…a slight jump in their test scores at schools where teachers received performance-based pay… [emphasis added]
via Texas study suggests teacher merit pay works [St. Petersburg Times Gradebook blog].
Anyone see the mismatch between the headline and the text? Reminds me of a common stock-market reporting tactic: "The Dow Jones fell more than 14 points today!" Yeah, right, and what was that in percentage terms? The Dallas Morning News story includes appropriate caveats to the research, but at first blush, I'd bet that even in this study (and plenty of studies show no effect), the cost-effectiveness of merit pay is about the same or lower than the cost-effectiveness of reducing class sizes.