The interest in promoting food and water security through development projects has led to the need to evaluate the impact of these projects. This study evaluates the impact from transitioning to a modern irrigation technology. Deciding to adopt or not an alternative irrigation technology (sprinklers) is not necessarily a random determination. Therefore, selection bias can be present and this can lead to biased estimates. In this study, we apply methodological specifications to correct for self-selectivity biases. Then, we measure and compare the technical efficiency scores from adopters and non-adopters. The empirical application uses data covering 56 small-scale greenhouse farms in Crete (Greece) for the cropping years 2009-2013. The results reveal that the average technical efficiency for farmers who adopted sprinkler irrigation is lower than the group of non-adopters, when the presence of selectivity bias cannot be rejected. This implies that either the farmers need more time to incorporate the know-how of the newly acquired technology or they become less motivated after the adoption. Consequently, agricultural water saving technologies need to be promoted in combination with support to the farmers, so they can cope with the lower performance in the first years after adoption.

通过开发项目促进粮食和水安全的兴趣导致需要评估这些项目的影响。本研究评估了过渡到现代灌溉技术的影响。决定采用或不采用替代灌溉技术（喷灌）不一定是随机决定的。因此，可能存在选择偏差，这可能导致估计有偏差。在这项研究中，我们应用方法论规范来纠正自我选择性偏差。然后，我们测量和比较采用者和非采用者的技术效率得分。实证应用程序使用的数据涵盖了克里特岛（希腊）2009-2013 年种植年的 56 个小型温室农场。结果表明，在不能拒绝存在选择性偏差的情况下，采用喷灌的农民的平均技术效率低于不采用的农民。这意味着要么农民需要更多时间来吸收新获得的技术的专业知识，要么在采用后他们的积极性降低。因此，农业节水技术的推广需要结合对农民的支持，使他们能够应对采用后最初几年的低绩效。