Water to Market Farmer Training 2007-2011 - Armenia

Abstract
---------------------------

The Farming Practices Survey (FPS) was commissioned by MCC to evaluate the impact of Water-to-Market (WtM) activities, particularly farmer training, on rural farmers in Armenia. Fielded by a consortium of AREG, an Armenia-based NGO, and Jen Consult, the FPS is a longitudinal survey of farming households interviewed at three points in time. FPS1 was conducted in 2007, before farmer training began. A second round was conducted one year after training began. Data from the second round is not included in this package. FPS3, the final round, was conducted three years after training began. This public-use file includes de-identified data from respondents to FPS3 and FPS1.

Households were selected for FPS1 interviews based on their likelihood of participating in WtM training, as assessed by mayors using criteria provided by the survey team. This process was used so that the surveyed households would include a high proportion of WtM participants. Each round of the FPS asked each household about their cropping patterns, irrigation and agricultural practices, crop yields, agricultural revenues and costs, other household expenditures, household employment, and other sources of household income.

Geographic coverage
---------------------------

Regional

Analysis unit
---------------------------

Households

Universe
---------------------------

The survey covered farming households in rural communities that were included in the evaluation sample for the Water-to-Market impact evaluation.

Kind of data
---------------------------

Sample survey data [ssd]

Sampling procedure
---------------------------

The evaluation design for the WtM activities dictated the sampling frame and approach to the FPS. The target was to complete interviews with approximately 25 farmers in each of 189 village clusters that was selected to be in the evaluation of WtM training. Village clusters consist of up to 4 small, neighbouring villages, and the 189 selected village clusters cover 211 villages. The village clusters are indicated in the variable "clusteringcode_b".

The baseline survey did not randomly sample respondents from the village clusters. The field team identified respondents for the FPS by working with village mayors to identify farmers who were likely to participate in WtM training so that a high proportion of farmers who were interviewed would have participated in training. The criteria were designed to align with the characteristics of farmers participating in ACDI's training programs, most notably, being actively engaged in farming, having modest farm area, living in the community for several years, and being between 25 and 70 years old.

AREG updated the sample list with the assistance of village mayors and marz officials, either at the marz offices or in the village itself. AREG and mayors targeted the households of farmers who were actively engaged in farming and had lived in the community for several years. Ultimately, a total of 4,715 farming households were interviewed for FPS1 in relevant communities. These same households were targeted for FPS3.

Sampling deviation
---------------------------

Three villages that were originally sampled for the FPS were not surveyed at final follow-up. Two villages that were surveyed at baseline were not surveyed at final follow-up because they were found to have almost no active farmers. A third village was not accessible for the baseline FPS due to heavy snow. The rest of the villages in these WUAs were surveyed at baseline and final follow-up according to the sample design.

For FPS3, MCA-Armenia also added the objective of surveying recipients of MCA credit. As a result, the FPS3 was administered to 33 new farmers who had not been interviewed in FPS1 and had received MCA credit.

Mode of data collection
---------------------------

Face-to-face [f2f]

Cleaning operations
---------------------------

Data edit:

After interviewers completed each questionnaire, the interviewers reviewed the questionnaire entries and submitted them to the field coordinator for cross-editing. During data entry in SPSS, mistakes were corrected using visual and program control. In the analysis phase, subsequent edits were made to logically impute data where appropriate. The data was entered in SPSS format by 4 specialists. Each set of responses for a questionnaire was entered by 2 specialists independently to cross-check skips and prevent mechanical mistakes. The first thousand and final five thousand entries were reviewed by Mathematica and MCA-Armenia, who compared the data entries to the hardcopy questionnaires and provided feedback on the data entry process. These data were transmitted to Mathematica for analysis.

Data processing:

After receiving the data, Mathematica merged the FPS3 and FPS1 data. While analysing the data, Mathematica identified several inaccurate records of farming households. These farmers were identified systematically based on their reported amounts harvested and sold at baseline versus follow-up. First, Mathematica identified specific crop harvests and amounts sold where the farmer's report changed by over 200 tons from baseline to follow-up. This identified fourteen farmers with harvests and sale amounts for barley, grape, peach, sweet cherry, potato, red beet, haricot, and gramma. None of the 14 identified harvests and sale amounts were accompanied by large changes in crop land area or revenues. Mathematica concluded that these results were likely to be outliers and replaced the outlying number based on the information about land and crop revenues. For many of these 14 harvests, this consisted of treating a reported amount sold as the revenues for that crop. This is plausibly a data recording error in that the FPS3 records crop revenues next to crop harvest amounts. Seven additional records were similarly recoded because they implied implausible prices per unit sold.

A second approach was used to address outliers for which there was insufficient evidence to conclusively determine if the reported value was accurate. The approach was to systematically censor outcome and baseline measures of income, production, cultivated land area at the 98th percentile for each measure, or the 2nd-highest value for that measure if the 98th percentile was less than or equal to zero. This process also helps de-identify any individuals with especially large amounts of income, production, or land.

Response rate
---------------------------

A response rate of 75%.

Sampling error estimates
---------------------------

Impacts of the WtM training program were estimated within a regression framework that controlled for baseline measures. Standard errors for the impact estimates were clustered at the village cluster level using Huber-White style "sandwich" estimators. Standard errors for key impact estimates are provided in Appendix B of the Water-to-Market Evaluation report, which is provided as a resource document.

Data appraisal
---------------------------

The censored variables were used to constructed nonresponse weights to adjust for differences in observed characteristics between households who did and did not respond to the FPS3. Nonresponse weights were calculated using the procedure described in Appendix A of the Water-to-Market Evaluation report. The code to construct these weights are located in the Stata program "1_armenia_construct.do". These materials are provided as external resources.

摘要
---------------------------
农田实践调查（FPS）由MCC委托，旨在评估水至市场（WtM）活动，尤其是农民培训，对亚美尼亚农村农民的影响。该调查由在亚美尼亚的NGO AREG和Jen Consult组成的联合体执行，FPS是对农田家庭在三个时间点的纵向调查。FPS1于2007年进行，在农民培训开始之前。培训开始后一年进行了第二轮调查。本数据包不包括第二轮的数据。FPS3，最后一轮调查，在培训开始三年后进行。此公共用途文件包括FPS3和FPS1受访者的去标识化数据。

农户选择参加FPS1访谈，是基于市长根据调查团队提供的标准评估其参与WtM培训的可能性。采用此过程，以确保调查的农户中包含高比例的WtM参与者。FPS的每一轮都询问每个家庭关于其种植模式、灌溉和农业实践、作物产量、农业收入和成本、其他家庭支出、家庭就业和其他家庭收入来源。

地理覆盖范围
---------------------------
区域

分析单元
---------------------------
农户

总体
---------------------------
该调查涵盖了包含在水至市场影响评估样本中的农村社区的农田家庭。

数据类型
---------------------------
样本调查数据 [ssd]

抽样程序
---------------------------
WtM活动的评估设计决定了FPS的抽样框架和方法。目标是完成约25名农民在189个村庄集群中的访谈，这些村庄集群被选为WtM培训评估的对象。村庄集群由最多4个小、相邻的村庄组成，189个选定的村庄集群覆盖了211个村庄。村庄集群在变量“clusteringcode_b”中指示。基线调查没有从村庄集群中随机抽样受访者。田野团队通过与村庄市长合作，识别FPS的受访者，以确定可能参与WtM培训的农民，以确保访谈的农民中有较高比例的培训参与者。这些标准旨在与ACDI培训项目的农民特征相一致，最值得注意的是，他们积极参与农业，拥有适度的农田面积，在社区居住多年，年龄在25至70岁之间。

AREG在村庄市长和州官员的帮助下更新了样本名单，无论是在州办公室还是在村庄本身。AREG和市长针对的是那些积极参与农业并在社区居住多年的农民家庭。最终，在相关社区对4,715个农田家庭进行了FPS1的访谈。这些相同的家庭被定位为FPS3的目标。

抽样偏差
---------------------------
最初为FPS抽样的三个村庄在最终随访时没有进行调查。在基线调查中调查的两个村庄在最终随访时没有进行调查，因为发现它们几乎没有活跃的农民。第三个村庄由于大雪而无法进行基线FPS调查。这些WUAs中其他村庄的基线和最终随访都按照抽样设计进行了调查。对于FPS3，MCA-Armenia还增加了调查MCA信用接受者的目标。因此，FPS3被施用于33名在FPS1中未接受访谈并已获得MCA信用的新农民。

数据收集方式
---------------------------
面对面 [f2f]

数据清洗操作
---------------------------
数据编辑:

在访谈员完成每份问卷后，访谈员审查问卷条目并将它们提交给现场协调员进行交叉编辑。在SPSS数据输入过程中，使用视觉和程序控制更正了错误。在分析阶段，对适当的数据进行了逻辑推断的后续编辑。由4名专家以SPSS格式输入数据。每份问卷的回答由2名专家独立输入以交叉检查跳过并防止机械错误。前一千条和最后的五千条输入由Mathematica和MCA-Armenia审查，他们比较了数据输入与硬拷贝问卷，并就数据输入过程提供了反馈。这些数据被传输到Mathematica进行分析。

数据处理:

在收到数据后，Mathematica合并了FPS3和FPS1数据。在分析数据时，Mathematica识别了几种农田家庭的不准确记录。这些农民是根据他们在基线和后续调查中报告的收获和销售量系统地识别的。首先，Mathematica确定了农民报告的特定作物收获和销售量，从基线到后续调查的变化超过200吨。这确定了十四位种植大麦、葡萄、桃子、甜樱桃、土豆、红萝卜、鹰嘴豆和gramma的农民。这14位农民的收获和销售量没有伴随作物种植面积或收入的大幅变化。Mathematica得出结论，这些结果可能是异常值，并根据关于土地和作物收入的信息替换了异常值。对于这些14次收获中的许多次，这包括将报告的销售量作为该作物的收入处理。这可能是数据记录错误，因为FPS3将作物收入记录在作物收获量旁边。

第二种方法用于解决缺乏足够证据来确定报告的值是否准确的外界异常值。该方法是对收入、生产和种植土地面积在每种措施的第98百分位数进行系统性地封存，或者如果第98百分位数小于或等于零，则为该措施的第二高值。此过程还有助于去识别收入、生产或土地特别大的个人。

响应率
---------------------------
75%的响应率。

抽样误差估计
---------------------------
在水至市场评估报告中，使用回归框架估计了WtM培训计划的影响，并控制了基线措施。影响估计的标准误差使用Huber-White风格的“三明治”估计量在村庄集群水平上聚类。关键影响估计的标准误差提供在水至市场评估报告附录B中，该报告作为资源文件提供。

数据评估
---------------------------
被封存的变量被用来构建非响应权重，以调整对FPS3做出响应的家庭和未做出响应的家庭之间观察到的特征之间的差异。非响应权重使用水至市场评估报告附录A中描述的程序计算。构建这些权重的代码位于Stata程序“1_armenia_construct.do”中。这些材料作为外部资源提供。