Notes on Coronavirus: How is America Feeling? Part 5
Another week of learnings pulled from our data thanks to our amazing CoFounder & Chief Data Scientist, Bob Filbin.
Part 5: How is America Feeling?
1. Quarantine is the top trending word in our conversations.
Much of our data this week reflects the impact of quarantines, even more than the impact of the virus itself.
2. Essential workers. This past week, 19% of texters report being an essential worker.
What Industry? The two biggest categories: 34% work in food/service, 29% in healthcare.
- Fear of the virus: Almost half (45%) of essential workers mention a fear of contracting the virus as a major source of distress, compared to 37% of other texters.
- Financial stress: 56% mention financial issues as a major source of distress, compared to 44% of other texters.
3. America is feeling more depressed, more isolated, less anxious.
-
As quarantine, grief, and financial stresses continue or grow, we’re seeing the start of a shift from anxiety to growing isolation and depression.
-
Feelings of isolation are increasing (+5 points, to nearly ¾ of texters)
-
Feelings of depression are increasing (+4 points, to ⅔ of texters)
-
Feelings of anxiety decreased slightly (-3 points)
-
-
The data is based on the PHQ scale, a standard clinical measure of anxiety and depression: “For each of the following, how have concerns around coronavirus affected how you feel, compared to how you normally feel?”
4. Domestic violence is up.
In Feb 10.5% of conversations directly mentioned hit, violence, or abuse. In April, 13% of conversations mention those words.
5. Good news? America is NOT feeling more suicidal.
– In the last month, overall, fewer texters are expressing suicidal ideation (22% vs. 28% normally)
– Texters mention virus, COVID, or quarantine are ~½ LESS likely to be experiencing suicidal ideation (13% vs. 25% of other texters).
People have asked where we get this data? It is pulled by our CoFounder & Chief Data Scientist, Bob Filbin, from implicit and explicit data. Implicit data: we can pull word associations from conversations, track mentions, etc. Explicit data: because we’re always looking to learn and improve, we send an optional survey at the end of conversations. A whopping 19% of texters complete it. (And we had an imputation study conducted that confirmed the 19% responders are representative of the whole.) I’ve CC’d Bob here. Maybe you have ideas for queries for Part 6?
