I would like to test the impact of ICT on firm performance. Different firms of my sample deploy ICT on different time. If 2015 is implementation year (0), 2014 is pre-implementation (-1), followed by 2016 as post-implementation stage (+1).My study time 2010 to 2018. I took the variables namely ROA, ROS, ROE. I make dummy variable for high-tech firm as 1 and non-hightech firm as 0. Firm size is the control variable. To remove economy and industry effect I calculate my varibale (V) in following ways:
[(V1firm-post - V1 firm-pre) - (V1industrypost -V1industypre)]
Thus I got single performance measure ROA, ROS,and ROE corresponding to a firm, which is used as a dependent variable. High-tech dummy is used as random factor and firm size is used as covariate. I need to run GLM three time for each measures ROA,ROS and ROE.
How do I get coefficient using GLM(General Linear Model) that show firm performance ( ROA,ROS and ROE) increase from (-1) to (+1)?
See any text on Applied Linear Statistical Models. I prefer Kutner et al. Best, D. Booth