CDC PLACES · 168 New York City ZIP codes
This is the whole file, all semester. Every question we ask comes back to it. Sort it, search it, and look at how the answer changes depending on which column you rank by.
Same 168 neighborhoods, same file. Press one, then the other, and watch which places move to the top.
Read this before you quote a number
These are modeled small-area estimates, not a count of people who were actually asked. The CDC uses a national survey to estimate what each ZIP code probably looks like. So say about and roughly, never exactly. We take this apart properly in Week 12.
| Identifying the place | |
| zip | ZIP code. A label, not a quantity — ZIP 11436 is not "more" than 11432. |
| borough | Which of the five boroughs the ZIP sits in. |
| adults | Estimated number of adults living there. This is the "out of whom." |
| Frequent mental distress | |
| distress_pct | Percent of adults reporting 14+ days of poor mental health in the past 30. Our main measure. |
| distress_cases | Estimated number of adults, not a percent. Big neighborhoods have big numbers. |
| distress_low distress_high | The range the estimate could plausibly sit in. When two ranges overlap, be careful about claiming a difference is real — that's Week 9. |
| Depression | |
| depression_pct depression_cases depression_low depression_high | Same four shapes, for diagnosed depression. Our secondary measure. |
| Everything else — for comparison only | |
| no_insurance_pct poor_health_pct no_checkup_pct inactive_pct smoking_pct binge_pct short_sleep_pct diabetes_pct | Other health measures for the same ZIP codes. We use these to ask whether something else might explain a pattern — that's confounding, in Week 11. |
Download it: nyc_mental_health_by_zip.csv — 168 rows, 19 columns.
To load it in Colab, this is the line:
import pandas as pd
df = pd.read_csv("https://ph320.vercel.app/data/nyc_mental_health_by_zip.csv")
df.head()You are not required to run anything. Notebooks are prepared and shown in class.