README for 2021 Grassroots Hollywood Homeless Count raw data ------------------------------------------------------------ Credits & Context: The count was organized by the Hollywood4WRD coalition and hosted by The Center In Hollywood. Funding was provided by The Central Hollywood Neighborhood Council, with special data collection by professional outreach teams from The Center In Hollywood and Covenant House. All data were obtained 25 Feb 2021 except tract 1919.02, surveyed 16 March. Data collection took place from 7:00p to 10:00p except for professional-counted tracts, which were surveyed the afternoon of 25 Feb. See Technical Report at hollywood4wrd.live/2021-homeless-count for more details. The count covered the LAHSA-defined "Hollywood" and "East Hollywood" *Communities*, which differ from the LAHSA-defined Hollywood *SPA*. Together, these comprise 40 US Census tracts (22+18, respectively). Our results are therfore comparable to official summaries at [1] or [2], but NOT [3]. Of the 31 tracts counted by volunteers, all except 1925.20 have at least two useable sets of observations. The remaining 9 tracts were surveyed once by professional outreach teams. [1] lahsa.org/data?id=45-2020-homeless-count-by-community-city [2] lahsa.org/documents?id=4686-2020-greater-los-angeles-city-community-homelessness-report-service-planning-area-4.pdf [3] lahsa.org/documents?id=4699-2020-greater-los-angeles-homeless-count-hollywood.pdf Data description: The data comprise a 16x75 element matrix where each line represents one survey of a given census tract. Column IDs: 1) Timestamp -- string; data entry time; set to "2/26/2021 11:28:00" for pro counts. 2) Team ID -- string; designation for counting team (32); "PRO" denotes professional outreach team. 3) Team Size -- int; estimated number of people in volunteer team. "-1" denotes unknown. 4) Tract -- float; US Census tract surveyed. 5-13) Number of persons or dwellings identified -- int; "Adults" are individuals 25+; "TAY" are 18-24 yr olds (Transition Aged Youth); "Minors" are individuals under 18; "Families" are at least 1 adult with at least one minor. 14. Verified? -- bool; True if digital and paper records cross-checked (True for all tracts.) 15. Flag -- bool; True indicates poor data quality (True for one survey of tract 1925.20). 16. Notes -- string; pertinent comments. East Hollywood Tract IDs: 1916.10, 1916.20, 1905.20, 1911.10, 1911.20, 1925.10, 1925.20, 1926.20, 1926.10, 1915.00, 1912.04, 1912.03, 1912.01, 1913.02, 1913.01, 1914.10, 1914.20, 1927.00. Instructions: 1. Average columns 5-13 over multiple measurements of each tract, rejecting lines flagged as poor quality as necessary. This should yield a 9x40 element matrix-- "COUNTS"--containing the best estimate of the number of persons + dwellings in each tract as estimated on 25 Feb (or 16 Mar for tract 1919.02). 2. Sum COUNTS over rows to obtain total tallies for persons and dwellings (9). 3. Sum COUNTS over columns to obtain all people+dwellings identified in each tract (40). 4. To obtain population estimates, the dwelling columns must be weighted by the relevant mean occupancy (ppl per dwelling). Columns 4 to 8 of COUNTS should thus be multiplied by the weight vector: wts = [C,V,R,T,M] = [1.51,1.77,1.42,1.48,1.68] where each entry is the average number of people estimated to live in a car, van, RV, tent, or makeshift structure, respectively. Our analysis assumed the 2020 SPA4/CD13 CVRTM values enumerated above (see: [4]). T was subsequently re-estimated by SELAH Hollywood volunteers and found to be consistent with the above based on 47 tents (38 responses) in Hollywood. Our results can be reproduced using the above procedure with the above weights, but the inferred unsheltered poplulation will depend on the [C,V,R,T,M] weight values. As such, the user may wish to employ updated weights based on information available to them, but not us. [4] lahsa.org/documents?id=4635-usc-2018-2020-multipliers-and-estimates-overview Contact: Please send any questions or comments to Louis Abramson at labramson.chnc@gmail.com.