Understanding Information Systems Continuance

From Driscollwiki

Jump to: navigation, search


Anol Bhattacherjee. (2001) Understanding Information Systems Continuance: An Expectation-Confirmation Model. MIS Quarterly, Vol. 25, No. 3, Sep., pp. 351-370

Contents

Previous research

Previous work looks only at first time use of information systems (IS).

Studies of the variables that motivate individuals to accept a new IS:

  • Innovation diffusion theory, Rogers, 1995
  • Technology Acceptance Model (TAM), Davis et al, 1989
  • The Theory of Planned Behavior, Ajzen, 1991

Continuance concerns long-term viability, eventual success

"IS continuance at the individual user level is also central to the survival of many business-to-consumer electronic commerce firms, such as Internet service providers (ISPs), online retailers, [etc.]" (352)

Previous continuance-like studies:

  • Implementation, Zmud, 1982
  • Incorporation, Kwon and Zmud, 1987
  • Routinzation, Cooper and Zmud, 1990

Continuation in Innovation Diffusion Theory

"Innovation diffusion theory, in its five-stage adoption decision process [...] suggests that adopters reevaluate their earlier acceptance decision during a final "confirmation" stage and decide whether to continue or discontinue using an innovation (Rogers 1995)." (352)
  • Assumes continuance covaries with acceptance
  • Can't explain some users' discontinuation after initial acceptance (352)
    • "acceptance-discontinuance anomaly"
  • Model doesn't account for changing psychological motivates after acceptance
"[Current] acceptnace models provide a limited explanation of, and may sometimes contradict, observed continuance behaviors." (352)

Research Questions

  1. What are the salient motivations underlying IS users' intention to continue using an IS after its initial acceptance?
  2. How do these motivations influence continuance intention?

Expectation-confirmation Theory (ECT)

  • Oliver, 1980
  • Used in Consumer behavior literature to study consumer satisfaction, post-purchase behavior, and service marketing in general (353)

Summary

  1. Consumer forms expectation of a specific product, service
  2. Accept and use the product
  3. Following period of use, form a perception about its performance
  4. Perceived performance v expectation yield confirmation of initial expectation
  5. Satisfaction is determined based on their confirmation and expectation
  6. Finally, form a repurchase intention or discontinue use

Satisfaction

"The summary psychological state resulting when the emotion surrounding disconfirmed expectations is coupled with the consumer's prior feelings about the consumption experience (Oliver 1981, p 29)." (353-4)
  • Repurchase intent determined by satisfaction with prior use of product or service
  • Lower expectation/ higher performance leads to greater confirmation, positive satisfaction, continuance intention

Expectation

  • High expectation tends to enhance one's satisfaction
  • Low expectation reduces consequent satisfaction (354)

Critique

Ignores changes in expectation following consumption experience

  • Pre-consumption expectation is based on mass media, social net information
  • Post-consumption is "more realistic" (354) and might include discovery of new uses, services, benefits (+ vice-versa)

Initial studies presented varying, conflicting conceptualizations of satisfaction (Yi 1990) (354)

  • Some saw satisfaction as synonymouns with attitude and emotion
  • Hunt (1977) argues that attitude is an emotion; one many have pleasureable experience but still feel dissatisfied if it does not meet expectation

Conceptualization of expectation also differs across studies

  • Sometimes pre-consumption belief, anticipated performance
  • Beliefs about the levelof product, individual beliefs

Adapting Expectation-Confirmation Model to IS Continuance

Continuance is similar to repurchase because

  • Follow an initial decision (acceptance)
  • Influenced by initial use experience
  • Can potentially lead to ex post reversal of initial decision

Rational users go through a non-trivial decision process

  • Because of monetary, non-monetary costs of continuance

Focus only on post-acceptance variables

  • Pre-acceptance variables are captured within the confirmation and satisfaction constructs (355)

Only examine effect of pre-consumption expectation

  • Especially important with phenomena in which expectations change with time
    • Characteristic of IS use (355)

Ex post Expectation represented by ex post Perceived usefulness

Affirmed by the TAM literature:

  • Perceived usefulness is a cognitive belief salient to IS use (Davis et al 1989)
  • Only belief to consistently influence user intention across temporaly stages of IS use (Davis et al 1989, Karahanna et al 1999)

Satisfaction is the primary determinant of IS continuance

  • Satisfaction is an affect
  • Theorized, validated in TAM-based studies as important predictor of intention concerning IS use (Davis et al 1989, Karahanna et all 1999, Taylor and Todd 1995)

Two constructs determine satisfaction:

  • Expectation of the IS
  • Confirmation of expectation following actual use

Technology Acceptance Model (TAM) Constructs

Perceived ease of use

Empirical studies comparing relative effects during pre- and post-acceptance:

  • Usefulness impacts attitude substantively and consistently in both stages
  • Ease of use has inconsistent effect on attitude in the initial stages, seems to further subside and become non-significant in later stages (Davis et al 1989, Karahanna et al 1999)
    • People get trained, comfortable, care more about efficiency

Perceived usefulness

  • IS often viewed as a means to rewards
  • Such striving for rewards is independent of the timing or stage of such behavior

Hypotheses

  • H1: Users' level of satisfaction with initial IS use is positively associated with their IS continuance intention
  • H2: Users' extent of confirmation is positively associated with their satisfaction with IS use.
  • H3: Users' perceived usefulness of IS use is positively associated with their satisfaction with IS use.
  • H4: Users' IS continuance intention is positively associated with their perceived usefulness of IS use.
  • H5: Users' extent of confirmation is positively associated with their perceived usefulness of IS use.

Comparing TAM with Continuance Model

Similarities

Continuance model

  1. employs individual cognitive factors for predicting continued IS use
  2. reflects the belief-affect-intention causality characteristic of most IS use theories.

Differences

  1. Focuses on continued use rather than initial acceptance
    • Temporally and conceptually distinct phases of IS use
  1. Based on ECT which is a theoretically richer model because of temporally-based, post-acceptance variables (satisfaction, confirmation)
  2. TAM cannot explain acceptance-discontinuance anomaly

Empirical testing of the model

Data collected via cross-sectional field survey of online banking users

  • Banking is info-intensive
  • Banking industry is aggressive in deploying IS (Tan and Teo 2000)
  • Survey respondants use the online banking division (OBD) of a large US bank
    • N=1000, Only 122 useable responses, population=>1million
    • E-mail requesting participation in survey
  • Single-data collection site (one bank) to control for variation among banks' OBDs
  • Advantage of online-survey
    • lower cost
    • faster response
    • geographically unrestricted sample (Tan and Teo 2000)
    • in this case, appropriate to the population under investigation
  • Single round of data collection as OBD didn't want to be seen as "spammer" (358)

Demographics

  • Age
  • Gender
  • Incoe
  • Profession
  • Education level
  • Length of time with online accounts
  • Online account balance
  • Whether or not they also had trad'l bank acct
  • Previously changed banks and why

Operationalization of constructs

IS continuance intention

  • User's intention to continue using OBD
  • Extended Mathieson's 1991 behavioral intention scale
  • Likert scales
  • 3 items
    • 2 measuring respondents' intention to continue
    • 1 measuring overal discontinuance intention (negative wording control for potentional common method bias)

Satisfaction

  • Users' affect with (feelings about) prior OBD use
  • Adapted from Spreng et al's 1996 overall satisfaction scale
  • Seven-point semantic differential scales
    • "Very displeased/very pleased"
    • "Very dissatisfied/very satisfied"
    • "Very frustrated/very contented"
    • "Absolutely terrible/absolutely delightful"
  • Affect is best measured along bipolar evaluative dimensions (e.g. good/bad) (Ajzen and Fishbein 1977)

Perceived usefulness

  • Users' perception of the expected benefits of OBD use
  • Adapted from Davis et al's 1989 perceived usefulness scale
  • Likert scales
  • Four item PU scale
    • First three items tap into performance, productivity, effectiveness of OBD
    • Fourth assesses overall usefulness

Confirmation

  • Users' perception of the congruence between expectation of OBD use and its actual performance
  • New scale developed
  • Likert scales
  • Perceived confirmation empirically better predictor of satisfactoin because of its temporal proximity to satisfaction and because human intentions are guided by perceptions of confirmation even if they are biased or inaccurate (Ajzen and Fishbein 1977)
  • New scale created to avoid overlaps with other constructs
    • Two items examined perceived congruence of user experience and service level
    • Third item assessed respondents' overall extent of confirmation

Data analysis, results

Scale Validation

Confirmatory factor analysis (CFA) using EQS program assessing construct validity (Bentler 1989) (361)

Hypothesis testing

Structured Equation Modeling (SEM) approach using EQS

Discussion of results

Findings

  • Satisfaction foundt o be strongest predictor of users' continuance intention. (364)
  • Perceived usefulness was a predictor of continuance but
    • Size of effect appears to decrease over time
"Ignoring post-acceptance user satisfaction can have disastrous consequences for user retention (continuance.)" (364)

IS firms and other supply-side institutions should adopt a two-fold strategy for maximizing their return on investments in customer training

  • New users should be informed of potential benefits
  • Old (continued) users should learn how to use IS effectively to maximize confirmation and satisfaction (364)

Acceptance-discontinuance anomaly

Little understood phenomenon in TAM

  • Dissatisfied users (due to disconfirmation of expectation) may discontinue IS use, despite having positive perceptions of its usefulness
  • Dissatisfaction and not perceived usefulness is the necessary condition for IS discontinuance

Explaining Satisfaction with IS use

  • Satisfaction predicted primarily by confirmation of expectations
    • Secondarily by perceived usefulness of initial use
    • Satisfaction may have add'l salient predictors than those identified here
    • Meeting/exceding expectations more salient than functionality in predicting continued use
    • Confirmation also effects perceivd usefulness

Confirmation is a cognitive belief:

  • The extent to which users' expectation of IS use is realized during actual use
  • Influences subsequent IS use via the satisfaction and intention constructs

Suggested feedback loop?

Belief-affect-intention-behavior-belief

  • Continuous refinement of belief, affect/satisfaction, intention
  • Leads to long-term continuance/discontinuance

Limitations of the study

  • Low response rate might lead to non-response bias
  • Novelty of online data collection might have lead to novelty bias
  • Survey of current users means that recent discontinuers were not polled
  • Systematic difference between more recent adopters of OBD

Ideal empirical design

  • Longitudinal comparison of customers' pre-acceptance and post-acceptance perceptions
  • Cross-sectional nature does not enable temporal comparison
Retrieved from "http://www.kevindriscoll.org/wiki/Understanding_Information_Systems_Continuance"
Personal tools