SPSS AMOS INVARIANCE ANALYSIS ACROSS CULTURES CASE STUDY

SPSS AMOS INVARIANCE ANALYSIS ACROSS CULTURES CASE STUDY: INTERNET BANKING ACCEPTANCE Bander Al. Sajjan and CHARLES DENNIS

AGENDA • Importance of Internet Banking • Theory of Reasoned Action • Technology acceptance models • Theory of Planned Behaviour • Cross-cultural comparison • Conclusions

INTERNET BANKING (IB) • Banking is a natural service for the Internet: – Intangible – Information-based • 85 % of UK bank customers use IB • Take-up is slow in many countries • Saudi Arabia is 25 th ranked country by GDP • Only 20 % of Saudi Arabia’s bank customers use IB (Grais & Kantur, 2003)

THEORY OF REASONED ACTION (TRA) • ‘Simple basis for identifying how to target consumers’ behaviour change attempts’ (Sheppard et al, 1988) • Peoples behaviour is governed by: – Attitudes – Social factors and – Intention towards performing the behaviour (Ajzen & Fishbein, 1980)

TECHNOLOGY ACCEPTANCE MODEL • Models adoption of IS in the workplace • Users do not use a system for its own sake but rather: – Evaluate the benefits and expect a certain utility – Less the disutility of using the system • TAM refers to these as: – Usefulness – Ease of use (EOU) (Davis, 1989)

PERCEIVED USEFULNESS • Perceived usefulness affects attitudes and intentions (Shih & Fang, 2004) H 1 Perceived usefulness has a positive effect on attitudinal intention

Improves banking performance Easier to conduct banking Useful for banking transactions Perceived Usefulness

ATTITUDINAL INTENTIONS • Intentions have two components – Positive evaluation of the behaviour and – Facilitating conditions – equivalent to Perceived Behavioural Control (PBC) from the Theory of Planned Behaviour – or Perceived ease of use (from TAM) (Venkatesh et al, 2003) • Analogous to loyalty: – Attitudinal loyalty – the preference vs – Behavioural loyalty – subject to external constraints (Beerli et al, 2004)

ATTITUDINAL INTENTIONS • Positive evaluation and attitude are conceptually equivalent, which we term: Attitudinal Intentions

PERCEIVED MANAGEABILITY • Attitude and social factors cannot be the sole determinants of behaviour when control is incomplete • Theory of Planned Behaviour (TPB) aims to improve on TRA by adding ‘perceived behavioural control (PBC)’ (Ajzen, 1991) • PBC is equivalent to computer ‘self efficacy’ (Argarwal et al, 2000) • We combine these concepts as ‘Perceived Manageability’

PERCEIVED MANAGEABILITY • Perceived ease of use (PEOU) operates mainly through perceived usefulness (Ha & Stoel, 2008) H 2 Perceived manageability has a positive effect on users’ perceived usefulness • PEOU affects trust (Gefen et al, 2003) H 3 Perceived manageability has a positive effect on users’ users trust

TRUST

TRUST A willingness to rely on an exchange partner in whom one has confidence (Moorman et al, 1992) • Trust is central to e-commerce intentions (Fortin et al, 2002) • Lack of trust is an obstacle to e-commerce intentions (Liu et al, 2004)

TRUST

TRUST A willingness to rely on an exchange partner in whom one has confidence (Moorman et al, 1992) • Trust is central to e-commerce intentions and attitudes (Fortin et al, 2002; Pavlou, 2003) • Lack of trust is an obstacle to e-commerce intentions (Liu et al, 2004) H 4 Trust has a positive effect on users’ attitudinal intentions towards IB

TRUST • At higher levels of trust, customers perceive a website as more useful (Stewart, 2003) H 5 Trust has a positive effect on users’ perceived usefulness

SOCIAL FACTORS • ‘Subjective norms’ also influence intentions – Belief that important others dictate whether one should eshop – Motivation to comply with important others

SUBJECTIVE NORMS (SN) • The effect of SN tends to be fully mediated (Vankatesh & Davies, 2000) • As a social force, we argue that SN acts through a social belief - trust H 6 SN has a positive effect on users’ trust

SUBJECTIVE NORMS (SN) • Important others influence perceptions about manageability ‘I will give it a try if my significant other says that it’s easy to do’ H 7 SN has a positive effect on users’ perceived manageability

Perceived Manageability H 2 Perceived Usefulness H 1 H 3 H 5 Attitudinal Intentions H 7 H 4 Subjective Norms H 6 Trust Belief

METHOD • Sample: undergraduate students – Younger and better educated, matching typical Internet early adopters • Paper questionnaires • Arabic questionnaire back-translated • In capital cities

QUESTIONNAIRE ITEMS • Intention (Venkatesh & Davis, 2000) • Attitude (Suh & Han, 2002) • PU & PEOU (Wang et al, 2003) • Self-efficacy & Control (Wang et al, 2003) • Trust (Mc. Knight et al, 2002)

SAMPLE UK Saudi Arabia N 232 386 Response rate 49 82 % Female 44 43 Years Internet experience 4. 3 3. 3

PRINCIPLE COMPONENTS ANALYSIS • 6 factors extracted at Eigen > 1 – Matching the hypothesised constructs • No cross-construct loadings > 0. 3 – Good discriminant validity • Factor loadings > 0. 5 – (except for 2 items, which were dropped) – Convergent validity

AMOS STRUCTURAL EQUATION MODEL FIT UK Saudi Arabia Across groups Chi ² 181 212 396 df 107 216 CFI 0. 96 0. 97 0. 96 RMSEA 0. 055 0. 050 0. 037

AMOS STRUCTURAL EQUATION MODEL • The path coefficients for H 1 – H 7 are significant – p < 0. 03 • Mediators: – Perceived usefulness and – Trust fully mediate the impact of – Perceived manageability and – Subjective norm

Perceived Manageability 0. 6 Perceived Usefulness UK 0. 5 KSA 0. 7 0. 5 0. 3 Attitudinal Intentions 0. 3 KSA 0. 18 UK 0. 6 Subjective Norms 0. 16 Trust Belief

INVARIANCE ANALYSIS Multiple Group Analysis Parameter subsets Measurement weights Structural weights Models √ √ √ OK

INVARIANCE ANALYSIS • Do the measures have the same meaning for the two groups of respondents? • Assuming that the unconstrained model is correct Measurement weights are invariant across groups: – Δ chi ² – Δ df – p = 21 = 12 = 0. 08 • Only the 2 nd perceived usefulness item reduces fit – Chosen as the constrained value

REGRESSION PATHS INVARIANCE • Since we have metric invariance, we can test regression path invariance • Comparing UK and KSA with all regression paths constrained equal Significantly different, i. e. non-invariant across groups: – Δ chi ² – Δ df – p = 16 = 7 = 0. 02

INVARIANCE ANALYSIS Manage Models Model Name Structural weights Parameter constraints b 1_1 = b 1_2 b 2_1 = b 2_2 b 3_1 = b 3_2 b 4_1 = b 4_2 b 5_1 = b 5_2

INVARIANCE ANALYSIS Manage Models Model Name Structural weights Parameter constraints b 3_1 = b 3_2

REGRESSION PATHS INVARIANCE • Which regression paths are significantly different? • Each regression path in turn constrained equal across the 2 groups • All others freely estimated

NON-INVARIANT REGRESSION WEIGHTS Δ Chi ² Δ df p Perceived usefulness → Attitudinal intentions 5. 9 1 0. 02 Trust belief → Attitudinal intentions 6. 5 1 0. 01

NON-INVARIANT REGRESSION WEIGHTS • Saudi customers perceive system usefulness as more important than do UK customers • UK customers perceive a bank’s trustworthiness as more important than do Saudi customers

CULTURE DIFFERENCES (Hofstede, 1980) • Saudi customers perceive system usefulness as more important than do UK customers • In countries that tend to be more ‘feminine’, personal relationships are more valued (KSA), – Need more convincing of usefulness • In the more ‘masculine’ and assertive UK, the low social presence media are more accepted – Usefulness of Internet Banking may be taken for granted (Straub et al, 1997)

CULTURE DIFFERENCES (Hofstede, 1980) • UK customers perceive a bank’s trustworthiness as more important than do Saudis • People in high collectivism cultures (KSA) tend not to trust people outside their group – More likely to derive trust from relationships (Yamagishi & Yamagishi, 1994) • Individualistic cultures (UK) trust others – Rely on the environment to determine whether it is in another’s (i. e. the banks) interest to behave well (Bhawuk & Brislin, 1992)

CONCLUSIONS • Cross-cultural questionnaires are feasible • The Internet Banking model has relevance across cultures • Behavioural models may vary psychometrically between cultures

CONCLUSIONS • Perceived manageability forms a single construct, aggregating the previous: – Perceived ease of use – Self-efficacy – Perceived control • Attitude is implicit in behavioural intentions • 80 % of the variance in Attitudinal Intention is explained by: – Perceived usefulness and – Trust

MANAGERIAL IMPLICATIONS • Internet Banking is influenced by important others • Marketing communications need to consider word of mouth – Social networks?

SPSS AMOS INVARIANCE ANALYSIS ACROSS CULTURES CASE STUDY: INTERNET BANKING ACCEPTANCE Bander Al. Sajjan and CHARLES DENNIS