Herath, P.H.M.U.2012-07-312012-07-311998Bulletin of the Rubber Research Institute of Sri Lanka37pp.75-82https://dl.nsf.gov.lk/handle/1/9292This study indicates that the rubbersector could be divided intensively into non overlapping groups. The division is tally with the sector performances as well as the administrative structure. Therefor, stratified random sampling procedure would be employed effectively to obtain representative data of the rubber growers population for socio economic studies. The study describes the sampling procedure for the different srata.SociologySmall industryStatistical analysisSampling techniquesArticle