Data Scientist

Some of the examples of work but not limited to, in this role are:

Data Quality Assessment - At intake, the Data Scientist shall assess the quality of the data. Data quality assessment is a continual process and should include qualitative and quantitative assessment to include but not limited to summarizing all variables and providing statistics such as minimum, maximum, mean, median and number of missing values. Prior to providing preliminary and final results, the Data Scientist is responsible for conducting a data quality assessment to resolve all errors and data inconsistencies.

Data Cleaning - The process of transforming, parsing, aggregation and mapping data from one "raw" data form into another format with the intent of making it more appropriate and valuable for a variety of downstream purposes such as analytics. The goal of this process is to ensure the quality and usefulness of data. The process goal is for analysts typically spend the majority of their time in the process of data wrangling compared to the actual analysis of the data.

Data File Review and Cleaning Results - The Contractor shall document its understanding of each file review and cleaning task in its project management plan. During the file review the Contractor shall scrutinize the accuracy of large loan file submissions from respondents and summarize observations in a memo. This summary should include, but is not limited to, missing values, logically inconsistent values, or implausible values by comparing it to other sources of information, such as underlying loan documents. Subsequently, the Contractor shall clean data by handling missing values, logically inconsistent values, implausible values and as well as recode, create, or impute any variables needed. When required, the Contactor shall merge multiple databases, for example, combining a respondents administrative data with U.S. Census or Home Mortgage Disclosure Act (HMDA) data. The contractor shall provide a detailed data file that allows the government to review and replicate steps taken to analyze and clean datasets. Examples of steps taken include inputting data, merging datasets, creating variables, and conducting analysis. The contractor shall have the capability to perform data analysis to achieve the performance objectives in the PWS utilizing the following software:

• SAS (required)
• R and STATA
• Alternative mapping software that is approved by the government.

Data Analysis Reports Review:

• Operating Policy Review Results: The Contractor shall review each of the respondents written policies or guidelines and summarize the respondents history of fair lending violations, fair lending compliance program and policies, procedures for pricing, underwriting, referrals, compensations, and marketing. In addition, the Contractor shall summarize the institutions loan product offerings and loan activity and highlight issues of concern as it pertains to the Fair Housing Act.
• Peer Comparison Results: The Contractor shall identify similarly situated peers of potential fair lending violators and compare their loan activity among borrowers of various race/ethnicity or geographic areas of various racial/ethnic concentrations. When selecting peers, consideration should be made and documented regarding the structure of the institution, the type of products offered, marketing and lending footprint. The Contractor shall include appropriate legal, academic, policy and statistical resources and accompany the analysis with appropriate statistical hypothesis tests, goodness-of-fit statistics, and confidence intervals.
• Descriptive Analysis Results: The Contractor shall provide a descriptive summary of HMDA, U.S. Census, or other data to inform fair housing investigations. This could include, but is not limited to, providing demographics of a respondents housing development or mapping to illustrate a lenders market area. When appropriate, the Contractor shall use sophisticated statistical techniques such as kernel density estimation or statistical and geospatial graphics.
• Economic Damages Results: The Contractor shall utilize counterfactual analysis along with advanced statistical and regression techniques to evaluate the likelihood of a specific group being harmed by a particular housing or policy. These groups include protected classes under the Fair Housing Act, but may also include renters, surviving spouses, or those with low credit scores. The Contractor shall recommend methodologies for calculating economic damages, justify the rationale for the approach, and simulate economic damages based on the proposed approach.
• Replication Analysis Results: The Contractor shall propose methods to replicate respondents policies and practices, such as their underwriting or pricing policies, for the purpose of determining if a respondent systematically treats similarly situated individuals differently based on one or more of the prohibited fair housing bases. The Contractor shall justify its approach by citing relevant literature or applications in case law.

Requirements:

Experience working with Housing and Urban Development is preferred.

3 years professional experience in the field.

Should be proficient in using SAS and R or STATA software for statistical and regression analysis.

Job Type: Full-time

Pay: $113,000.00 - $143,000.00 per year

Benefits:

• 401(k)
• 401(k) matching
• Dental insurance
• Health insurance
• Paid time off
• Vision insurance

Experience level:

• 3 years

Schedule:

• Monday to Friday

Ability to Relocate:

• Rockville, MD: Relocate before starting work (Required)

Work Location: In person