Working with Missing Data: Imputation of Nonresponse Items in Categorical Survey Data with a Non-Monotone Missing Pattern

Table 1

Results of logistic regression substantive analyses on 10 imputed data sets. The far right column reports the percentage of the imputed data sets which resulted in rejecting the null hypothesis that the parameter is equal to zero versus the two-sided alternative. The other columns represent the parameter estimates and values for the -test described in Section 1.6. , .

Index						% Reject
Child labor						100
English language proficiency						100
Parents’ education						100
Race						100
Sex						100

Dependent variable: socioeconomic success (0, 1).