Sampling Frames

AAIC has been in operation since 1983. The main area of AAIC's expertise is the construction of national and project level area frames (AFs) and the operation of agricultural and environmental surveys. These activities are carried out with the full participation of foreign governments and by design involve a considerable component of technology transfer. Established AAIC systems generate agricultural production statistics in operational settings with less than 5% error. Some important AAIC activities include:

  • Conducting data user requirement studies generally involving: Ministries of Agriculture, Environment and Natural Resources, Commerce, Economics and Planning, the Central Bank, commodity traders, farmer’s associations, farm suppliers, and other users. This activity is essential for proper survey design.
  • Constructing area frames (AF) using satellite imagery, topographic maps, and aerial photos. Adaptation of existing digital materials for the purposes of the AF is another option.
  • Constructing list frames and implementing multiple frame sampling (MFS). During this stage the survey timing and the sample approach for the list frames are defined.
  • Calculating sample allocation and sample selection using optimum design. This system works in principle as a nested analysis of variance. The design is based on probability sampling and it is robust enough to allow for optimal sample allocation, emphasizing cost reduction and control of nonsampling errors.
  • Designing questionnaires tailored to particular conditions and timing of the survey, type of sampling segment, and variables to be researched. Questionnaires are written in the local language and allow for easy computer data entry.
  • Training personnel provided by client governments for managerial and supervision posts, as well as providing workshops for field interviewers.
  • Defining logistic needs to be provided by client governments, including working facilities, computer needs, and timing of vehicles and field expenditures.
  • Conducting surveys in cooperation with client governments. Surveys are conducted for planting and harvesting times. AAIC also manages other specialized surveys such as crop yield and economic surveys.
  • Analyzing data and integrating survey results into a usable form as requested in the data user requirement study. This activity includes such items as definition of content and timing of basic agricultural production, specialized information, statistics, farmer’s bulletin, etc. The qualitative aspects of the data collected, summarized, stored, and analyzed are primary project objectives. Typically, an efficient data collection system evolves after several interactions of survey operations. Survey operations are fine tuned using estimators and variances. Improved methods include data collection, summarization, and storage.
  • Developing an agro-economic analysis of production and trends when requested by client governments.
  • Ensuring quality control of the project at every stage.
  • Integrating satellite imagery that allows small area estimates to be made with more accuracy than can be accomplished without satellite imagery.

Agricultural Assessments International Corporation designs and implements information systems to support effective agricultural and environmental decision making and planning. The type of data needed for the decision making process varies according to the requirements of each project and forms a basic set of objectives for information system development. While the specific decisions and data are project dependent, the following list represents some typical data requirements and program objectives:

  • crop characteristics planted area, harvested area, inputs, yields total production
  • tree crops number of trees, area planted, yield per tree
  • livestock total number of animals, breakdown by ages, cattle on feed, birds for laying
  • farm production expenditures costs of inputs, labor, fertilizer, seed etc.
  • socio-economic data health, food consumption, income, labor, gender data, nutrition
  • demographic data
  • environmental indicators water quality, wildlife habitat, timber volume, range carrying capacity, soil capacity
  • integration of environmental, and socioeconomic data with agricultural production data.
  • data summary levels national, state, and project level data

In case you've never heard of the terminology of frame sampling, it's a list of sampling units that cover an entire population. It gives access to a population to take a probability sample. It may seem like the name, "frame sampling," implies that a frame is only for sampling purposes, but that's not entirely true. It's needed for a census as well, which requires you to collect data for all units of the frame. When conducting an agricultural survey, it's essential to include a sampling frame to gain a proper agriculture census. A sampling frame is when you pick a target population and develop a sample frame around it. For example, if your target population is Fall vegetables, your sample frame for that population could be lettuce, kale, parsley, etc. When you pick a sample, it shouldn't be just some simple random sample. Knowing about the sampling design provides necessary context that helps the researchers create a proper estimate and come up with correct inferences. In the following paragraphs, we will be explaining why we only survey a small sample rather than the whole population, multiple frame sampling, and area frame sampling.

Why do we do Sample Framing? The whole idea of picking a sample from an entire population of crops is to help reduce the cost and amount of work that goes into surveying the agricultural group. Surveying samples can also help provide more accurate assessments of present and future agricultural production. The sampling frame must be up to date, complete, affordable, and easy to use. It should include key features of all units in the part of the population you are sampling. When you are picking a sample frame, you need to decide all the units that you are interested in studying. It's important to be as specific as possible with the sampling units you have chosen.

Multiple Frame Sampling: When you've gathered a survey sample, most of the time, they are chosen from one sample frame and covers all units in a target population. Multiple-frame (MF) sampling is a survey that uses two or more frames and independent samples that must be taken respectively from each frame. Despite being simple in concept, the Multiple-frame design is difficult to implement. Whenever data is collected, combined, and duplicates are removed, this is what's considered to be agricultural multiple frame sampling. Multiple frame design may be pretty complex, but having access to the list and area samples are the biggest advantage.

Area Frame Sampling: The goal of an area frame is to divide the entire land into sampling units and to select a sample of such segments. The typical process is to select the sample in stages. This approach avoids having to divide the whole population into segments. You'll want to consider an area frame sampling for general purpose sampling and designed for getting information about variables associated with the land (ie. crops, livestock, etc.). A few advantages for area frame samples are that it's permanent for 10 years or more. Also, the same segments employed in the area frame are used for many surveys and measure change accurately because of that. However, some disadvantages relate to how expensive it is to set up the system. Luckily, over a 10-year period, agricultural area frame sampling is more cost-effective than other sampling frames.

Area framing is made to collect data from the farmland and from households located inside the chosen segments. When contributing to agricultural data, you should take the first survey immediately after you've planted the seeds. When it's almost time to harvest your crops, the one who is employed in taking the census of a population will go back to the same fields and perform a crop cutting survey. This estimate yields total production.

The life of a farmer would be an extremely difficult one if it weren’t for agricultural frame sampling. Because of these methods, it’s a lot easier for farmers to keep up with their crops. For years, AAIC has specialized in constructing national and project level area frames. We believe that it's super important that we always keep track of how the production is progressing with our farmlands. With multiple frame sampling and area frame sampling, all the tools that we need to ensure that our farmers are kept up to speed with all of our crops each growing season are given to us.