- 10- The farm records provide information which is used for longer run evaluation on changes in practices and yields and comprise a more representative sample than of only those farmers who participated in Farmers' Tests. Ultimately, a cotipletely randomized scrmp of -1i target farmers will need to be conducted to determine adoption of tochnclogics, but this has not been undertaken in anj' area to date. COORDINATION WITH OTHER ENTITIES Figure 1 shows schematically this integrated, multidisciplinary approach to the generation 2nd promotion of technology for small, traditional farmters. Three factors in this figure were not discussed previouslye 1) the inputs to the s stem from international centers, universities, industry, etc.*. 2) the product from the agro-,sociocconomic studies which goes to the other entities within the Public Agricultural Sectorand 3) the relationship to other entities, both public and private, with respect to the transfer of tho technology to the target farmers and for other purposes. The two public agencies with which ICTA has the closest relationship are DIGESA (extension and credit assistance) and BANDESA (credit). Coordination at the inter-institutional level has been weak, but is now strengthening. The area of greatest emphasis is to create closer cooperation between ICTA's Farmers' Tests and initial extension tests or demonstration plots. Beginning in 1973, some DIGESA personnel will work under ICTA supervision in Farmers' Tests so they are familiar with the technology before it is placed in their control, At the same tine, the DIGESA personnel will be familiarized with the technology generating process and the new technology being evaluated in the Farm Trials. EXAMPLES FROM GUATEMALA In an area in eastern Guatemala_ the agro-sociocon omic survey provided information indicating that the tw' controllable factors most important in limiting' product-ion of the traditional farmers on the steed hillsides were the availability of labor in the short planting