Page 9 - rohana_journal_No_13-2020-PRINT_03_of_03 -edit-2023_07-corre

Research Journal of the University of Ruhuna, Sri Lanka- Rohana 13, 2021

A reliability analysis is required to ascertain the degree of consistency between

multiple measurements of a variable (Black, 2009). The most commonly used

method of measuring reliability is the internal consistency measure, (Hair et al.,
2006); current study calculates this using Cronbach's Alpha. However assessing PLS

models, using Cronbach's alpha is disputed, rather using indicators with composite

reliabilities above 0.7 is recommended (Bagozzi & Yi, 1988). In order to be robust
current study calculates both these values.

Table 1: Operationalization of Variables

Variable Items Adopted from
Physical Easiness of Complain 4 Cai & Chi, 2018

Cognitive Easiness of Complain 4 Cai & Chi, 2018
Perceived Complaint Resolution 3 Stauss, 2002
Satisfaction
Brand Image 3 Low & Lamb, 2000

Attitudinal Loyalty 3 Bandyopadhyay & Martell, 2007 and
Chiou & Droge,2006

Behavioral Loyalty 6 Bandyopadhyay & Martell, 2007 and
Chiou & Droge,2006

Source: Survey Data, 2018.

Table 2: Construct Reliability and Validity

Cronbach's Composite Avg. Var. Extrac
Alpha Reliability (AVE)
Behavioral Loyalty 0.842 0.883 0.558

Attitudinal Loyalty 0.698 0.832 0.624
Brand Image 0.695 0.831 0.622

Cognitive Ease 0.834 0.889 0.668
Perceived 0.697 0.832 0.623

Complaint
Resolution
Satisfaction

Physical Ease 0.768 0.852 0.59
Source: Survey Data, 2018.

v