Dude, Where’s My Treatment Effect? Errors in Administrative Data Linking and the Destruction of Statistical Power in Randomized Experiments

Tahamont, Sarah; Jelveh, Zubin; Chalfin, Aaron; Yan, Shi; Hansen, Benjamin

doi:10.1007/s10940-020-09461-x

Cited by 18 publications

(15 citation statements)

References 131 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Due to the absence of global identifiers across records systems, we used a combination of the first three letters of a defendant's first name, the first three letters of a defendant's last name, and year of birth to link assessments to jail and court records within each county. Prior studies have shown that exact name matching can result in data loss and lower statistical power (Tahamont et al, 2020). We have found that local records often contain slight name misspellings, and this shortened identifier can detect the same individuals with good accuracy in a given jurisdiction.…”

mentioning

confidence: 82%

Five-County Validation of the Indiana Risk Assessment System – Pretrial Assessment Tool (IRAS-PAT) using a Local Validation Approach

et al. 2020

View full text Add to dashboard Cite

show abstract

mentioning

confidence: 82%

Five-County Validation of the Indiana Risk Assessment System – Pretrial Assessment Tool (IRAS-PAT) using a Local Validation Approach

et al. 2020

View full text Add to dashboard Cite

show abstract

“…When referring to links within a dataset, this is more commonly known as deduplication. Record linkage has received much attention in recent years with advances in computational techniques for linking data by matching on identifying information (Tahamont et al 2021).…”

Section: Quantifying Consolidationmentioning

confidence: 99%

“…Although a single entity might own many properties, its name or contact information might not be entered the same way in every record -creating errors that needed to be corrected. Fortunately, advances in text matching methods have significantly increased our ability to deal with the issues of messy data (Tahamont et al 2021).…”

Section: Introductionmentioning

confidence: 99%

Linking Landlords to Uncover Ownership Obscurity

Hangen¹,

O’Brien²

2022

Preprint

View full text Add to dashboard Cite

Defining the ownership of rental housing can be a difficult task. In recent years there has been an increasing obscurity of ownership in administrative records as more property owners use Limited Liability Companies (LLCs) on deeds and in tax assessment records. In many cases, this obscures the nature and scale of ownership and makes research into property ownership, investors, and landlords more challenging. To overcome these challenges, we compare different text-matching methods within property tax assessment records in Boston, MA from 2004-2019. We show that the source of the difficulty in creating an accurate knowledge of landlords and their portfolios of properties has shifted in the past decade from the scale of data and the messy nature of administrative data to an intentional strategy of obscurity through LLCs. To do so, we incorporate linking to corporate records to uncover intentional ownership obscurity. We assess the prevalence of obscurity among landlords as well as the extent to which it is undermining our ability to observe patterns in rental housing in ways that cannot be accounted for solely with text-matching. These include how obscurity hides not only an increasing consolidation of property ownership in the past decade, but also concentrations of disorder and evictions. In doing so, we demonstrate a comprehensive method for uncovering this obscurity and show how this representation of property ownership can form the basis for understanding inequities in rental housing and the effects of property consolidation.

show abstract

“…Recent work has also shown that even small amounts of linkage error can result in large effects on false negative (Type II) error rates in research studies. This is especially the case with small sample sizes that can occur with the rare effects that are often sought to be identified via record linkage from large population databases [78]. Such errors can even lead to research studies that are based on a small linked subpopulation to become useless [8].…”

Section: Misconceptions Due To Linking Datamentioning

confidence: 99%

“…One example is the increasing use of medical databases for public health research, compiled from the electronic health records of patients in hospitals or doctor's clinics, and potentially linked with databases that contain other personal information such as education or employment details of these patients. While analysing such linked data can provide exciting new insights into the effects of people's social status upon their health, not considering any potential bias or other limitations of such data can lead to wrong conclusions [10,26,43,78,83].…”

Section: Introductionmentioning

confidence: 99%

Common Misconceptions about Population Data

Christen¹,

Schnell²

2021

Preprint

View full text Add to dashboard Cite

Databases covering all individuals of a population are increasingly used for research studies in domains ranging from public health to the social sciences. There is also growing interest by governments and businesses to use population data to support data-driven decision making. The massive size of such databases is often mistaken as a guarantee for valid inferences on the population of interest. However, population data have characteristics that make them challenging to use, including various assumptions being made how such data were collected and what types of processing have been applied to them. Furthermore, the full potential of population data can often only be unlocked when such data are linked to other databases, a process that adds fresh challenges. This article discusses a diverse range of misconceptions about population data that we believe anybody who works with such data needs to be aware of. Many of these misconceptions are not well documented in scientific publications but only discussed anecdotally among researchers and practitioners. We conclude with a set of recommendations for inference when using population data.

show abstract

Dude, Where’s My Treatment Effect? Errors in Administrative Data Linking and the Destruction of Statistical Power in Randomized Experiments

Cited by 18 publications

References 131 publications

Five-County Validation of the Indiana Risk Assessment System – Pretrial Assessment Tool (IRAS-PAT) using a Local Validation Approach

Five-County Validation of the Indiana Risk Assessment System – Pretrial Assessment Tool (IRAS-PAT) using a Local Validation Approach

Linking Landlords to Uncover Ownership Obscurity

Common Misconceptions about Population Data

Contact Info

Product

Resources

About