Package 'SBdecomp'

Estimation of the Proportion of SB Explained by Confounders

Description

Uses parametric and nonparametric methods to quantify the proportion of the estimated selection bias (SB) explained by each observed confounder when estimating propensity score weighted treatment effects. Parast, L and Griffin, BA (2020). "Quantifying the Bias due to Observed Individual Confounders in Causal Treatment Effect Estimates". Statistics in Medicine, 39(18): 2447- 2476 <doi: 10.1002/sim.8549>.

Details

The DESCRIPTION file:

Package:	SBdecomp
Type:	Package
Title:	Estimation of the Proportion of SB Explained by Confounders
Version:	1.2
Date:	2021-11-14
Author:	Layla Parast
Maintainer:	Layla Parast <[email protected]>
Description:	Uses parametric and nonparametric methods to quantify the proportion of the estimated selection bias (SB) explained by each observed confounder when estimating propensity score weighted treatment effects. Parast, L and Griffin, BA (2020). "Quantifying the Bias due to Observed Individual Confounders in Causal Treatment Effect Estimates". Statistics in Medicine, 39(18): 2447- 2476 <doi: 10.1002/sim.8549>.
License:	GPL
Imports:	stats, twang, graphics, survey
Config/pak/sysreqs:	make libjpeg-dev libpng-dev
Repository:	https://laylaparast.r-universe.dev
RemoteUrl:	https://github.com/laylaparast/sbdecomp
RemoteRef:	HEAD
RemoteSha:	4d6422cb329f31f228296bbe6a72616f084b441a

Index of help topics:

SBdecomp-package        Estimation of the Proportion of SB Explained by
                        Confounders
bar.sbdecomp            Creates a Bar Plot
petsdata                Dog ownership dataset
sbdecomp                Selection Bias Decomposition

This packge provides a function that decomposes the estimated selection bias to quantify what proportion of the estimated selection bias is explained by each observed confounder used in the propensity score model; the function is sbdecomp. The function offers two approaches - confounder inclusion or removal, and offers two estimation approaches - parametric or nonparametric. These methods allow one to identify the most important confounder when estimating a propensity score weighted treatment effect in the presence of selection bias.

Author(s)

Layla Parast

Maintainer: Layla Parast <[email protected]>

References

Parast, L and Griffin, BA (2020). "Quantifying the Bias due to Observed Individual Confounders in Causal Treatment Effect Estimates." Statistics in Medicine, In press.

Examples

data(petsdata)
sbdecomp(outcome = petsdata$genhealth, treatment = petsdata$gotdog, confounders = 
as.data.frame(petsdata[,c(2:13)]), type = "inclusion", estimation = "parametric")

sbdecomp(outcome = petsdata$genhealth, treatment = petsdata$gotdog, confounders = 
as.data.frame(petsdata[,c(2:13)]), type = "inclusion", estimation = "parametric", 
Bonly =FALSE, balance = TRUE) 

sbdecomp(outcome = "genhealth", treatment = "gotdog", confounders = c("age",
"ismale", "race_coll","hhsize","ownhome", "married",  "ontanf", "hhincome",
"fulltime","spouse_fulltime" ,"liveinhouse", "ruralurban"), data = petsdata, 
type = "inclusion", estimation = "parametric", Bonly =FALSE, balance = TRUE)

---

Creates a Bar Plot

Description

This function creates a bar plot to illustrate the estimated proportion of the observed selection bias explained by each observed confounder used in the propensity score model.

Usage

bar.sbdecomp(output.list, main)

Arguments

output.list	List outputed from main sbdecomp function
main	Title for plot, optional

Value

A bar plot showing the estimated proportions of the estimated selection bias explained by the observed confounders used in the propensity score model is created.

Author(s)

Layla Parast

References

Parast, L and Griffin, BA (2020). "Quantifying the Bias due to Observed Individual Confounders in Causal Treatment Effect Estimates." Statistics in Medicine, In press.

Examples

data(petsdata)
output = sbdecomp(outcome = petsdata$genhealth, treatment = petsdata$gotdog, confounders = 
as.data.frame(petsdata[,c(2:13)]), type = "inclusion", estimation = "parametric",
balance = FALSE) 

bar.sbdecomp(output)

---

Dog ownership dataset

Description

This dataset is a random subset of the publicly available 2003 California Health Interview Survey data; the data consists responses from 2,102 adults. For the purpose of illustrating the functions in this package, the goal is to investigate the effect of dog ownership on general health. Dog ownership was assessed with the question “Do you have any dogs that you allow inside your home?"; 29.0% of respondents owned a dog. General health status of the individual was measured as the self-reported response to the question “Would you say that in general your health is excellent, very good, good, fair or poor?" Responses were coded from 1 through 5 with 5 indicating “Excellent." Available individual characteristics i.e., confounders, in this dataset include age, gender, race/ethnicity, household size, marriage status, whether the individual received TANF (Temporary Assistance for Needy Families), household annual income, whether the individual worked full time, whether the individual had a spouse that worked full time, whether the individual lived in a house, and a rural/urban measure (1= urban; 2= 2nd city; 3 = suburban; 4 = town and rural) for the individual's address.

Usage

data("petsdata")

Format

A data frame with 2102 observations on the following 14 variables.

gotdog

indicates whether individual reported owning a dog where 1 is yes and 0 is no (binary)

age

age of the individual (numeric)

ismale

indicates the sex of the individual where 1 is yes and 0 is no (binary)

race_coll

ace/ethnicity of the individual, collapsed into categories where 1 is Latino, 4 is Asian, 6 is White, and 7 is a combined category of Pacific Islander/Native American/African American/other race/multiple race (factor/categorical)

hhsize

household size (numeric count)

ownhome

indicates whether the individual owns the home they live in where 1 is yes and 0 is no (binary)

married

indicates whether the individual is married where 1 is yes and 0 is no (binary)

ontanf

indicates whether the individual received TANF where 1 is yes and 0 is no (binary)

hhincome

household income, log-transformed (numeric)

fulltime

indicates whether the individual works full time where 1 is yes and 0 is no (binary)

spouse_fulltime

indicates whether the individual has a spouse that works full time where 1 is yes and 0 is no (binary)

liveinhouse

indicates whether the individual lives in a house (as opposed to an apartment, duplex, or mobile home) where 1 is yes and 0 is no (binary)

ruralurban

the rural/urban measure of the home address of the individual where 1= urban, 2= 2nd city, 3 = suburban, 4 = town and rural (numeric)

genhealth

self-reported general health of the individual, numeric 1-5 scale where 5 is excellent (numeric)

References

California Health Interview Survey (2005). CHIS 2003 Methodology Series. Los Angeles: UCLA Center for Health Policy Research.

Examples

data(petsdata)
dim(petsdata)
names(petsdata)

---

Selection Bias Decomposition

Description

This function decomposes the estimated selection bias to quantify what proportion of the estimated selection bias is explained by each observed confounder used in the propensity score model when estimating propensity score weighted treatment effects. The function offers two approaches - confounder inclusion or removal, and offers two estimation approaches - parametric or nonparametric.

Usage

sbdecomp(outcome, treatment, confounders, data=NULL, type = "inclusion", estimation 
= "parametric", Bonly = T, balance = T, n.trees = 20000, interaction.depth = 4, 
shrinkage = 0.005, verbose = FALSE, stop.method = c("es.max"), cv.folds = 0,
standard.error = F, boot.rep=500)

Arguments

outcome	Continuous outcome; numeric vector or name of the variable in supplied data
treatment	Binary treatment indicator, 0 or 1; numeric vector or name of the variable in supplied data
confounders	a data frame of confounders with names or a vector containing the names of the confounders in supplied data (each confounder must be either numeric or factor)
data	data containing all variables; required if names provided in outcome, treatment, and/or confounders arguments above
type	The type of method for selection bias deocomposition; options are inclusion or removal, default is inclusion
estimation	The type of estimation that should be used; options are parametric (based on logistic regression) or nonparametric (based on generalized boosted models [GBM]), default is parametric
Bonly	TRUE or FALSE; whether the user only wants the quantities B which are the proportions of the selection bias explained by each confounder, default is TRUE
balance	TRUE or FALSE; whether the user wants balance information, default is TRUE
n.trees	The n.trees for the nonparametric GBM approach, if using nonparametric estimation; default is 20000
interaction.depth	The interaction depth for the nonparametric GBM approach, if using nonparametric estimation; default is 4
shrinkage	The shrinkage for the nonparametric GBM approach, if using nonparametric estimation; default is 0.005
verbose	TRUE or FALSE for the nonparametric GBM approach, if using nonparametric estimation; default is FALSE
stop.method	The stopping method for the nonparametric GBM approach, if using nonparametric estimation; default is "es.max"
cv.folds	The cross-validation folds for the nonparametric GBM approach, if using nonparametric estimation; default is 0
standard.error	TRUE or FALSE; whether the user wants standard error estimates for the resulting estimates, obtained using boostrapping; default is FALSE; note that if using nonparametric estimation, obtaining the standard error estimates will be very time-intensive
boot.rep	The number of bootstrap replications to be used when calculating the standard error; default is 500

Details

Details can be found in: Parast, L and Griffin, BA (2020). "Quantifying the Bias due to Observed Individual Confounders in Causal Treatment Effect Estimates". Statistics in Medicine, In press. Please email [email protected] for a current version.

Value

A list is returned:

delta.naive	Estimated naive treatment effect
p.value.delta.naive	P-value for estimated naive treatment effect
conf.int.delta.naive	95% confidence interval for estimated naive treatment effect
delta.fully.adjusted	Estimated treatment effect after adjusting for all variables
p.value.delta.fully.adjusted	P-value for estimated treatment effect after adjusting for all variables
conf.int.delta.fully.adjusted	95% confidence interval for estimated treatment effect after adjusting for all variables
B	The proportion of the selection bias explained by each variable
estimated.selection.bias	The estimated selection bias, only if Bonly = FALSE
lambda	The defined selection bias movement when including/removing a variable, only if Bonly = FALSE
delta.each	The adjusted delta including only each variable or dropping each variable (depending on type), only if Bonly = FALSE
balance.naive.mean	mean of the standardized effect size differences across all variables between the two groups; only if balance = TRUE
balance.naive.max	max of the standardized effect size differences across all variables between the two groups; only if balance = TRUE
balance.fully.adjusted.mean	mean of the standardized effect size differences across all variables between the two groups, after weighting with propensity score weights calculated using all variables; only if balance = TRUE
balance.fully.adjusted.max	max of the standardized effect size differences across all variables between the two groups, after weighting with propensity score weights calculated using all variables; only if balance = TRUE
balance.mean	mean of the standardized effect size differences across all variables in the relevant propensity score model between the two groups, after weighting with relevant propensity score weights; if using type = "inclusion" and variable included is a factor, output will be the mean effect size difference across all levels; only if balance = TRUE
balance.max	max of the standardized effect size differences across all variables in the relevant propensity score model between the two groups, after weighting with relevant propensity score weights; if using type = "inclusion" and variable included is a factor, output will be the max effect size difference across all levels; only if balance = TRUE
B.standard.error	The estimated standard error for the estimated proportion of the selection bias explained by each variable; only if standard.error = TRUE
lambda.standard.error	The estimated standard error for the estimated selection bias movement when including/removing a variable; only if standard.error = TRUE

Author(s)

Layla Parast

References

Parast, L and Griffin, BA (2020). "Quantifying the Bias due to Observed Individual Confounders in Causal Treatment Effect Estimates." Statistics in Medicine, In press.

Examples

data(petsdata)
sbdecomp(outcome = petsdata$genhealth, treatment = petsdata$gotdog, confounders = 
as.data.frame(petsdata[,c(2:13)]), type = "inclusion", estimation = "parametric",
balance = FALSE)

sbdecomp(outcome = petsdata$genhealth, treatment = petsdata$gotdog, confounders = 
as.data.frame(petsdata[,c(2:13)]), type = "inclusion", estimation = "parametric", 
Bonly =FALSE, balance = TRUE) 

sbdecomp(outcome = "genhealth", treatment = "gotdog", confounders = c("age",
"ismale", "race_coll","hhsize","ownhome", "married",  "ontanf", "hhincome",
"fulltime","spouse_fulltime" ,"liveinhouse", "ruralurban"), data = petsdata, 
type = "inclusion", estimation = "parametric", Bonly =FALSE, balance = TRUE)

---