(FORTUNE Magazine) – As every employer knows, the work force is divided into two classes of people: those who view midwinter blizzards as providential excuses for taking a day off, and those who view them as challenges requiring desperate measures to make it to the plant or store or office. An interesting question is whether it is possible to figure out who's in which camp at the point of hiring. Based on a forthcoming fascinating report in the Journal of Applied Psychology, we judge it surprisingly possible. The report, by University of Iowa industrial psychologists Deniz S. Ones, Chockalingam Viswesvaran, and Frank L. Schmidt, is about integrity tests -- the paper-and-pencil tests used by many employers, mainly in an effort to keep thieves and sociopaths off the payroll. Like most folks first reading about such tests, your servant started out highly skeptical of their validity. Why, after all, would a dishonest individual write in a truthful answer when asked, ''Please indicate the total amount of money you have stolen in your last job?'' (to cite one popular question). And yet the data show convincingly that the tests are pretty good at catching out the low-integrity population. For one thing, thieves have a propensity to believe that ''everybody does it'' and, accordingly, reason that it would be implausible to claim that they never do it themselves. In many circumstances, they also half-suspect that their track records are known to the prospective employer, and so once again they confess to misdeeds (while typically understating the depravity involved). For whatever reason, Schmidt mentioned the other day, ''it is very difficult for dishonest people to fake honesty.'' What is most interesting about the integrity tests, however, is not their ability to predict dishonest behavior in job applicants. Some tests do in fact claim only an ability to predict employee theft. Some more expansively claim an ability to predict worker productiveness. Some others are endeavoring to deliver personality profiles. But the Iowa psychologists are telling us that many standard integrity tests actually appear to deliver more than their designers claim for them. The authors believe that a sizable number of the tests measure what they call ''general conscientiousness.'' They have conducted a sweeping ''meta-analysis'' of more than 40 tests and of data reporting on test results for more than half a million testees. They report that employers giving many of the tests to job applicants can have substantial success in predicting and screening out the kind of irresponsible and counterproductive behavior that drives bosses crazy: disciplinary problems, disruptiveness on the job, chronic tardiness, excessive absenteeism. The tests tend to have quite substantial ''criterion validities,'' with test scores typically correlating significantly and positively with supervisors' subsequent reports on the behavior being predicted. There are some indications that the correlations are highest in more responsible jobs, where, of course, the employer has the most at stake. The Iowa authors come out recommending that integrity tests be combined with mental-ability tests, arguing that the combination optimizes employers' chances of landing the most productive workers. IQ and other mental-ability tests have had a rocky road in the workplace because they persistently show advantages for whites over blacks. But there is no black-white difference on integrity tests, and combining them with IQ tests would increase the proportion of minority hires. Also the proportion who show up in a blizzard.