Matt VanEseltine

Sociologist, Programmer, Data Enthusiast

I am a sociologist with the Institute for Research on Science and Innovation, part of the Institute for Social Research at the University of Michigan. IRIS is a national consortium of research universities organized around an administrative data repository. We develop data to understand, explain, and improve the public value of higher education.
I study the effects of scientific infrastructure, training, and initiatives, and lately I'm especially interested in open science and reproducible research. I also contribute to open source projects in Python, and I maintain nominally (a maximum-strength name parser for record linkage).

Writing

2021, preprint: This pilot demonstrates that substantive evaluations of even relatively small-scale career development programs are possible using administrative data from universities with low burden on specific programs.

2018: The relative neglect of replication research in criminology is somewhat surprising […] opportunities for organizing something akin to a reproducibility project are nothing short of excellent in our field.

2014: If marriage affects antisocial behavior by increasing responsibilities, decreasing leisure time and time with friends, and changing identities and priorities […] our findings imply that cohabitation does not do these things.

2013: Our findings should come as a relief to parents and teachers who may worry about teenage employment most youth rate their jobs favorably, even though some jobs are clearly better than others.

2012: Criminal activity drops significantly following the birth of a first child. After several years, however, this drop fades. By the time the child is around age nine, the parent’s crime is, on average, no different than we would have expected if they had remained childless.

2011: Youth involved in intensive work are ‘at risk’ for engaging in behaviors that may lead to early—or, some might say, premature—family formation.

2009: Personal and contextual characteristics pertaining to social bonds, prior deviant involvement, self-protection, and negative self-appraisal all were predictive of subsequent tattoo acquisition.

2008: Close to half of the likelihood of imprisonment for parolees may be attributed to parole surveillance.

2008: Among incarcerated offenders, mental health problems were more strongly associated with assaultive violence and sexual offenses than with other types of crimes.