Constructs and variables in environmental research: a secondary review using Canonical Correlation Analysis

This study examines whether environmental disclosure and
environmental performance variables express distinct constructs using canonical correlation analysis technique and analyses secondary data in Cong and Freedman (2011) article as a case example. The analysis finds that the environmental performance variables express different dimensions of the performance construct, and the variables differ in their expression of the construct. This study applies canonical correlation analysis to the published Cong and Freedman (2011) article using the statistical mean, standard deviation, and the correlation table published in that article and examined using SPSS syntax outlined by Abeysekera (2014) article. The canonical correlation results indicated that MHPR and RRR variables express another dimension of the environmental performance construct. The MHPR variable contributed very little in 2003 and 2005 but contributed sufficiently in the 2004 year to the performance construct. Overall, the canonical correlation analysis of the data published by Cong and Freedman (2011) revealed that the environmental disclosure variable (SCORE) expresses a different construct from the environmental performance variable (PBR). PBR is the most representative variable of the environmental performance construct.