I refer to this question about error propagation in calibration curves.
The reason for closure was not enough effort, but I disagree with this statement: the OP clearly explains his confusion and thought process, and I feel that this distinguishes it from the ubiquitous zero-effort questions that are normally closed for this reason.
In addition, I'd like to know whether this question is off-topic. While error propagation is certainly a statistical concern, we encounter it here in a chemical context, and this seems to be sufficient rationale for its belonging here. The Help Center states that questions pertaining to experimental techniques and technology are relevant, and serial dilution surely belongs in this category.