We've been using the ACS5 data for some time and have found it very useful. I recently started using new fields for age and sex, and found what appear to be anomalies in the data.
For example, 253 of the 33,774 zip codes are entirely one sex. For example, zip 89070 with population 5,032; zip 93581 with population 3611; zip 75884 with population 3535 - all entirely male.
Regarding age, zip 13674 has median ages of 10.6 and 90.1 for male & female respectively; zip 30454 has median ages of 13.9 and 82.1 respectively. Maybe it is great grandmothers who are caregivers for children, but it seems strange.
Is this the real data, or maybe I did something wrong when downloading and processing it.
I would check the group quarters tables (B26XXX) and see if there are any group quarters (non-household) populations there - military and correctional facilities will skew very male. Nursing/skilled nursing…
I would check the group quarters tables (B26XXX) and see if there are any group quarters (non-household) populations there - military and correctional facilities will skew very male. Nursing/skilled nursing facilities will skew very female. University student housing will skew mildly female.
Thank you Diana. We tried B26001_001E and it was highly useful. Most of the age/gender disparities are in zip codes where this field has a high value in terms of % of total population. Conversely, for 90% of the zip codes this field is less than 5% of the population.