Olaseni Muritala Okunola and Jennifer Rowley
( 16.5 %), private sector( 32.8 %), self‐employed( 23.9 %), unemployed( 6.3 %), retiree( 0.9 %) and students( 19.6 %). Respondents classified themselves in terms of income as: low income earners( 44.2 %), medium income earners( 33.3 %) and high income earners( 22.5 %).
Table 1: Key variables included in the questionnaire
Questionnaire |
Number of |
References |
variables |
Items |
|
Ease of Use |
10 |
Davis et al.( 1989), Hung et al.( 2006), Wangpipatwong et al.( 2005, 2008), Mohammed et al.( 2009), |
Information Quality |
10 |
Eldon( 1997), Aladwani and Palvia( 2002), Hung et al.( 2006), Barnes and Vidgen.( 2007), Wangpipatwong et al.( 2005, 2008), Mohammed et al.( 2009), Karunasena, and Deng( 2012) |
Benefits |
11 |
Parasuraman et al.( 2005), Wangpipatwong et al.( 2005, 2008), Shareef et al( 2011) |
Barriers |
11 |
Sweeney et al.( 1999), Zhang and Prybutok( 2005), Udo et al( 2008) |
Trust |
11 |
McKnight et al.,( 2002), Hu et al.( 2009), Shareef et al( 2011), Papadomichelaki and Mentzas( 2012) |
Website Quality |
10 |
Yoo and Donthu( 2001), Wolfinbarger and Gilly( 2002), Zeithaml( 2002), Parasuraman et al.( 2005), Zhang and Prybutok( 2005), Udo et al( 2010), Shareef et al( 2011) |
Satisfaction |
9 |
Davis,( 1993), Olorunniwo et al.( 2006), Zhang and Prybutok( 2005), Udo et al.( 2010), Hu et al( 2009) |
3.5 Data analysis
Data was loaded into SPSS 19.0 for analysis. First, data was cleaned to remove any incomplete or inaccurate responses. Next, negatively worded items were re‐coded. Descriptive statistics were then generated both to profile the sample in terms of demographic data and to generate descriptive statistics for inspection prior to undertaking further analytical data analysis. Finally, data relating to the statements designed to collect perceptions of user experience with the NIS website were subjected to Exploratory Factor Analysis, using Principal Components Analysis, in order to eliminate items from the measurement scale and to identify the dimensions of the user experience in the context of the NIS website
4. Findings
Exploratory Factor Analysis was chosen for the creation of a measurement scale as this tool is suitable in complex sets of data for identifying the correlation among variables( Pallant 2010). The user experience statements were subjected to Principal Component Analysis( PCA).
Prior to conducting the PCA, the suitability of the data for this test was assessed through various means. Pallant( 2010) recommends that Kaiser‐Meyer‐Olkin Measure of Sampling Adequacy should be above. 6 and Bartlett’ s Test of Sphericity value is significant and should be. 05 or less to show that data set is suitable for factor analysis. In this analysis, Kaiser‐Meyer‐Olkin Measure of Sampling Adequacy value is. 917 and Bartlett’ s Test of Sphericity value is significant at. 000 and which confirmed the data set used is suitable for factor analysis.
Kaiser’ s criterion was applied and only the factors that have eigenvalues of 1.0 or more were retained; the scree plot test was used to validate the number of retained factors. The Cronbach’ s Alpha for each factor was calculated to ensure that scales are reliable and consistent and to ensure all the items measured the same underlying construct( Pallant 2010). The factors’ Cronbach’ s Alpha coefficients range from. 931 to. 859 which shows each factors’ scale to be either very good to excellent; Hair et al.( 2007 pg. 244) recommends that Cronbach’ s Alpha coefficients should be at least at least. 7. Once the number of factors is decided, the next step is to rotate the factors. In this study the rotation used was Varimax with Kaiser Normalisation. The Pattern Matrix showed a clear structure with meaningful strong loadings of variables onto only one component for most of the components. However, we have removed ten items because they do not load sufficiently strongly onto any specific factor. We have then named the factors in accordance with the items that load onto them and in Table 2 present the user experience scale developed by this research.
383