a series of increasingly abstract representational levels (p 591)

1 cognitille inllOlliement:the degree of personal relevanceofmessage contents

or issue based on the brand's functional performance (utilitarian motive) (p.12)

2 ajfectille inllOlliement:the degree of personal relevance of a message based

on emotional or aesthetic appeals to one's motive to express an actual or ideal self-image to the outside world (value-expressive motive) (p 12)

3 low inllOlliement:a lack of focused mental processing in evaluating a mes-sage (p 14)

1 adllertising-messageinllOlliement: a motivational construct embodying the amount of cognitive effort directed by the consumer at processing the contents of an advertising message (p 75)

2 adllertising-execution inllOlliement:a motivational construct embodying the degree of cognitive effort directed by the consumer at processing the executional properties of an advertising stimulus without regard to their brand relatedness (p 80)

audienceinlloillement: "the level of attention and depth of processing(i.e.,

focus on sensory versus semantic features) should serve to define the level

of audience involvement." (pA)

The general level of interest in the object or the centrality of the object to the person's own ego-structure (p 45)

productinlloillement: a direct outgrowth of the potential benefits or rewards the product holds for the consumer (p 615)

enduringinlloillement: reflects the strength of the pre-existing relationship between an individual and the situation in which the behavior will occur

Involvement is best thought of in terms of two underlying components: a) norrnzmve importance:how connected or engaged a product class is to an individual's values (p 68)

b) commitment: the pledging of binding of an individual to his/her brand choice (p 68)

productinllOlliement: a construct which affects consumer behavior on an ongoing basis (p 62)

A person's perceived relevance of the object based on inherent needs, values, and interests (p 342)

intrinsicsources ofpersonalrelevsnce (ISPR): are relatively stable, enduring structures of personally relevant knowledge, derived from past experience and stored in memory (p 212)

Trang 7

A primary element in theaudience/

process-involvement stream of research

is the notion of audience involvement

As defined by Greenwald and Leavitt

(1984, p 591), " audience

involve-ment is the allocation of attentional

ca-pacity to a message source, as needed

to analyze the message at one of a series

of increasingly abstract representation

levels." These increasingly abstract and

complex representations are viewed as

levels of involvement in this research

stream The idea is that

low-involve-ment levels (e.g., preattention, focal

at-tention) use little capacity, while higher

levels of involvement (e.g.,

comprehen-sion, elaboration) require greater

ca-pacity Therefore, research proposing

definitions of involvement as a

three-stage processing sequence (Leavitt,

Greenwald, and Obermiller 1981), or

as a four-level process (Greenwald and

Leavitt 1984), is included in the

audi-ence/process-involvement research

stream Studies defining involvement

in processing terminology, such as

cog-nitive responses, connections,

cogni-tive effort, or cognicogni-tive and affeccogni-tive

needs (e.g., Baker and Lutz 1987; Batra

and Ray 1983; Krugman 1966-1967;

and Park and Young 1986) have also

been included in this research stream

"Connections" and other cognitive

ac-tivity generated from advertising differ

from the "personal relevance of

adver-tising information" in that personal

relevance (e.g., due to a planned

pur-chase) can lead to a higher state of

in-volvement, and subsequently, to a

greater number of "connections" and

increased cognitive-response activity

As indicated by Krugman (1966-1967,

pp 587-589), "connections" represent

the number of actual

thoughts/refer-ences per minute that relate advertising

information to a person's own life

These "connections" (as an

involve-ment consequence) should be more

prevalent when the information in the

ad is personally relevant (an

involve-ment antecedent)

Finally, the product-involvementfield

(e.g., Bloch 1981; Bowen and Chaffee

1974; Day 1970; Zaichkowsky 1985) has

been integrated with enduring

involve-ment(Celsi and Olson 1988; Houston

and Rothschild 1978) because the pre-existing experience and knowledge-structure terminology found in endur-ing-involvement definitions also plays

an important role in product-involve-ment conceptualizaitons (ef., Bloch 1981) In this research stream, individ-uals are viewed as having relatively sta-ble and enduring involvement levels with a particular stimulus (e.g., a prod-uct class) This notion differs from

oth-er definitions of involvement (e.g., attention/processing strategies; per-sonal/situational involvement) that view involvement assituationally-spectficor transitory in nature

the primary

problem in defining the domain of the involvement construct

is that numerous antecedents and consequences of involvement have been confused with involvement per se.

Table 2 provides a summary exami-nation and comparison of each involve-ment research stream on the basis of its origin, evolution, theoretical basis, pri-mary focus, state versus process, dimen-sionality, and primary contribution and relationships to our proposed frame-work presented in Figure1

Interesting-ly, several commonalities emerge from the comparisons, especially in the

ex-amination of the theoretical bases for the streams For example, Kahneman's (1973) theory of attention (with his arousal and capacity dimensions), and Greenwald's (1968) cognitive-response theory play important roles in both at-tention/processing strategies and audi-ence/process-involvement research streams, even though the streams differ

with respect to their state v process ori-entation Also, the enduring/product-involvement field has relied upon the other three streams for its theoreticai development (e.g., its major theoretical work includes Greenwald and Leavitt 1984; Mitchell 1979; Petty and

Caciop-po 1981a; and Sherif and Cantril 1947; ef.,Celsi and Olson 1988; Zaichkowsky 1985) However, most commonalities end here, leaving clear differences across the streams on their primary focus, con-tribution and relationship to our frame-work (see Table 2)

Measurement Issues Perhaps more troublesome is the review of involve-ment measures that have been devel-oped and applied in advertising and consumer research Table 3 provides a sample of these measures, incorporat-ing two perspectives: measurement scales and manipulation checks These measures have also been organized around the four streams of research in involvement The enduring/product-involvement stream accounts for most involvement measurement scales listed, while contributions from the other three streams are found in the sample

of involvement manipulation checks

Measurement scales. An examina-tion of measurement scales in Table 3 reveals that the product-involvement-scale approaches are quite diverse, ranging from ingenious methods to tap product interest (Buchanan 1964), to the use of "proxy" measures to infer product involvement (Bowen and Chaf-fee 1974; Tyebjee 1979), to scales em-ploying rigorous procedures suggested

by Churchill (1979) in developing, pu-rifying, and validating scales (cf., Bloch 1981; Zaichkowsky 1985) The latter validation attempts are preferred in or-der to have better confidence in the degree to which our measures assess the involvement construct However, the validation of multi-item scales of product involvement is not without its problems For example, the construc-tion of general measures of product in-volvement (e.g., Zaichkowsky's 1985 PIl) aid in the generalizability of such scales across product categories

Trang 8

TABLE 2

A Comparison of Four Research Streams In Involvement

ResearchStreams

Distinguishing Attention/Processing Personal/Situational Audience/Process

Origin: Gardner, Mitchell, Sherif and Cantril Krugman

Greenwald, and Ob-ermiller (1981)-3 processing sequences Evolution: Mitchell (1979; 1981); Apsler and Sears Greenwald and

Leav-Gardner, Mitchell, (1968); Petty and Ca- itt (1984; Bloch and Russo (1985);La- cioppo (1981a; 1986) 1985)-4 levels of

Grossbart (1989) Theoretical HI:Wright (1973) Social Judgment The- Attention:

Kahne-Basis: Greenwald (1968) ory; Elaboration Likeli- man (1973)

Theory: Greenwald (1968)

Info Processing Stages: McGuire 1969) Primary stimuli within issues, situations, adver- processing stages,

Process:

Dimensionality: Bi-dimensional: Level Unidimensional:indi- Unidimensional:

in-of attention, direction vidual's level of person- creasing capacity

message- or Issue-relat- increasingly complex

ed information representations

Primary con- involvement state involvement anteced- involvement

conse-tribution and ents (e.g., personal rele- quences requiring

Enduring/Product

InllOlvement Product: Day (1970) Enduring: Houston and Rothschild (1978)

Product: Bloch (1981); Zaichkowsky (1985) Enduring: Celsi and Olson (1988) Product: Greenwald and Leavitt (1984; Mitchell (1979); Petty and Cacioppo (1981a)

Enduring: Sherif and Cantril (1947)

product-class rele-vance, the individual state measures (e.g., Zaichkowsky's 1985 PII)

Primarily a unidimen-sional intent (Bloch 1981; Zaichkowsky 1985), however some are bi-dimensional (Lastovicka and Gard-ner 1979)

involvement anteced-ents

(v product-specific scales, e.g., Bloch

1981) Recently, however, questions

have arisen regarding the

unidimen-sionality of such general

product-in-volvement measures (cf., McQuarrie

and Munson 1987; Mittal 1989;

Zaich-kowsky 1987) This is due, in part, to

the problem of separating affect from

product involvement in such measures

Định dạng
Số trang	14
Dung lượng	1,96 MB