Abstract
The objective of this research is to examine, validate, and recommend techniques for handling the problem of missingness in observational data. We use a rich observational data set, the Nielsen HomeScan data set, which allows us to effectively combine elements from simulated data sets: large numbers of observations, large number of data sets and variables, allowing elements of "design" that typically come with simulated data, and its observational nature. We created random 20% and 50% uniform missingness in our data sets and employed several widely used methods of single imputation, such as mean, regression, and stochastic regression imputations, and multiple imputation methods to fill in the data gaps. We compared these methods by measuring the error of predicting the missing values and the parameter estimates from the subsequent regression analysis using the imputed values. We also compared coverage or the percentages of intervals that covered the true parameter in both cases. Based on our results, the method of single regression or conditional mean imputation provided the best predictions of the missing price values with 28.34 and 28.59 mean absolute percent errors in 20% and 50% missingness settings, respectively. The imputation from conditional distribution method had the best rate of coverage. The parameter estimates based on data sets imputed by conditional mean method were consistently unbiased and had the smallest standard deviations. The multiple imputation methods had the best coverage of both the parameter estimates and predictions of the dependent variable.
| Original language | English |
|---|---|
| Title of host publication | Advances in Econometrics |
| Pages | 313-337 |
| Number of pages | 25 |
| Volume | 27 A |
| DOIs | |
| Publication status | Published - 2011 Dec 1 |
Publication series
| Name | Advances in Econometrics |
|---|---|
| Volume | 27 A |
| ISSN (Print) | 07319053 |
Fingerprint
Keywords
- Missingness
- Multiple imputation
- Nielsen homescan data
- Nonresponse
- Single imputation
ASJC Scopus subject areas
- Economics and Econometrics
Cite this
A missing variable imputation methodology with an empirical application. / Kyureghian, Gayaneh; Capps, Oral; Nayga, Jr, Rodolfo M.
Advances in Econometrics. Vol. 27 A 2011. p. 313-337 17004573 (Advances in Econometrics; Vol. 27 A).Research output: Chapter in Book/Report/Conference proceeding › Chapter
}
TY - CHAP
T1 - A missing variable imputation methodology with an empirical application
AU - Kyureghian, Gayaneh
AU - Capps, Oral
AU - Nayga, Jr, Rodolfo M.
PY - 2011/12/1
Y1 - 2011/12/1
N2 - The objective of this research is to examine, validate, and recommend techniques for handling the problem of missingness in observational data. We use a rich observational data set, the Nielsen HomeScan data set, which allows us to effectively combine elements from simulated data sets: large numbers of observations, large number of data sets and variables, allowing elements of "design" that typically come with simulated data, and its observational nature. We created random 20% and 50% uniform missingness in our data sets and employed several widely used methods of single imputation, such as mean, regression, and stochastic regression imputations, and multiple imputation methods to fill in the data gaps. We compared these methods by measuring the error of predicting the missing values and the parameter estimates from the subsequent regression analysis using the imputed values. We also compared coverage or the percentages of intervals that covered the true parameter in both cases. Based on our results, the method of single regression or conditional mean imputation provided the best predictions of the missing price values with 28.34 and 28.59 mean absolute percent errors in 20% and 50% missingness settings, respectively. The imputation from conditional distribution method had the best rate of coverage. The parameter estimates based on data sets imputed by conditional mean method were consistently unbiased and had the smallest standard deviations. The multiple imputation methods had the best coverage of both the parameter estimates and predictions of the dependent variable.
AB - The objective of this research is to examine, validate, and recommend techniques for handling the problem of missingness in observational data. We use a rich observational data set, the Nielsen HomeScan data set, which allows us to effectively combine elements from simulated data sets: large numbers of observations, large number of data sets and variables, allowing elements of "design" that typically come with simulated data, and its observational nature. We created random 20% and 50% uniform missingness in our data sets and employed several widely used methods of single imputation, such as mean, regression, and stochastic regression imputations, and multiple imputation methods to fill in the data gaps. We compared these methods by measuring the error of predicting the missing values and the parameter estimates from the subsequent regression analysis using the imputed values. We also compared coverage or the percentages of intervals that covered the true parameter in both cases. Based on our results, the method of single regression or conditional mean imputation provided the best predictions of the missing price values with 28.34 and 28.59 mean absolute percent errors in 20% and 50% missingness settings, respectively. The imputation from conditional distribution method had the best rate of coverage. The parameter estimates based on data sets imputed by conditional mean method were consistently unbiased and had the smallest standard deviations. The multiple imputation methods had the best coverage of both the parameter estimates and predictions of the dependent variable.
KW - Missingness
KW - Multiple imputation
KW - Nielsen homescan data
KW - Nonresponse
KW - Single imputation
UR - http://www.scopus.com/inward/record.url?scp=84874142128&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84874142128&partnerID=8YFLogxK
U2 - 10.1108/S0731-9053(2011)000027A015
DO - 10.1108/S0731-9053(2011)000027A015
M3 - Chapter
AN - SCOPUS:84874142128
SN - 9781780525242
VL - 27 A
T3 - Advances in Econometrics
SP - 313
EP - 337
BT - Advances in Econometrics
ER -