Exploring the Determinants of Internet Searches for Abortion-Related Information in the Midwest United States: A Google Trends Analysis Open Access
Nelson, Hans-Kristof (Spring 2023)
Abstract
Background:
Google Trends (GT) analysis first became available in 2004 after petitioned improve collection of this data as well as to release this data to the public. GT analysis is used in various public health research applications including in areas of reproductive health access and justice.
Methods:
Interest_by_city is a variable that was drawn from aggregated Google Trends data using the package ‘gtrendsR’. Data is returned as geolocated cities with a ‘hits’ value indicated that cities Google searching interest value (hits) relative to other cities in their state. We used this data as a measure of interest and examined this data against county level data from where the ‘hit_city’ exists. Examples of these county data are teen birth rate, metropolitan or non-metropolitan class, and percent of total poverty of each county. Simple t-tests, correlation tests, linear regression models and chi-squared tests are all methods of statistical analysis utilized. Table and spatial joins in ESRI ArcMap allowed for distance to nearest provider analysis.
Results:
For analysis the response variable (Y) was always ‘hits’ and the predictor variable (X) would change, there were significant as well as insignificant findings. Significant p-values are returned in the analysis of the county percent total poverty, and teen birth rate. Given the significant findings, there still exists a large limitation of the model. The project failed to control each hit city value by that cities population; giving a hit per capita variable is suggested to improve the balance of the analysis. Without controlling for population, a small, rural town measuring a 75 GT hits value is weighted the same as a large capital city measuring a 75 also. There may exist selection and sampling bias due to this fact.
Conclusions:
Even with the potential for selection bias, the model still was able to detect that of the hit cities that exist in counties with higher percent total poverty or teen birth rates, there could be increased Google search interest ‘hits’. The findings highlight a few things; the abortion access landscape is changing week to week it seems, real time surveillance of Google traffic related to abortion could be more important now than ever. Not to assume much further than the 5 states in the Midwest but counties therein with initiatives combatting poverty or those with historically high teen birth rates should be evaluated for the equity of their health care access and financial support systems.
Table of Contents
INTRODUCTION AND RATIONAL.. 1
The background. 1
Research Questions - 2
Significance Statement - 3
Definition of Terms - 3
REVIEW OF LITERATURE: 3
Justification - 5
Body. 5
Conclusion - 9
METHODS: 9
ANALYSIS & RESULTS: 13
DISSCUSSION AND MAPS OF DETERMINANTS OF ANALYSIS: 23
STRENGTHS. 29
LIMITATIONS. 30
FUTURE RESEARCH IN THIS AREA.. 30
About this Master's Thesis
| School | |
|---|---|
| Department | |
| Degree | |
| Submission | |
| Language |
|
| Research Field | |
| Keyword | |
| Committee Chair / Thesis Advisor |
Primary PDF
| Thumbnail | Title | Date Uploaded | Actions |
|---|---|---|---|
|
|
Exploring the Determinants of Internet Searches for Abortion-Related Information in the Midwest United States: A Google Trends Analysis () | 2023-04-22 12:09:34 -0400 |
|
Supplemental Files
| Thumbnail | Title | Date Uploaded | Actions |
|---|