v1.0 data
See the code used that analyzes the data in this repository at wwbp/robust_spatiotemp
Contains the generated 2020 mental health scores generated in Robust language-based mental health assessments in time and space through social media. These scores control for 2019 findings.
lbmha_yw_cnty.csv
: depression and anxiety scores for 2020 measured from Twitter data. This data controls for the 2019 data, so 2019 is not included in this file.id
unique row identifier [0, 1, 2, ...]yearweek_cnty
concatenation of the year, week number (ISO format), and county (fips code) feat: either DEP_SCORE (depression) or ANX_SCORE (anxiety), derived from an adapted lexicon [2019_01:36091, 2020_52:22069]score_category
Depression (DEP_SCORE) or Anxiety (ANX_SCORE)n_lexwords
total count of words signalling the feat from all users in that yearweek_cntyscore
weighted average of scores signalling the feat from all users in that yearweek_cnty, weights are meant to better align with actual demographics of population and only occurs when sufficient users are presentscore_unweighted
average of scores signalling the feat from all users in that yearweek_cnty are presentscore_std
standard deviation of scores signalling the feat from all users in that yearweek_cnty are presentmoderate_reliability_criteria
whether the score is considered reliable based on the criteria of UT 50 [1 if true; 0 if false]high_reliability_criteria
whether the score is considered reliable based on the criteria of UT 200 [1 if true; 0 if false]n_users
number of unique users measured in that yearweek_cntycnty
redundant column for easy sortingcounty_name
expanded name of the countystate_abbr
standard state 2-letter abbreviation of statedate
The Wednesday of the ISO weekdate_range
Monday-Sunday of the ISO weeksyearweek
redundant column for easy sorting
lbmha_yw_cnty_undifferenced.csv
depression and anxiety scores for 2019 and 2020 measured from Twitter data. This data does not include any correction for other years. The data format is shared fromlbmha_yw_cnty.csv
.