following the ddCT method, you'll need calibrators as one or more housekeeping genes to ensure a basal and stable expression line, and samples you'll take hoping the average won't move in contrast to the phenotype you're studying. follow the method at Article Analysis of Relative Gene Expression Data using Real-Time Qu...
To have a more bias-free reference you might use two or three reference (aka housekeeping) genes. I would recommend you to read about the standard curve method, we've just published a web tool that you can use with this method to help you analyse the data (Article QuantGenius: Implementation of a decision support system for...
).
Also, check MIQE guidelines for reporting (http://clinchem.aaccjnls.org/content/55/4/611.long) that many journals nowadays require you to follow if you want to publish.