The Partial Least Squares Spline Model for Public Health Surveillance Data

Table 1

Regression estimates of finally fitted PLS-spline model based on odds scale with one knot to select influential factors of infant mortality.

Selected factor	Estimate
Mother’s age	0.156
Region of residence	-0.191
Type of place of residence	0.257
De facto place of residence	0.258
Selected for domestic violence module	-0.164
Household has motorcycle/scooter	-0.133
Relationship of mother to household head	0.125
Reading newspaper or magazine	-0.108
Watching television	-0.222
Type of cooking fuel	0.133
Wealth index	-0.146
Number of births in last five years	0.103
Getting medical help for self: problem due to distance to health facility	0.197
Getting medical help for self: problem having to take transport	0.185
Getting medical help for self: not wanting to go alone	0.255
Awareness of tuberculosis	-0.126
Mother’s occupation	0.129
Person who usually decides on respondent’s health care	0.247
Person who usually decides on visits to family or relatives	0.170
Person who usually decides what to do with money husband earns	0.253
Beating justified if wife neglects the children	-0.191
Beating justified if wife argues with husband	-0.178
Beating justified if wife burns the food	-0.106
Preceding birth interval	-0.126
Succeeding birth interval	0.100
Blood relation with husband	0.153
Awareness about hepatitis	-0.147