Browse by



Recent Submissions

  • Errors in Variables or Bad Leverage at Some Observations ? 

    Blankmeyer, Eric (2011-10-01)
    Errors-in-variables is a long-standing, difficult issue in linear regression; and progress depends in part on new identifying assumptions. I characterize measurement error as bad-leverage points and assume that fewer than ...
  • A Brief Exploration in Statistics and Quantum Mechanics 

    Blankmeyer, Eric (2011-05-01)
    Quantum mechanics has some probabilistic or statistical features that have been considered paradoxical or exotic; at least, this impression is frequently conveyed in introductory textbooks on the subject and informal ...
  • Log-linear Indexes of Nominal Wage Rates and Employment in Eight U. S. Regions 

    Blankmeyer, Eric (2008-02-11)
    This paper illustrates the use of log-linear indexes to estimate and test hypotheses about interregional differences in nominal wages and employment. A two-way analysis of variance is applied to cross sections of U. S. ...
  • Improving the small-sample efficiency of a robust correlation matrix: a note 

    Blankmeyer, Eric (2007-04-23)
    This note reports small-sample simulations of a correlation matrix estimated robustly by P. J. Rousseeuw's MCD algorithm. It appears that the statistical efficiency of MCD can be improved significantly if a pairwise-difference ...
  • How robust is linear regression with dummy variables ? 

    Blankmeyer, Eric (2006-11-03)
    Researchers in education and the social sciences make extensive use of linear regression models in which the dependent variable is continuous-valued while the explanatory variables are a combination of continuous-valued ...
  • Percentiles of an Inflation Index by Quantile Regression 

    Blankmeyer, Eric (2006-06-01)
    This paper gives a methodology for estimating an inflation index using the quantile regression of Bassett and Koenker. The regression, which is orthogonal in the logarithmic price changes, is computed by linear programming ...