Sometimes numbers do lie. That's the major theme of a breakout story in Sunday's USA Today. Reporters at the paper found more than 1600 cases where sudden surges in student test scores were fishy, to say the least. The big news was that, in the vast majority of cases, officials did nothing to investigate the anomalies.

Such evidence of cheating is without a doubt very shocking, especially when teachers and principals seem to be the culprits. But problems with state test results go far beyond cheating. Cut scores on state tests are often so low that the results of those tests are well nigh meaningless.

In the six states where USA Today looked for evidence of cheating, for example, 75 percent of fourth graders, on average, cleared the proficiency bar on their state math tests. Yet only 37 percent of those same students cleared the bar on the National Assessment of Educational Progress, which many experts see as the gold standard for US assessments. We can't blame cheating for that gap, but we can blame low expectations.

Until recently, for example, the bar in New York State was set so low that many students could pass by simply guessing on all of the questions. The State Board of Regents raised their cut scores last year and saw their state's student proficiency rates plummet. The Regents took the courageous and principled stand that an honest assessment of student knowledge and skills was worth the bad press.

Yet in many other states, state tests continue to set a low standard. A mother who hears that her child is doing just fine is in for a rude awakening when that child tries to make it in college or a career.

The vast majority of teachers and principals would never dream of cheating. It's the tests themselves that may not be telling the truth.