Sample for survey validation?

I would like to clarify that I am trying to generate a survey instrument to assess attitude and perception towards research from previous published studies. But the surveys used in these studies have not been validated.
I would like to validate my instrument before using it. Normally its recommended that the sample size for such studies be atleast 10 participants per item or 300 participants. I would like to know if this rule always holds?
I was thinking that since I am not creating an instrument from scratch but only compiling it from previous surveys, will I still need the same sample size or could I do it with a smaller sample size.