I cannot get all of your data without a sign-on to your library. Using the Texas State library, I did look at the first report (the HGN review by Rubenzer and Stevenson). Recognizing my bias toward the police, I must say I think the article started from an anti-police or anti-SFST point of view. Several of the arguments in it were based on the presumption that the officer would change his scoring of clues on the HGN based on what the portable breathalyzer said. The authors note that the HGN was verified in testing but claim the test was incorrect because it compared HGN results solely to alcohol and not including other possible causes. They also note the SFST battery verification was a field test based on stopping people for poor driving and seeing if the test predicted intoxication properly. They claim this is flawed because of the selection bias of only stopping suspected drunks.
The problems with their claims are that if a cop is going to lie, no test is valid. I don't think it will happen but concede it is possible. The other two claims have flaws in logic that are similar. If I was testing HGN to use it as a diagnostic tool, they would be right. But if I am testing it to see if it predicts alcohol impairment, the tests seem valid to me. They acknowledge that officers are taught that there are other possible causes of HGN, and even of other instability problems for the other tests, but fail to acknowledge that the test has been validated for what the officers are attempting to do.
The test is used to help confirm the officer's prior suspicion that the person is intoxicated based on other factors present - like his driving mistakes. This is the key to the validity of the SFST, in my opinion. It is a factor used in conjunction with many other factors and it is never considered conclusive. It leads to the officer making an opinion based on his observations and training. The opinion may be confirmed later by a breathalyzer, or may not. If the breathalyzer does not confirm, the officer can give his opinion in court and the defendant can give their opinion and the jury can chose who they think is more accurate.
But, since you asked for reports of scientific testing, I thought you deserved more than just my opinion. In addition to advising you to check the article's references (he lists several validation studies), here is an older copy of the SFST instruction manual to read (how the tests are given and the officer's are taught):
http://breathtest.wsp.wa.gov/SupportDoc ... 202004.pdf" onclick="window.open(this.href);return false;
And here is the NHTSA page on a report on how to manage training for SFSTs, including some of the verification of the tests:
http://www.nhtsa.gov/People/injury/alco ... uction.htm" onclick="window.open(this.href);return false;