In matchit(), setting method = "subclass" performs
subclassification on the distance measure (i.e., propensity score).
Treatment and control units are placed into subclasses based on quantiles of
the propensity score in the treated group, in the control group, or overall,
depending on the desired estimand. Weights are computed based on the
proportion of treated units in each subclass. Subclassification implemented
here does not rely on any other package.
This page details the allowable arguments with method = "subclass".
See matchit() for an explanation of what each argument means in a general
context and how it can be specified.
Below is how matchit() is used for subclassification:
Arguments
- formula
a two-sided formula object containing the treatment and covariates to be used in creating the distance measure used in the subclassification.
- data
a data frame containing the variables named in
formula. If not found indata, the variables will be sought in the environment.- method
set here to
"subclass".- distance
the distance measure to be used. See
distancefor allowable options. Must be a vector of distance scores or the name of a method of estimating propensity scores.- link
when
distanceis specified as a string, an additional argument controlling the link function used in estimating the distance measure. Seedistancefor allowable options with each option.- distance.options
a named list containing additional arguments supplied to the function that estimates the distance measure as determined by the argument to
distance.- estimand
the target
estimand. If"ATT", the default, subclasses are formed based on quantiles of the distance measure in the treated group; if"ATC", subclasses are formed based on quantiles of the distance measure in the control group; if"ATE", subclasses are formed based on quantiles of the distance measure in the full sample. The estimand also controls how the subclassification weights are computed; see the Computing Weights section atmatchit()for details.- discard
a string containing a method for discarding units outside a region of common support.
- reestimate
if
discardis not"none", whether to re-estimate the propensity score in the remaining sample prior to subclassification.- s.weights
the variable containing sampling weights to be incorporated into propensity score models and balance statistics.
- verbose
logical; whether information about the matching process should be printed to the console.- ...
additional arguments that control the subclassification:
subclasseither the number of subclasses desired or a vector of quantiles used to divide the distance measure into subclasses. Default is 6.
min.nthe minimum number of units of each treatment group that are to be assigned each subclass. If the distance measure is divided in such a way that fewer than
min.nunits of a treatment group are assigned a given subclass, units from other subclasses will be reassigned to fill the deficient subclass. Default is 1.
The arguments
exact,mahvars,replace,m.order,caliper(and related arguments), andratioare ignored with a warning.
Details
After subclassification, effect estimates can be computed separately in the
subclasses and combined, or a single marginal effect can be estimated by
using the weights in the full sample. When using the weights, the method is
sometimes referred to as marginal mean weighting through stratification
(MMWS; Hong, 2010) or fine stratification weighting (Desai et al., 2017).
The weights can be interpreted just like inverse probability weights. See vignette("estimating-effects") for details.
Changing min.n can change the quality of the weights. Generally, a
low min.w will yield better balance because subclasses only contain
units with relatively similar distance values, but may yield higher variance
because extreme weights can occur due to there being few members of a
treatment group in some subclasses. When min.n = 0, some subclasses may fail to
contain units from both treatment groups, in which case all units in such subclasses
will be dropped.
Note that subclassification weights can also be estimated using WeightIt, which provides some additional methods for estimating propensity scores. Where propensity score-estimation methods overlap, both packages will yield the same weights.
Outputs
All outputs described in matchit() are returned with
method = "subclass" except that match.matrix is excluded and
one additional component, q.cut, is included, containing a vector of
the distance measure cutpoints used to define the subclasses. Note that when
min.n > 0, the subclass assignments may not strictly obey the
quantiles listed in q.cut. include.obj is ignored.
References
In a manuscript, you don't need to cite another package when
using method = "subclass" because the subclassification is performed
completely within MatchIt. For example, a sentence might read:
Propensity score subclassification was performed using the MatchIt package (Ho, Imai, King, & Stuart, 2011) in R.
It may be a good idea to cite Hong (2010) or Desai et al. (2017) if the treatment effect is estimated using the subclassification weights.
Desai, R. J., Rothman, K. J., Bateman, B. . T., Hernandez-Diaz, S., & Huybrechts, K. F. (2017). A Propensity-score-based Fine Stratification Approach for Confounding Adjustment When Exposure Is Infrequent: Epidemiology, 28(2), 249–257. doi:10.1097/EDE.0000000000000595
Hong, G. (2010). Marginal mean weighting through stratification: Adjustment for selection bias in multilevel data. Journal of Educational and Behavioral Statistics, 35(5), 499–531. doi:10.3102/1076998609359785
See also
matchit() for a detailed explanation of the inputs and outputs of
a call to matchit().
method_full for optimal full matching and method_quick for generalized full matching, which are similar to
subclassification except that the number of subclasses and subclass
membership are chosen to optimize the within-subclass distance.
Examples
data("lalonde")
# PS subclassification for the ATT with 7 subclasses
s.out1 <- matchit(treat ~ age + educ + race + nodegree +
married + re74 + re75, data = lalonde,
method = "subclass", subclass = 7)
s.out1
#> A `matchit` object
#> - method: Subclassification (7 subclasses)
#> - distance: Propensity score - estimated with logistic regression
#> - number of obs.: 614 (original), 614 (matched)
#> - target estimand: ATT
#> - covariates: age, educ, race, nodegree, married, re74, re75
summary(s.out1, subclass = TRUE)
#>
#> Call:
#> matchit(formula = treat ~ age + educ + race + nodegree + married +
#> re74 + re75, data = lalonde, method = "subclass", subclass = 7)
#>
#> Summary of Balance for All Data:
#> Means Treated Means Control Std. Mean Diff. Var. Ratio eCDF Mean
#> distance 0.5774 0.1822 1.7941 0.9211 0.3774
#> age 25.8162 28.0303 -0.3094 0.4400 0.0813
#> educ 10.3459 10.2354 0.0550 0.4959 0.0347
#> raceblack 0.8432 0.2028 1.7615 . 0.6404
#> racehispan 0.0595 0.1422 -0.3498 . 0.0827
#> racewhite 0.0973 0.6550 -1.8819 . 0.5577
#> nodegree 0.7081 0.5967 0.2450 . 0.1114
#> married 0.1892 0.5128 -0.8263 . 0.3236
#> re74 2095.5737 5619.2365 -0.7211 0.5181 0.2248
#> re75 1532.0553 2466.4844 -0.2903 0.9563 0.1342
#> eCDF Max
#> distance 0.6444
#> age 0.1577
#> educ 0.1114
#> raceblack 0.6404
#> racehispan 0.0827
#> racewhite 0.5577
#> nodegree 0.1114
#> married 0.3236
#> re74 0.4470
#> re75 0.2876
#>
#> Summary of Balance by Subclass:
#>
#> - Subclass 1
#> Means Treated Means Control Std. Mean Diff. Var. Ratio eCDF Mean
#> distance 0.1141 0.0674 0.9626 1.0379 0.2771
#> age 24.4815 28.6536 -0.6501 0.3615 0.1150
#> educ 10.7407 10.2590 0.2544 0.4213 0.0401
#> raceblack 0.0370 0.0030 0.1802 . 0.0340
#> racehispan 0.2963 0.1506 0.3191 . 0.1457
#> racewhite 0.6667 0.8464 -0.3812 . 0.1797
#> nodegree 0.5556 0.5723 -0.0337 . 0.0167
#> married 0.2593 0.5904 -0.7555 . 0.3311
#> re74 3205.1533 6429.1274 -0.4426 1.0648 0.2177
#> re75 1983.7216 2631.3105 -0.2054 0.9066 0.0980
#> eCDF Max
#> distance 0.4934
#> age 0.2446
#> educ 0.1169
#> raceblack 0.0340
#> racehispan 0.1457
#> racewhite 0.1797
#> nodegree 0.0167
#> married 0.3311
#> re74 0.4266
#> re75 0.2263
#>
#> - Subclass 2
#> Means Treated Means Control Std. Mean Diff. Var. Ratio eCDF Mean
#> distance 0.4522 0.3958 0.6193 0.8134 0.1707
#> age 29.1154 29.5135 -0.0603 0.3486 0.1308
#> educ 9.1538 9.7568 -0.2064 0.9233 0.0599
#> raceblack 0.8846 0.7027 0.5694 . 0.1819
#> racehispan 0.1154 0.2973 -0.5694 . 0.1819
#> racewhite 0.0000 0.0000 0.0000 . 0.0000
#> nodegree 0.6923 0.6757 0.0360 . 0.0166
#> married 0.5000 0.5405 -0.0811 . 0.0405
#> re74 6754.9455 5617.3241 0.1467 1.5744 0.0817
#> re75 3312.9595 3257.5161 0.0116 1.5170 0.0581
#> eCDF Max
#> distance 0.2994
#> age 0.3129
#> educ 0.1674
#> raceblack 0.1819
#> racehispan 0.1819
#> racewhite 0.0000
#> nodegree 0.0166
#> married 0.0405
#> re74 0.2453
#> re75 0.1746
#>
#> - Subclass 3
#> Means Treated Means Control Std. Mean Diff. Var. Ratio eCDF Mean
#> distance 0.5998 0.6097 -0.4590 1.2565 0.1421
#> age 26.6923 25.9000 0.1037 0.2696 0.2119
#> educ 10.3846 8.8000 0.7416 0.3682 0.1731
#> raceblack 1.0000 1.0000 0.0000 . 0.0000
#> racehispan 0.0000 0.0000 0.0000 . 0.0000
#> racewhite 0.0000 0.0000 0.0000 . 0.0000
#> nodegree 0.6923 0.5000 0.4167 . 0.1923
#> married 0.5000 0.0000 1.0000 . 0.5000
#> re74 2416.9731 1748.8415 0.1851 2.6567 0.1324
#> re75 1350.8111 1004.3709 0.1863 1.8285 0.0677
#> eCDF Max
#> distance 0.3615
#> age 0.5077
#> educ 0.3846
#> raceblack 0.0000
#> racehispan 0.0000
#> racewhite 0.0000
#> nodegree 0.1923
#> married 0.5000
#> re74 0.2769
#> re75 0.2077
#>
#> - Subclass 4
#> Means Treated Means Control Std. Mean Diff. Var. Ratio eCDF Mean
#> distance 0.6582 0.6588 -0.0428 0.9551 0.0452
#> age 22.2963 24.2941 -0.4195 0.2181 0.1053
#> educ 10.4074 11.0588 -0.3613 0.5596 0.0787
#> raceblack 1.0000 1.0000 0.0000 . 0.0000
#> racehispan 0.0000 0.0000 0.0000 . 0.0000
#> racewhite 0.0000 0.0000 0.0000 . 0.0000
#> nodegree 0.4815 0.5294 -0.0959 . 0.0479
#> married 0.0370 0.2353 -1.0498 . 0.1983
#> re74 663.6473 1733.6054 -0.6576 0.1297 0.0617
#> re75 498.2558 1882.9978 -1.0580 0.1094 0.1153
#> eCDF Max
#> distance 0.1307
#> age 0.1808
#> educ 0.1765
#> raceblack 0.0000
#> racehispan 0.0000
#> racewhite 0.0000
#> nodegree 0.0479
#> married 0.1983
#> re74 0.1678
#> re75 0.2048
#>
#> - Subclass 5
#> Means Treated Means Control Std. Mean Diff. Var. Ratio eCDF Mean
#> distance 0.7013 0.6979 0.2860 0.8887 0.0765
#> age 24.3846 22.2941 0.2995 0.7142 0.1296
#> educ 10.1154 9.8824 0.1297 1.3668 0.0637
#> raceblack 1.0000 1.0000 0.0000 . 0.0000
#> racehispan 0.0000 0.0000 0.0000 . 0.0000
#> racewhite 0.0000 0.0000 0.0000 . 0.0000
#> nodegree 0.7308 0.8235 -0.2091 . 0.0928
#> married 0.0385 0.0000 0.2000 . 0.0385
#> re74 932.1795 812.5195 0.0403 3.9223 0.0899
#> re75 1551.4350 586.9098 0.1945 23.9287 0.0601
#> eCDF Max
#> distance 0.2805
#> age 0.3032
#> educ 0.0928
#> raceblack 0.0000
#> racehispan 0.0000
#> racewhite 0.0000
#> nodegree 0.0928
#> married 0.0385
#> re74 0.2783
#> re75 0.2014
#>
#> - Subclass 6
#> Means Treated Means Control Std. Mean Diff. Var. Ratio eCDF Mean
#> distance 0.7396 0.7425 -0.2126 1.4075 0.1214
#> age 23.8846 22.4615 0.1747 0.6952 0.1058
#> educ 10.3462 11.3846 -0.8436 0.4905 0.1202
#> raceblack 1.0000 1.0000 0.0000 . 0.0000
#> racehispan 0.0000 0.0000 0.0000 . 0.0000
#> racewhite 0.0000 0.0000 0.0000 . 0.0000
#> nodegree 0.9231 0.7692 0.5774 . 0.1538
#> married 0.0000 0.0000 0.0000 . 0.0000
#> re74 591.5328 517.2464 0.0596 2.0185 0.0545
#> re75 974.8401 479.1179 0.2851 4.6722 0.0769
#> eCDF Max
#> distance 0.2692
#> age 0.3462
#> educ 0.3077
#> raceblack 0.0000
#> racehispan 0.0000
#> racewhite 0.0000
#> nodegree 0.1538
#> married 0.0000
#> re74 0.1538
#> re75 0.1923
#>
#> - Subclass 7
#> Means Treated Means Control Std. Mean Diff. Var. Ratio eCDF Mean
#> distance 0.7837 0.7774 0.2999 2.7433 0.1407
#> age 29.8889 25.6667 0.6648 0.3002 0.2333
#> educ 11.2222 10.6667 0.3763 6.5385 0.0679
#> raceblack 1.0000 1.0000 0.0000 . 0.0000
#> racehispan 0.0000 0.0000 0.0000 . 0.0000
#> racewhite 0.0000 0.0000 0.0000 . 0.0000
#> nodegree 0.8889 1.0000 -0.3536 . 0.1111
#> married 0.0000 0.0000 0.0000 . 0.0000
#> re74 190.2635 281.4813 -0.1585 1.3943 0.0889
#> re75 1091.6910 1912.6611 -0.3734 2.5440 0.2189
#> eCDF Max
#> distance 0.2963
#> age 0.6667
#> educ 0.1481
#> raceblack 0.0000
#> racehispan 0.0000
#> racewhite 0.0000
#> nodegree 0.1111
#> married 0.0000
#> re74 0.2222
#> re75 0.7407
#>
#> Sample Sizes by Subclass:
#> 1 2 3 4 5 6 7 All
#> Control 332 37 10 17 17 13 3 429
#> Treated 27 26 26 27 26 26 27 185
#> Total 359 63 36 44 43 39 30 614
#>
# PS subclassification for the ATE with 10 subclasses
# and at least 2 units in each group per subclass
s.out2 <- matchit(treat ~ age + educ + race + nodegree +
married + re74 + re75, data = lalonde,
method = "subclass", subclass = 10,
estimand = "ATE", min.n = 2)
s.out2
#> A `matchit` object
#> - method: Subclassification (10 subclasses)
#> - distance: Propensity score - estimated with logistic regression
#> - number of obs.: 614 (original), 614 (matched)
#> - target estimand: ATE
#> - covariates: age, educ, race, nodegree, married, re74, re75
summary(s.out2)
#>
#> Call:
#> matchit(formula = treat ~ age + educ + race + nodegree + married +
#> re74 + re75, data = lalonde, method = "subclass", estimand = "ATE",
#> subclass = 10, min.n = 2)
#>
#> Summary of Balance for All Data:
#> Means Treated Means Control Std. Mean Diff. Var. Ratio eCDF Mean
#> distance 0.5774 0.1822 1.7569 0.9211 0.3774
#> age 25.8162 28.0303 -0.2419 0.4400 0.0813
#> educ 10.3459 10.2354 0.0448 0.4959 0.0347
#> raceblack 0.8432 0.2028 1.6708 . 0.6404
#> racehispan 0.0595 0.1422 -0.2774 . 0.0827
#> racewhite 0.0973 0.6550 -1.4080 . 0.5577
#> nodegree 0.7081 0.5967 0.2355 . 0.1114
#> married 0.1892 0.5128 -0.7208 . 0.3236
#> re74 2095.5737 5619.2365 -0.5958 0.5181 0.2248
#> re75 1532.0553 2466.4844 -0.2870 0.9563 0.1342
#> eCDF Max
#> distance 0.6444
#> age 0.1577
#> educ 0.1114
#> raceblack 0.6404
#> racehispan 0.0827
#> racewhite 0.5577
#> nodegree 0.1114
#> married 0.3236
#> re74 0.4470
#> re75 0.2876
#>
#> Summary of Balance Across Subclasses
#> Means Treated Means Control Std. Mean Diff. Var. Ratio eCDF Mean
#> distance 0.3104 0.2972 0.0585 1.0249 0.0454
#> age 25.1099 27.1896 -0.2272 0.3247 0.0996
#> educ 10.5836 10.3105 0.1106 0.5537 0.0332
#> raceblack 0.4040 0.3898 0.0371 . 0.0142
#> racehispan 0.0894 0.1233 -0.1139 . 0.0340
#> racewhite 0.5066 0.4869 0.0498 . 0.0197
#> nodegree 0.5194 0.6163 -0.2049 . 0.0969
#> married 0.2688 0.4150 -0.3257 . 0.1462
#> re74 2489.2647 4627.1084 -0.3614 0.5922 0.1342
#> re75 1532.5437 2199.1568 -0.2047 0.9287 0.0962
#> eCDF Max
#> distance 0.1674
#> age 0.2203
#> educ 0.0969
#> raceblack 0.0142
#> racehispan 0.0340
#> racewhite 0.0197
#> nodegree 0.0969
#> married 0.1462
#> re74 0.3399
#> re75 0.1839
#>
#> Sample Sizes:
#> Control Treated
#> All 429. 185.
#> Matched (ESS) 332.11 42.66
#> Matched 429. 185.
#> Unmatched 0. 0.
#> Discarded 0. 0.
#>