My main question is: can I compare the number of visits from one week to another without making any adjustments or should I adjust those numbers based on the home-panel-summary or normalization-stats files?

Help
,
#1

First off - thank you so much for making this data available. I have a question about Weekly Patterns and supporting files. I have read the documentation (Weekly Patterns | SafeGraph Docs) but am still unclear as to what the normalization stats are showing. My main question is: can I compare the number of visits from one week to another without making any adjustments or should I adjust those numbers based on the home-panel-summary or normalization-stats files?

#2

Hi Marianne, I believe you should be able to compare the data week to week, but that may depend on what exactly you are trying to do. I will link a normalization sheet from safegraph that may help. https://drive.google.com/file/d/1shkAwx5c9CSwXIQBqAh-bCB50QTX09KG/view

there is also the chance to do your own calculations with this:

Panel Growth
• The panel has grown significantly since its inception. As such, it is important to normalize the data when doing time series analysis across long periods of time or multiple releases.
• We have seen success by normalizing visits by the total number of visits in the SafeGraph Panel, month by month. It is also worth exploring normalizing based on state or census block group. With each delivery, we provide you with the Panel Overview Data files to enable you to do these calculations.
I will ping @ross_epstein_safegraph to further clarify this, but I hope this helps.

#3

Hi Jack - thanks for your reply. I will read the linked PDF later today when I’m not in meetings :slightly_smiling_face:. I’m looking to compare the number of visits to businesses of a certain type (like restaurants) within a geographic area week to week starting March 1 of this year. Based on your response (without having read the PDF yet), it sounds like it would be ok to do that. Do you agree? This would be one example of the output.

#4

I believe due to only starting on March 1st, the data should be fine (it is so recent), but I would like that to be confirmed by @Ryan_Fox_Squire_SafeGraph just to be double sure.

#5

hi @Marianne_Cardwell_Indiana_University

There is no single correct answer here. The data is a panel dataset which varies day to day in size as devices leave and enter the panel. A device may be in the panel on Day 1 and Day 3 but not on Day 2.

There are many different ways to approach how to handle this. The Six Flags whitepaper is one way. MOre recently we’ve been using this methodology for our Covid-19 Dashboards, which I recommend over the Six Flags whitepaper

Across a short timer period, like within 1 month, you may not need to worry too much about normalization, but there is no clear cut answer.

here are some previosu threads that are relevant to this discussion:

https://safegraphcovid19.slack.com/archives/C0109NPA543/p1585758004307400

https://safegraphcovid19.slack.com/archives/C0109NPA543/p1587683795086100?thread_ts=1587675555.073800&cid=C0109NPA543

#6

Thanks Ryan! I’ll read the papers you linked and see what makes sense in our case.

#7

@Marianne_Cardwell_Indiana_University let me know if you have additional questions, and please post your findings and work-in-progress in #research-and-results