A Multiple-Item Scale for Assessing Electronic Service Quality

A Multiple-Item Scale for Assessing Electronic Service Quality

A. Parasuraman et al. 2005

Journal of Service Research

課程指導：任維廉 教授報 告 人：黃保源 日 期： 2009.12.17

2

Outline

Traditional Service Quality Versus Electronic Service Quality

Research on e-SQDevelopment & Refinement of a Scale to Meas

ure e-SQ Additional Assessment of the Refined ScalesRelative Importance of the E-S-QUAL Dimensi

onsDiscussion

3

Traditional Service Quality Versus Electronic Service Quality

Traditional Service Quality (a) A comparison of actual service performance

.

(b) The five SERVQUAL dimensions of reliability, responsiveness, assurance, empathy, and tangibles capture.

(c) Customer assessments of SQ are strongly linked to perceived value and behavioral intentions.

4

SERVQUAL

The five service quality dimensions collapses to three with online service quality (Gefen, 2002): Tangibles A combined dimension of responsiveness, relia

bility, and assurance Empathy

5

Research on e-SQ

2000 Loiacono WebQual (a scale for rating Web sites on 12 di

mensions)

2001 Yoo and Donthu A Nine-item SITEQUAL

2003 Wolfinbarger & Gilly Used online and offline focus groups, a sorting t

ask, and to develop a 14-item scale called eTailQ

6

Process Employed inDeveloping the Scale to Measure e-SQ

Step 1. Definition and Domain of e-service quality.

Step 2. Develop a preliminary scale.

Step 3. Sample Design and Data Collection.

Step 4. Data Analysis and Scale Reduction.

Step 5. Reliability and Validity Assessment.

Step 6. Additional Assessment of the Refined Scales.

7

Development & Refinement of a Scale to Measure e-SQ (Step 1) (1/3)

e-SQ is defined broadly to encompass all phases of a customer’s interactions with a Web site: The extent to which a Web site facilitates efficient and

effective shopping, purchasing, and delivery.

8

Development & Refinement of a Scale to Measure e-SQ (Step 1) (2/3)

Search engines,

One click ordering

Perceived ease of finding, perceived transaction speed

Ease of navigation, responsiveness to customer needs

Overall perceived quality and value

9

Development & Refinement of a Scale to Measure e-SQ (Step 1) (3/3)

The perceptual attribute level and categorized them into 11 e-SQ dimensions (Zeithaml, Parasuraman, & Malhotra, 2000): Reliability Responsiveness Access Flexibility Ease of navigation Efficiency Assurance/trust Security/privacy Price knowledge Site aesthetics Customization/personalizaiton

10

Development & Refinement of a Scale to Measure e-SQ (Step 2)

Preliminary Scale 121 items, Likert-type scale and low-high perfor

mance anchors, Reword items to improve clarity, Eliminate redundant items

Revised to 113 items with 5-point scales.

11

Development & Refinement of a Scale to Measure e-SQ (Step 3)

Sample Design & Data Collection Respondents: used the Internet at least 12

times during the past 3 months and made at least three purchases within that period.

Qualified respondents listed three most familiar sites.

One third of the respondents evaluated their favorite sites, another third evaluated their second-favorite sites, and the rest evaluated their third-favorite site.

549 completed questionnaires were collected.

12

Development & Refinement of a Scale to Measure e-SQ (Step 4) (1/3)

Data Analysis & Scale Reduction Reliability analysis: group the items according

to the 11 a priori conceptual dimensions. Examine corrected item-to-total correlations

and delete items whose elimination improved reliability coefficient alpha.

Eliminate items of high proportion of missing data.

Examination of dimensionality through exploratory factor analysis (EFA)

13

Development & Refinement of a Scale to Measure e-SQ (Step 4) (2/3)

E-S-QUAL Scale, consisting of 22 items on 4 dimensions. Efficiency: The ease and speed of accessing

and using the site. Fulfillment: The extent to which the site’s

promises about order delivery and item availability are fulfilled.

System availability: The correct technical functioning of the site.

Privacy: The degree to which the site is safe and protects customer information

14

Development & Refinement of a Scale to Measure e-SQ (Step 4) (3/3)

E-RecS-QUAL, e-Recovery Service Quality Scale, consisting of 11 items on 3 dimensions. Responsiveness: Effective handling of proble

ms and returns through the site. Compensation: The degree to which the site c

ompensates customers for problems. Contact: The availability of assistance through

telephone or online representatives.

15

Development & Refinement of a Scale to Measure e-SQ (Step 5)

Reliability and Validity Assessment The coefficient alpha values range from .83 to .

94 for E-S-QUAL and .77 to .88 for E-RecS-QUAL, exceeding the conventional minimum of 0.7.

The various fit indices for the CFA are also very good.

The root mean square error of approximation (RMSEA), which is slightly above the cutoff value of .06

16

17

18

Additional Assessment of the Refined Scales (1/3)

Administered the final scales via online surveys to representative samples of customers of Amazon.com (n=653) and Walmart.com (n=205) to:

re-confirm the scales’ reliability and validity and

Assess the relative importance of the various e-service quality dimensions in influencing consumers’ overall quality and value perception and loyalty intentions.

19

20

Additional Assessment of the Refined Scales (2/3)

Reassessment of Reliability and Validity The coefficient alpha values in both samples exceed th

e minimum standard of .7 The four dimensions of E-S-QUAL have consistently str

ong and positive correlations with perceived value (.52 to .72 for amazon.com and .48 to .73 for walmart.com) and loyalty intentions (.48 to .65 for amazon.com and .48 to .69 for walmart.com). These results attest to E-S-QUAL’s predictive validity.

21

Additional Assessment of the Refined Scales (3/3)

E-RecS-QUAL dimensions Perceived value and loyalty intentions are significant

for walmart.com, but only three of the six corresponding correlations are significant for amazon.com,

Perhaps because the number of respondents for several of these items was quite small for that site.

Moreover, as mentioned before, insufficient sample size for E-RecS-QUAL hindered reconfirming the scale’s factor structure and construct validity through CFA and SEM analyses.

22

23

24

Relative Importance of the E-S-QUAL Dimensions

E-S-QUAL assessments made at the dimensional level have an impact on higher-order evaluations.

To determine the extent to which each E-S-QUAL dimension contributes to this impact, we conducted three multiple regression analyses.

Results show that all four factor-score measures have significant positive effects on the three dependent measures in both samples.

25

26

Discussion

Set out to conceptualize, construct, refine, and test.

Authors created a subscale of E-S-QUAL—called E-RecS-QUAL

The basic E-S-QUAL (web site’s entire customer base) Scale is a four-dimensional, 22-item scale.

E-RecS-QUAL (the portion of the customer base with recovery service experience) is a three-dimensional, 11-item scale.

27

Comment

Recovery 的部份，如果不分為子量表，直接加在主量表是否可行。

是否可加入逆向物流，如：回收、維修…等因子。

Thanks for Your Attention