Research Symposium

24th annual Undergraduate Research Symposium, April 3, 2024

Eden Sobalvarro He/Him Poster Session 3: 1:30 pm - 2:30 pm /306


A0F23381-DE09-46D2-86C5-C4D1688BD9F1_1_201_a.jpeg

BIO


I am a second-year student working towards degrees in Economics and Psychology at FSU. Raised in Miami, Florida, I now study in Tallahassee where I wish to gain research experience as well as law experience. I am currently involved in the Undergraduate Research Opportunity Program as well as the FSU Men's Club Soccer team. After completing my undergraduate studies, I plan on applying to a law program where I can earn my JD as well as my MBA simultaneously., in order to go into corporate law.

Construction of a Historical Infrastructure Price Index

Authors: Eden Sobalvarro, Carl Kitchens
Student Major: Economics and Psychology
Mentor: Carl Kitchens
Mentor's Department: Department of Economics
Mentor's College: College of Social Science and Public Policy
Co-Presenters: Ibraheem Saqib Ellahi, Christopher Lynch, Abhik Saha, Jesse Valdes

Abstract


This research project aims to develop a comprehensive historical construction price index spanning from the 20th century onwards, recognizing significant shifts influenced by factors such as inflation, technological advancements, and efficiency improvements.
In the early 20th century, the pricing of American infrastructure construction lacked digitization. To address this, microfilms from the Engineering News-Record are digitized through scanning microfilm and processing to make them machine readable. Leveraging microfilm provides access to a historical journal with weekly editions dating back to the 19th century, enabling an examination of prices for various construction elements, job-related salaries, and awarded contracts. After processing the aggregate microfilm data, images are corrected for transcription errors, and weights are assigned to individual projects. The organized aggregate data is then categorized at the city-year-infrastructure type level.
The extraction process employs text parsing and image formatting techniques to unveil relevant construction pricing information. This involves identifying monthly awarded construction contracts based on regional parameters. Specifically, machine learning methods, including Amazon Textract and Python data scraping syntax, are utilized to efficiently extract construction pricing from thousands of pages at a time.
The findings from this project hold the potential to assist policymakers and those involved in constructing new buildings in estimating potential costs. By identifying trends among historical decisions, this information contributes to more informed decision-making regarding future construction expenses.

IMG_1716.jpg

Keywords: Economics, Python, Social Sciences