SAMPLE SURVEY STATISTICS TEACHING: AN ALMOST WORLDWIDE PROBLEM ON TEACHING AGRICULTURAL SURVEY METHODS

Most national or large-scale multiple-purpose Agricultural Surveys, because of their complex nature, should be based on multiple frame survey designs that use samples selected from an area frame, list frames of holdings and housing units, and point sampling methods. The problem is that, although such Agricultural Surveys imply important human and economic resources as well as time for their implementation and maintenance, the area and multiple frame survey methods already mentioned are simply not taught at most universities and statistical teaching institutions. In other words, it has been almost a worldwide tradition to teach in an inappropriate way agricultural survey methods. However those methods are widely used and involve specialized statistical knowledge for their implementation and analysis. In this paper we shall refer to the reasons of such lack of appropriate teaching as well as possible solutions.