Skip to Main Content

Purdue University

Data Management for Undergraduate Researchers

How to collect, save, document and preserve research

Data Quality Checking Guide

Data Credibility	Timeliness	Information Completeness	Data Accuracy	Data Consistency	Data Deduplication
Who created the data?	Are data outdated?	Are there missing elements in data records?	Are there data typos?	Are data formats consistent?	Are there repeated data records?
Who published the data?	When are the data captured and updated?	Are there missing data records?	Are data formats correct?	Are data units measurements consistent?	Are data entered more than once?
Who contributed to the data?	Is version control implemented to track revisions of a data set?	Are all information captured for their intended uses?	Are there data outliers which may not be recorded accurately?	Are types of data consistent?
Is contact information available?		Are data described to be findable and reusable?	Do data represent the information we intend to capture?	Are data synched within and across platforms?

This checklist is available under a CC BY 4.0 license, with attribution to Wei Zakharov, Purdue Libraries and School of Information Studies.

Example 1

Research Questions

What are the impacts of covid vaccines on the Tennessee’s new cases numbers?
Are people of a particular age vulnerable to coronavirus disease?

Two data sources:

TN Department of Health. (2021). Daily case information [Data file]. Retrieved from https://www.tn.gov/health/cedep/ncov/data/downloadable-datasets.html
TN Department of Health. (2021). COVID vaccine state summary [Data file]. Retrieved from https://www.tn.gov/health/cedep/ncov/data/downloadable-datasets.html

Data quality checking:

questions from the data quality check guide answered in relation to this example question

Example 2

Research question:

How different driving scenarios impact driver’s choice of accepting paper coupons by stores? Investigated factors include age group, weather, passenger, and income.

Data source:

UC-Irvine Machine Learning Repository. Data were collected via a survey on Amazon Mechanical Turk.

Data quality checking:

Data Credibility	Timeliness	Information Completeness	Data Accuracy	Data Consistency	Data Deduplication
Who created the data? Tong Wang Cynthia Rudin	Are data outdated? Created on 9/15/2020	Are there missing elements in data records? Yes	Are there data typos? No	Are data formats consistent? Yes	Are there repeated data records? No
Who published the data? The Journal of Machine Learning Research	When are the data captured and updated? A Survey on Amazon Mechanical Turk	Are there missing data records? Yes	Are data formats correct? Yes	Are data units measurements consistent? Yes	Are data entered more than once? No
Who contributed to the data? Faneli Doshi-Velez Yimin Liu Erica Klampfl	Is version control implemented to track revisions of a data set? No. 1-time survey	Are all information captured for their intended uses? Yes	Are there data outliers which may not be recorded accurately? No	Are types of data consistent? Yes
Is contact information available? Yes		Are data described to be findable and reusable? Yes	Do data represent the information we intend to capture? Yes	Are data synched within and across platforms? Yes