This book provides a concise yet thorough introduction to methods that linguists can use to study patterns of sentence acceptability in speech. Experimental Syntax shows how to design, execute and analyze an appropriate survey experiment for a well-formed question about a matter of fact relative to sentence acceptability. The book also examines variability and demonstrates a method by which an investigator can make principled decisions as to whether two or more informants who use atypical dialects are using the same atypical dialect.
An essential classic for syntacticians looking to understand or get into experimental work. The first 3-4 short chapters on big-picture judgment stability and variance are the most important for skeptical theoreticians who may believe experimental work is "besides the point". There are already many general introductions to experiments and stats textbooks that talk about variance and counterbalancing and such, using generic data about flowers and diseases and census data. But this book presents it all, concisely, in the context of a syntax judgment experiment that will make very much sense to a dyed-in-the-wool theoretician.
I don't think a researcher with no background in stats or experiments would be able to conduct rigorous experiments based on this short book alone -- for example, you'll still have to consult a statistician in the design of your first experiments -- but it'll give you the foundation to start understanding experimental literature and start asking productive questions about statistics and experiments.