Neighborhood pattern/summary data questions

Hi Community,

Our research is related to a longitudinal analysis based on Dewey’s neighbood pattern and summary data. The data consistency over time is therefore important for us.

Since Jan 2023, Advan has become the data provider instead of Safegraph. The way to collect data may be changed according to our observation. The figure below shows the medians of the sums of the fields “raw_device_counts”, “device_home_areas” and “device_daytime_area” across months (from pattern dataset), also the “number_devices_residing” and “number_devices_primary_daytime” (from summary dataset). (Advan doc: Documentation (Public) - Google Drive)

I wonder:

  1. If there’s any explanation about the data collecting/processing procedures and the relation between these variables/fields,
  2. Also the dramatic variances of “number_devices_residing”, “number_devices_primary_daytime”, “device_home_areas” and “device_daytime_area” after Dec 2022, while “raw_device_counts” is always consistent.

Thank you!

Hi @yunjieh , the Advan data should be available historically, as well. Are these graphs using SafeGraph until 2023 and then Advan? Or are they all Advan? There are differences in the underlying panels, check out this post for details.

Hi Ryan, the data are all from Advan. The data of the figure “Neighborhood summary” are from this link Neighborhood Patterns - US Home Panel Summary. If you dispay the data’s median by month, you can have the same results.

We just received a new restatement of the 2023 Neighborhood Patterns and Panel data. I don’t know how much has changed in the new data, but I encourage you to pull from the most recent refresh.

We don’t have much info on the supply changes these partners are dealing with but I know their data supply changes regularly and they’re always working hard to make the data as consistent as possible over time.