Wireless sensor nodes and its inconsistency in reporting sensory data information tend to inaccurate processing at sink. Imputing information of sensors with collocated sensory information is needed to provide intermittent less processing at sink. This work deals with regression model in selecting the nearest sensor based upon nature of dependant variable. Poisson Regression based Imputed Data Information (PRIDI) has been used when there are equidispersion criteria of dependant variable is observed with collocated sensors. Negative Binomial Distribution model using Imputed Data Information (NBDIDI) has been used when over dispersion criteria of dependant variable is observed with collocated sensors. Thus both protocols estimate the imputed information considering the nature of collocated sensor data and type of reported interval information acquired at sink. Simulation comparison has been done with discrete event simulator for proposed protocols in network simulator 2.