The Statistical Properties and Empirical Performance of Double Regression

Voting rights litigation requires ecological inference to estimatethe voting preferences of minority and nonminority groups withinthe electorate. Double regression has been the procedure mostcommonly employed for this purpose. This article presents thefirst formal examination of this procedure. The underlying structuralmodel reveals that double regression estimators are neitherunbiased nor consistent estimators of true within-group votepreferences or polarization. Simulations demonstrate that theycan substantially exaggerate the differences between minorityand nonminority vote choices when none are present, and dramaticallyunderstate them when differences exist. In sum, double regressioncannot meet conventional statistical standards for reliability.Consequently, it should be abandoned.     

Sentencing Using Statistical Treatment Rules: What We Don’t Know Can Hurt Us

The existing literature seriously misinterprets the available evidence on the predictability of high rate criminal offending and thus the potential value of statistical treatment rules that impose stiffer punishments on offenders with higher predicted risk of recidivism. The misinterpretation results from the failure to take account of the fact that the data used in existing risk assessment exercises come from environments characterized by informal (and sometimes formal) attempts by judges and other actors to base penal treatments on expected recidivism. Findings of little or no predictive power for baseline covariates may simply indicate the efficient use of the available information. We lay out the problem in detail, provide examples from several literatures and then consider general solutions to the problem.

Interrogating genomic diversity of E. coli O157:H7 using DNA tiling arrays

Here, we describe a novel microarray-based approach for investigating the genomic diversity of Escherichia coli O157:H7 in a semi-high throughput manner using a high density, oligonucleotide-based microarray. This microarray, designed to detect polymorphisms at each of 60,000 base-pair (bp) positions within an E. coli genome, is composed of overlapping 29-mer oligonucleotides specific for 60 equally spaced, 1000-bp loci of the E. coli O157:H7 strain EDL933 chromosome. By use of a novel 12-well microarray that permitted the simultaneous investigation of 12 strains, the genomes of 44 individual isolates of E. coli O157:H7 were interrogated. These analyses revealed more than 150 single nucleotide polymorphisms (SNPs) and several deletions and amplifications in the test strains. Pyrosequencing was used to confirm the usefulness of the novel SNPs by determining their allelic frequency among a collection of diverse isolates of E. coli O157:H7. The tiling DNA microarray system would be useful for the tracking and identification of individual strains of E. coli O157:H7 needed for forensic investigations.     

Voter response to congressional campaigns: new?techniques for analyzing aggregate electoral behavior

Scholarship on congressional elections typically analyzes either the sum or the division of the major party vote at the district level. Our main innovation is to model the votes received by each candidate as a separate dependent variable, utilizing statistical methods developed to analyze multiparty elections. Because we are dealing with a substantive area in which endogeneity problems are rampant, we synthesize an instrumental variables approach with a “compositional data” model. This approach allows one to study the mobilizing and persuasive effects of congressional campaigns at the aggregate level.     

Family Business or Social Problem? The Cost of Unreported Domestic Violence

Social interest in problems such as domestic violence is typically motivated by concerns regarding equity, rather than efficiency. However, we document that taking steps to reduce domestic violence by reporting it yields substantial benefits to external parties. Specifically, we find that although children exposed to as‐yet‐unreported domestic violence reduce the achievement of their classroom peers, these costs disappear completely once the parent reports the violence to the court. This suggests that the public has an interest in helping families to overcome their problems in general and to report domestic violence in particular. It also suggests that social and judicial interventions may help combat negative peer effects in schools.