Nt1310 Unit 1 Case Studies

PCA 1. The first PCA was conducted to construct the four categories of citizenship norms by using only data collected in 2004 (Table A11). It included the variables Actasoc, Buypol, HelpUSA, Helpwrld, Milserve, Obeylaws, Othreasn, Paytaxes, Voteelec, and Watchgov. Missing cases were deleted listwise, leaving a total of 1,406 valid cases from a population aged 18 to 89 years and older. The PCA was conducted using ten variables with orthogonal (varimax) rotation. An inspection of the correlation matrix revealed that all variables had at least one correlation coefficient r > .3. The KMO measure was .78, with individual measures exceeding the minimum level .5, confirming sample size adequacy. Furthermore, Bartlett’s test of spherity was statistically significant (p < .05), indicating that the data was suitable for PCA.
The initial analysis suggested three components with eigenvalues greater than one. Although a review of variance explained and a visual inspection of the scree plot indicated the retention of two factors, I decided to use four. Interestingly, the four-component solution also met the interpretability criterion and explained 63.14% of the total variance. Employing
The second PCA was carried out to construct the four categories of citizenship norms by using data collected in 2004 and 2014 (Table A11). It included the variables Actasoc, Buypol, HelpUSA, Helpwrld, Obeylaws, Othreasn, Paytaxes, Voteelec, and Watchgov. Missing cases were deleted listwise, leaving a total of 2,511 valid cases from a population aged 18 to 89 years and older. The PCA was conducted using nine variables with varimax rotation. An inspection of the correlation matrix revealed that all variables had at least one correlation coefficient r > .3. The KMO measure was .77, with individual measures exceeding the minimum level .5, confirming sample size adequacy. Again, Bartlett’s test of spherity was statistically significant (p < .05), indicating that the data was suitable for