Are you trying to compare the slopes from two subsets of your data, such as men and women? If so, then Harshvardhan has given you good advice. Alternatively, if you are more familiar with dummy variables than with ANCOVA, you can accomplish exactly the same thing by using a dummy variable and an interaction effect between that dummy and the linear variable.

Hi. the first link below is to a calculator that purports to do the test. I have not tested it. the second explains the interaction method and gives you R code that shows you what happens when there is and isn't a difference. Most Design of Experiments books cover this in detail in the case of analyzing interaction in a 2x2 factorial design.I also included some youtube videos. Best wishes.

http://www.danielsoper.com/statcalc3/calc.aspx?id=103

http://stats.stackexchange.com/questions/18391/how-to-calculate-the-difference-of-two-slopes

https://video.search.yahoo.com/yhs/search;_ylt=A0LEVvDmKNpWIEsAqzcnnIlQ;_ylu=X3oDMTEyMmFyOXVjBGNvbG8DYmYxBHBvcwMxBHZ0aWQDQjE3MTlfMQRzZWMDc2M-?p=Interaction+In+a+2x2+Factorial+Design&fr=yhs-mozilla-002&hspart=mozilla&hsimp=yhs-002

Daniel -

Good info above, but I'd suggest not just relying on a p-value to make a decision, as a p-value is driven by sample size. Note that you can't reasonably use the same level in all cases.

In a simple case of just comparing two slopes, you might find a confidence interval about the difference in the estimated slopes.

A great deal of information can be found on the internet by searching on "comparing slopes," though you should confirm what you find, some sources being better than others. But as your question does not include a great deal of detail, I wonder if you might be looking at something more complex, where you might want to search on "multilevel modeling," or "random coefficients modeling?" 

Cheers - Jim

PS - I notice a power analysis section in Harshvardhan's last attachment.  Glad of that.  

Dear Daniel,

You can use modified method of one-way analisys of variance for compare linear regression. Look at my following papers.

www.milanarambasic.com