The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.
staggR makes it easy to fit linear difference-in-differences models where interventions are staggered over time across multiple cohorts. The approach uses cohort- and time-since-treatment specific difference-in-differences parameters, and it provides convenience functions both for specifying the model and for flexibly aggregating coefficients to answer a variety of research questions.
install.packages("staggR")This is a basic example which shows you how to solve a common problem:
library(staggR)
#' Fit the staggered DiD model
sdid_hosp <- staggR::sdid(hospitalized ~ cohort + yr,
df = hosp,
intervention_var = "intervention_yr",
.vcov = sandwich::vcovCL,
cluster = hosp$county)
summary(sdid_hosp)
#>
#> Supplied formula:
#> hospitalized ~ cohort + yr
#>
#> Fitted formula:
#> hospitalized ~ cohort_5 + cohort_6 + cohort_7 + cohort_8 + yr_2011 +
#> yr_2012 + yr_2013 + yr_2014 + yr_2015 + yr_2016 + yr_2017 +
#> yr_2018 + yr_2019 + yr_2020 + cohort_5:yr_2010 + cohort_5:yr_2011 +
#> cohort_5:yr_2012 + cohort_5:yr_2013 + cohort_5:yr_2015 +
#> cohort_5:yr_2016 + cohort_5:yr_2017 + cohort_5:yr_2018 +
#> cohort_5:yr_2019 + cohort_5:yr_2020 + cohort_6:yr_2010 +
#> cohort_6:yr_2011 + cohort_6:yr_2012 + cohort_6:yr_2013 +
#> cohort_6:yr_2014 + cohort_6:yr_2016 + cohort_6:yr_2017 +
#> cohort_6:yr_2018 + cohort_6:yr_2019 + cohort_6:yr_2020 +
#> cohort_7:yr_2010 + cohort_7:yr_2011 + cohort_7:yr_2012 +
#> cohort_7:yr_2013 + cohort_7:yr_2014 + cohort_7:yr_2015 +
#> cohort_7:yr_2017 + cohort_7:yr_2018 + cohort_7:yr_2019 +
#> cohort_7:yr_2020 + cohort_8:yr_2010 + cohort_8:yr_2011 +
#> cohort_8:yr_2012 + cohort_8:yr_2013 + cohort_8:yr_2014 +
#> cohort_8:yr_2015 + cohort_8:yr_2016 + cohort_8:yr_2018 +
#> cohort_8:yr_2019 + cohort_8:yr_2020
#>
#> Residuals:
#> Min Q1 Median Q3 Max
#> -0.8007 -0.3788 -0.2139 0.4595 0.84
#>
#> Coefficients:
#> term estimate std_error t_value p_value
#> (Intercept) 0.228155340 0.031307912 7.2874659 3.2e-13 ***
#> cohort_5 0.088156223 0.028579679 3.0845771 0.00204 **
#> cohort_6 -0.159451272 0.016375109 -9.7374175 0.0e+00 ***
#> cohort_7 0.075133698 0.018413395 4.0803827 0.00005 ***
#> cohort_8 -0.237490795 0.003992926 -59.4778926 0.0e+00 ***
#> yr_2011 0.055426750 0.027992651 1.9800465 0.04771 *
#> yr_2012 0.025200365 0.048720589 0.5172426 0.60499
#> yr_2013 0.047663300 0.034037028 1.4003367 0.16142
#> yr_2014 0.110034046 0.035770010 3.0761536 0.00210 **
#> yr_2015 0.130370006 0.039124167 3.3322116 0.00086 ***
#> yr_2016 0.150645703 0.037030770 4.0681224 0.00005 ***
#> yr_2017 0.193950839 0.030086602 6.4464189 1.2e-10 ***
#> yr_2018 0.198984915 0.042004187 4.7372638 2.2e-06 ***
#> yr_2019 0.239084188 0.032639986 7.3248863 2.4e-13 ***
#> yr_2020 0.226537217 0.032775690 6.9117451 4.9e-12 ***
#> cohort_5:yr_2010 -0.075771022 0.048141117 -1.5739357 0.11551
#> cohort_5:yr_2011 -0.073492699 0.044324381 -1.6580649 0.09731 .
#> cohort_5:yr_2012 0.008214848 0.029858839 0.2751228 0.78322
#> cohort_5:yr_2013 0.018970874 0.035225443 0.5385560 0.59020
#> cohort_5:yr_2015 0.031183015 0.047308110 0.6591473 0.50981
#> cohort_5:yr_2016 0.073549064 0.039113486 1.8804016 0.06006 .
#> cohort_5:yr_2017 0.111184626 0.039801432 2.7934831 0.00522 **
#> cohort_5:yr_2018 0.181886621 0.033410927 5.4439263 5.3e-08 ***
#> cohort_5:yr_2019 0.188240613 0.019982991 9.4200420 0.0e+00 ***
#> cohort_5:yr_2020 0.257882482 0.043630603 5.9105871 3.4e-09 ***
#> cohort_6:yr_2010 0.192165497 0.040766852 4.7137683 2.4e-06 ***
#> cohort_6:yr_2011 0.083833784 0.021768788 3.8511001 0.00012 ***
#> cohort_6:yr_2012 0.130360273 0.017418812 7.4838784 7.4e-14 ***
#> cohort_6:yr_2013 0.103217201 0.054669436 1.8880239 0.05903 .
#> cohort_6:yr_2014 0.019231429 0.018934185 1.0156988 0.30978
#> cohort_6:yr_2016 0.039597597 0.014527663 2.7256687 0.00642 **
#> cohort_6:yr_2017 -0.017621795 0.045720916 -0.3854209 0.69993
#> cohort_6:yr_2018 -0.065811049 0.024388171 -2.6984823 0.00697 **
#> cohort_6:yr_2019 -0.071375212 0.028617604 -2.4941016 0.01263 *
#> cohort_6:yr_2020 -0.083590800 0.043178676 -1.9359278 0.05289 .
#> cohort_7:yr_2010 -0.028841089 0.040446429 -0.7130688 0.47581
#> cohort_7:yr_2011 -0.097944813 0.025984565 -3.7693458 0.00016 ***
#> cohort_7:yr_2012 -0.046622258 0.030682909 -1.5194862 0.12865
#> cohort_7:yr_2013 -0.036083533 0.037344341 -0.9662383 0.33393
#> cohort_7:yr_2014 -0.053870348 0.028528256 -1.8883155 0.05899 .
#> cohort_7:yr_2015 -0.042354696 0.030159042 -1.4043780 0.16022
#> cohort_7:yr_2017 0.123013287 0.013601784 9.0439086 0.0e+00 ***
#> cohort_7:yr_2018 0.131938968 0.032104648 4.1096531 0.00004 ***
#> cohort_7:yr_2019 0.159289976 0.015694481 10.1494260 0.0e+00 ***
#> cohort_7:yr_2020 0.189118922 0.037494013 5.0439765 4.6e-07 ***
#> cohort_8:yr_2010 0.169335455 0.030086602 5.6282679 1.8e-08 ***
#> cohort_8:yr_2011 0.162863929 0.014472854 11.2530626 0.0e+00 ***
#> cohort_8:yr_2012 0.231193914 0.024009151 9.6294083 0.0e+00 ***
#> cohort_8:yr_2013 0.185971220 0.018317313 10.1527563 0.0e+00 ***
#> cohort_8:yr_2014 0.166476981 0.011655637 14.2829587 0.0e+00 ***
#> cohort_8:yr_2015 0.050939971 0.009977242 5.1056167 3.3e-07 ***
#> cohort_8:yr_2016 0.059839177 0.007804591 7.6671765 1.8e-14 ***
#> cohort_8:yr_2018 0.009379666 0.013978605 0.6710016 0.50222
#> cohort_8:yr_2019 0.036917934 0.003425534 10.7772793 0.0e+00 ***
#> cohort_8:yr_2020 -0.003312873 0.016633407 -0.1991698 0.84213
#>
#> Significance codes: < 0.001: '**'; < 0.01: '**'; < 0.05: '*'; < 0.1: '.'
#> Residual standard error: 0.4632 on 30985 degrees of freedom
#> R^2: 0.118875596336873; Adjusted R^2: 0.117339991437799
#' What is the effect of the intervention in the post-intervention period?
ave_coeff(sdid = sdid_hosp,
coefs = select_period(sdid_hosp, period = "post"))
#> est se pval sign lb ub n
#> 1 0.09962262 0.0159212 3.890523e-10 *** 0.06841706 0.1308282 11745If you encounter a problem, please file an issue with a minimal reproducible example on GitHub.
These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.