Tobit analysis
a type of regression analysis used when a dependent variable with values above or below a certain threshold takes on the value of that threshold; that is, the exact value of the variable is unknown or ignored (see censored data). For example, in an analysis of aptitude in college students, the dependent variable could be a score on a standardized aptitude test that has an upper limit of 500 (the threshold value). Students who answer all questions on the test correctly receive a score of 500, even though these students may not have equal aptitude. In such a case, the influence of independent variables, such as reading and math scores, on the dependent variable of academic aptitude is more appropriately studied with Tobit regression analysis than with ordinary least squares regression because the threshold values would bias the slope of the obtained regression line. Also called censored regression;
Tobit model. [named by analogy with probit analysis, after James Tobin (1918–2002), U.S. economist]