Updates to OPM Strategic Measure_Vision Zero dataset and metric calculation - M.D.1

[mateoclarke]

Vision Zero team has to report certain metrics to the Office of Performance Management. Of focus is the yearly number of KSI persons per 100k. https://data.austintexas.gov/stories/s/Number-of-persons-and-number-of-persons-per-100-00/xgcc-d979

Currently, this metric is powered by a static dataset that Surbhi helped Xavier publish under a tight deadline from OPM when our dataset wasn't yet stable. We need to figure out the best solution to get the OPM metric consuming the same regularly updated data as what we display in VZV.

• [x] @JaceDeloney to schedule time with Xavier and Surbhi to discuss

• [x] Decide on an approach below: Approach 1: Add KA per 100k (by Year) to our comprehensive Open Crash data. This is the fast and simple solution but that attribute would be repetitive and the same value for crashes of the same year. It kind of feels out of place, but some hacks are good. OR Approach 2: Create a new Socrata dataset that is a more simple table with important aggregated crash metrics like "KA per 100k (by Year)" as columns and Years as rows

• [x] Understand the updates required on the OPM measurement side so that we can design a sustainable solution (ex, future years)

[JaceDeloney]

Surbhi and Mateo met to discuss this issue. Next steps:

1. Surbhi to scheudle meeting with Nicole, Lewis, Xavier, and myself to discuss the best way to update/change the two datasets.
2. Decide on one of the two options outlined above, team is leaning towards Approach 2
3. Decide on which population numbers to use. Most likely want to go with the 2020 population Forecast
4. Update VZV to use this new population forecast

[SurbhiBakshi]

Meeting with Nicole scheduled for 08/27/2020

[johnclary]

+1 for Approach 2

[JaceDeloney]

Since @SurbhiBakshi also has to create two new measures using population data, it was decided that Approach 2 makes the most sense. Will assign this to Surbhi and discuss action items from tomorrow's meeting.

[SurbhiBakshi]

Hi @JaceDeloney , slight correction to the comment above, I have two more VZ related metrics for this wave, but not related to population. I agree with Approach 2, but I would still need to know what population data to use.

[SurbhiBakshi]

@johnclary @JaceDeloney @mateoclarke - Here is the table and measure card I created to prep for the meeting tomorrow - It probably does not match the VZ Viewer or OPM metric charts because I am sourcing from the VZ dataset but using the population numbers used for the OPM metric.

1. Dataset
2. Measure Card

[SurbhiBakshi]

• updated the assets above after replacing Total Population with Full Purpose Population.

[SurbhiBakshi]

Meeting with Nicole, Lewis, Xavier, Jace, John

1. Get Xavier and Nicole to look at assets
2. Lewis will get confirmation from management that we can switch out the old dataset and measure card
3. Replace older links with links to new measure card and dataset
4. Include some way to navigate to VZ Viewer for most recently updated data / add text about how data is updated.

[JaceDeloney]

Next steps:

• [x] Document queries needed to replicate the methodology used in the By Population chart in VZV
• [x] Work with @mddilley to define a solution that enables the M.D.1 measure to update automatically for future years (do not use a static crash number table, use the VZ crash table)
• [ ] Create documentation with link to VZV, explanation of changes made to this measure, measure methodology, definitions, and disclaimers

[amenity]

Using this dataset ... not the one Xavier was looking at.

• [x] @JaceDeloney add query here
• [x] Update Socrata with new table
• [x] Adding crash numbers to👆
• [ ] Demo for Lewis
• [ ] Push to Production

[JaceDeloney]

After speaking with Lewis during this Tuesday's Sprint Review meeting, he mentioned that he preferred that crash numbers do not update throughout the year in a dynamic way. He was concerned that this would cause confusion with visitors on why the data was changing. So we will need to use static crash data for 2015-2019.

Next steps:

• [x] Pull fatality and serious injury numbers by year from this dataset
• [x] Add the yearly crash numbers to this dataset (or a similar dataset, I don't want to be prescriptive here) that includes the latest Austin Area Population Histories and Forecasts
• [x] Create the updated M.D.1 metric by summing the Death count with the serious injury count, dividing by population, and multiplying by 100,000
• [x] Add the following content to the M.D.1 metric page:

Additional Measure Insights In August 2020, the Austin Transportation Department released Vision Zero Viewer, which allows people to view crash data by month, year, transportation mode, demographic groups impacted, time of day, and geographic location. Visit Vision Zero Viewer here.

Measure Details and Definition 6) Updates to Yearly Data: Crash numbers may change between January and June of each year as the Austin Transportation Department reviews and refines Vision Zero crash data that comes from the State of Texas.

Date page was last updated: September 2020. Explanation of changes: The total number of persons who were killed or seriously injured was updated in September 2020 to reflect the latest information available to the City of Austin. Additional information regarding the Vision Zero Viewer was added.

[JaceDeloney]

Surbhi has had her hands full with other SD 2023 metrics during this sprint, so M.D.1 was moved to Sprint 34

[SurbhiBakshi]

• Made all the changes
  • asked Lewis to review. He had one small change to the content of the Story Page which I made.
• Wrote to Nicole to review