The GSA study referred to can be found here; it's an interesting read. Here's it's conclusion (in its entirety):
TSA has taken several positive steps to validate the scientific basis and strengthen program management of BDA and the SPOT program, which has been in place for over 6 years at a total cost of approximately $900 million since 2007. Nevertheless, TSA has not demonstrated that BDOs can consistently interpret the SPOT behavioral indicators, a fact that may contribute to varying passenger referral rates for additional screening. The subjectivity of the SPOT behavioral indicators and variation in BDO referral rates raise questions about the continued use of behavior indicators for detecting passengers who might pose a risk to aviation security. Furthermore, decades of peer-reviewed, published research on the complexities associated with detecting deception through human observation also draw into question the scientific underpinnings of TSA's behavior detection activities. While DHS commissioned a 2011 study to help demonstrate the validity of its approach, the study's findings cannot be used to demonstrate the effectiveness of SPOT because of methodological limitations in the study’s design and data collection.Ouch...
While TSA has several efforts under way to assess the behavioral indicators and expand its collection of data to develop performance metrics for its behavioral detection activities, these efforts are not expected to be completed for several years, and TSA has indicated that additional resources are needed to complete them. Consequently, after 10 years of implementing and testing the SPOT program, TSA cannot demonstrate that the agency's behavior detection activities can reliably an d effectively identify high-risk passengers who may pose a threat to the U.S. aviation system.