Understanding consumer online shopping behaviour from the perspective of transaction costs

 
Table 4.18 Perceived Enjoyment Measurement Items 
 
ENJ1. Shopping on the Internet is one of my favourite leisure activities.
ENJ2. When shopping online, I feel playful.
ENJ3. I would assess the experience of online shopping as enjoyment. 
ENJ4. I enjoy shopping on the Internet. 
ENJ5. Online shopping can bring me happiness. 
 
 
4.6.17 Control Variables: Demographics 
Several control variables were incorporated to enhance the robustness of findings. 
Demographic variables, including gender, age, education, and income were measured using a 

---

192
categorical format. Specifically, the demographic variables were operationalized and 
measured as follows: 

Gender: Female, male (nominal data). 

Age: Age in years is indicated by selecting the appropriate age range (ordinal data). 

Education: Highest level of formal education completed (nominal data). 

Income: Indicate by selecting the appropriate household income range (ordinal data). 
4.7 INSTRUMENT VALIDITY AND RELIABILITY 
It is important to assess the instrument validity of reliability in this study. Research validity 
determines whether the research truly measures what it was intended to measure, or how 
truthful the research results are (Campbell
 et al.
1963). To ensure the research rigor, issues 
related to internal validity, content validity, construct validity and reliability must be 
addressed. Without rigor, research is worthless, becomes fiction, and loses its utility. 
According to Straub (1989), internal validity raises the question of whether the observed 
effects could have been caused by or correlated with a set of unhypothesized and/or 
unmeasured variables. Internal validity also refers to the confidence that can be placed in the 
cause-and-effect relationship. Campbell et al. (1963) stated that internal validity is the basic 
minimum, without which any experiment is uninterpretable. Straub (1989) further 
emphasized the need for a good instrument to ensure internal validity. He noted that survey 
instrument design has two basic goals: to obtain information relevant to the purpose of study, 
and to collect this information with maximal reliability. 

---

193
Content validity measures the degree to which the survey items represent the domain or 
universe of the trait or property being measured (Straub 1989). To ensure content validity, 
previously used and validated measures were used in this study.
According to Straub (1989), the construct validity concerns with the degree to which the 
survey items measure the construct they were designed to measure. Most importantly, the 
theory underlying the construct to be measured must be considered (Harrington 2008). Since 
most of the survey instruments developed for this study have borrowed the survey items from 
previously validated and tested instruments, to some extent, it can reduce such threat. Factor 
analyses including exploratory and confirmatory factor analyses were adopted to test the 
construct validity.
Reliability of the instrument is considered as an important issue in ensuring the rigor of this 
research. As presented by Straub (1989), reliability is a statement about the stability of 
individual measures across replication from the same source of information. Straub (1989) 
further suggested that Cronbach’s Alpha is a reliability coefficient that measures how well a 
set of items (or variables) measure a single unidimensional latent construct. When data has a 
multidimensional structure, Cronbach’s Alpha will usually be low.