adoption of technology in public sector

dibakar · Post by **dibakar** » Fri Apr 16, 2010 8:43 am

Dear friend
I am trying to find out adoption of technology in government. most of the constructs (complexity, relative advantage, management support etc) would be measured in 7 point likert scale(1 strongly disagree, 7 strongly agree). We have one more construct type of government, which we think may affect adoption decision. this attribute is measured as 0 - Local Government, 1 - Provincial Govt, 2 Federal Government. As this scale is different from scales for other constructs I am not sure whether I can do a PLS regression here.
As I am very poor in stats., I would greatly appreciate if you can suggest me how to proceed (eg do I need to rescale).
One more point, the dependent variable is measured in a 6 point Guttman type scale.
Regards
Dibakar Ray

Diogenes · Post by **Diogenes** » Mon Apr 19, 2010 10:09 am

Hi,
The type of government is a categorical variable (0 - Local Government, 1 - Provincial Govt, 2 Federal Government.) that could be used as dummy variable (two), or a better way is to split the sample and run three models:
0 - Local Government,
1 - Provincial Govt,
2 Federal Government.

Some hints about model comparations are given in this forum, use search fuction.

Best regards,

Bido

dibakar · Post by **dibakar** » Wed Apr 21, 2010 7:50 am

thanks a lot. regards

dibakar · Post by **dibakar** » Wed Apr 21, 2010 4:10 pm

Dear Prof Bido, thanks for the reply. But I have a problem i want to use type of Government as one of the exogenous variable in the model (like size of farm, farm type). As samples under some groups are small i can't divide them for analysis either. Help would be highly appreciated.

Diogenes · Post by **Diogenes** » Thu Apr 22, 2010 1:22 am

Hi,

Transform the 3 exogenous variables in two:

............................ Dummy1 //...............Dummy2
Local Government, 0............//................0 (category of reference)
Provincial Govt, .........1............//............... 0
Federal Government.) .0...........//................1

Example: Y = b1X1 + b2X2 + 0,5*Dummy1 + 0,2 * Dummy 2

When Dummy1 = Dumm2 = 0 --> we have the equation to Local Gov.

When Dummy1 = 1 and Dummy2 = 0 --> we have the equation to provincial gov.

When Dummy1 = 0 and Dummy2 = 1 --> we have the equation to Federal gov.

Best regards,

Bido

dibakar · Post by **dibakar** » Wed Apr 28, 2010 4:52 pm

thanks, i am trying to work in that line. in case i face problem i would bother you again. regards