Parenting in Place
Location
National study
Year
2020-2022
Background
I started my job at ASU in 2019 and my first year there was also the beginning of the COVID pandemic. Mostly what I remember from this time was feeling overwhelmed with fear and anxiety, and the uncertainty of not knowing what would happen next. I began to think of what we were all experiencing as collective trauma and I knew that there would be many long-term mental health impacts.
I reached out to the colleagues I had created Tender Press Books with years before, and we collaborated on a coloring book that addressed children’s fears, worries, and other big feelings about COVID. We hoped this book would also support parents by providing language to speak with their kids about the pandemic.
The response we received from parents was overwhelming, and I received many emails about how caregivers were struggling in terms of not only having to deal with their own stress, but also manage the anxiety of their children and explain to them what was happening when we as adults could barely understand the situation ourselves.
After hearing from so many parents I wanted to gather more data to understand the issues parents were facing, in order to ultimately find the most effective ways to provide support.
Study Overview
National snapshot of maternal mental health during the COVID-19 pandemic.
My hope for the study was to encapsulate a specific point in time during the pandemic and gather enough data to represent a national sample in addition to having enough power for complex analyses.
I decided to focus on mothers because in the United States mothers are still responsible for the majority of caregiving.
I also wanted to use standardized mental health assessments, so we could compare mental health across various points in time as well as to other studies.
Research Questions
Does past trauma history impact maternal mental health in the context of the COVID-19 pandemic?
How does maternal mental health impact perceptions of infant behavior in the context of the COVID-19 pandemic?
Does caregiving support impact maternal mental health in the context of the COVID-19 pandemic?
Deciding who to Include
One child between 0-3 years old
Identified as
a mother
Agreed to provide follow-up contact information
Choosing a Data Collection Method
After researching various ways to recruit large samples of participants online, I decided to work with a panel research company, which was something I had never done before.
Panel research companies have individuals pre-identified for recruitment and can gather data in a matter of weeks that would have taken my team months. Once I decided on panel research I reached out to several companies and compared their services based on:
-
Cost
-
Transparency in recruitment practices & participant incentives
-
Data quality
-
Timeline
Based on these parameters I ultimately decided to work with Qualtrics.
Designing the Study
-
5 Key Dimensions
-
Mental health
-
Trauma history
-
Infant behavior
-
Pandemic experience
-
Social support
-
-
65 questions
-
Short answer
-
Multiple choice
-
Scale questions
-
-
Measures
-
Validated
-
Modified for the COVID-19 pandemic
-
Ensuring Data Quality
In order to ensure the quality of the data I piloted the survey twice and iterated based on the responses.
-
First soft launch
-
10% of final sample size (30 participants)
-
50% questionable data
-
Revisions
-
Wording changes
-
Validity checks
-
-
-
Second soft launch
-
10% of final sample size (30 participants)
-
20% questionable data
-
Identifying Groups of Mothers
To analyze the data I utilized latent profile analysis, a type of person‐centered statistical approach. This type of methodology identifies subgroups of people based on their similarities on a set of variables, which felt the most reflective of the actual data because I was grouping people with similar mental health struggles together.
-
I then used multiple regression to examine what factors were different between groups
-
Although I couldn't pinpoint causation because I was only looking at one time point, I was able to examine significant factors between these groups that may be contributing to their different mental health profiles
Research Findings
-
Four classes of mothers found
-
These groups significantly differed on several variables
-
Significant trauma history was found to be the main indicator of negative infant perception, or viewing their infant in a negative light (i.e., my infant is more difficult than other infants)
-
Social support was a major protective factor for mental health outcomes
-
Grief and financial insecurity were the two pandemic experiences most closely linked with negative mental health outcomes
Why is this Important?
This study used innovative methodologies to identify subpopulations of mothers impacted by the COVID pandemic. We demonstrated how quantitative surveys can be used to create more nuanced prevention and intervention approaches that focus on specific groups of individuals based on various criteria.
This study also provided the foundation for a follow-up qualitative Zoom study, that further probed the experiences of high risk mothers, and assessed the mental health resources accessible to them.
The experience of designing this study led to my participation in a symposium dedicated to standardizing the process of collecting high quality data online. I also led a discussion about how to apply trauma-informed principles while collecting large-scale data.
Challenges & Lessons
During this study I spent a lot of time at each stage thinking about the responsibility I had to the participants in asking them such personal questions, that could bring up painful experiences. I wanted to make sure that each question we asked them had a purpose, and that the data collection process was completed in the most respectful way possible. I also made sure that we provided information about national resources in case any of the participants needed support during the study.
This focus on ethics extended to data analysis as well. After learning a lot about the theory behind person-centered data analysis, I believe this is a more representative methodology. Person-centered analyses assume various distinct subpopulations within a given sample while variable-centered analyses, which assume all individuals in a given sample come from a single population. Person-centered analyses reflect the diversity in modern society. Researchers should consider analytic approaches that facilitate the modeling of complex, socially constructed phenomena in ways that are both scientifically replicable and quickly translatable to practice.