4Design methodology

50 %
50 %
Information about 4Design methodology

Published on December 29, 2007

Author: Eagle

Source: authorstream.com

Overall Objectives of the Livestock Breed Survey in Oromia Regional State:  Overall Objectives of the Livestock Breed Survey in Oromia Regional State To describe, identify and classify the indigenous livestock genetic resources in the region and to obtain reliable estimates of population size and distribution. In particular: To describe what breeds or types of animals exist, in what numbers and where they are. To describe what they look like. To define the environments in which different breeds are raised in terms of agro-ecological zone, disease, etc. To say for what purposes they are used, how they are bred and by which farmers. To determine farmers’ opinions on the main attributes of different breeds, in particular in terms of their adaptation to heat, drought and disease tolerance. and so on ……………….. To develop recommendations for utilisation of the livestock resources in Oromia Region. Not an easy task!:  Not an easy task! Such a survey requires careful planning to ensure that these objectives can be met. At least 6 months should be set aside for planning a survey in a region. This includes: seeking cooperation of partners agreeing on objectives for the survey planning how the survey will be executed and by whom finding out what ancillary (census) statistics may be available on household numbers, people and livestock. organising a sampling frame selecting zones, woredas and peasant associations to sample organising for training of supervisors and enumerators making sure that everyone is clear on final arrangements planning for data entry and analysis and making arrangements with those to be responsible ………… and so on. Types of sampling:  Types of sampling cluster representative stratified convenience random purposive Cluster Sampling :  Cluster Sampling When sampling a large area it is easier in terms of survey costs and survey administration for the area to be first divided into clusters. To do so use can be made of the administrative structure in a country. In Ethiopia this is region zone woreda peasant association. Sampling units can be selected at each administrative layer in turn. Slide5:  Cluster sampling Stratified sampling:  Stratified sampling The sampling units within a given administrative layer may vary in relation to a particular characteristic, e.g. agro-ecological zone, livestock density and household size. A completely random sample may miss woredas from a certain agro-ecological zone or a particular livestock density. It may also not adequately represent the population of households in a village. By efficiently stratifying according to agro-ecological zone, livestock density or household size a more precise estimate of the number of livestock in a population can be achieved. Slide7:  Stratified sampling Units selected Strata Sampling frame:  Sampling frame Definition The entire list of zones, woredas, peasant associations (P.A.) and households in a region. How large should a survey be and how many zones, woredas, P.A.s and households should be sampled? This depends on funds available, costs of organising the survey, available manpower, administrative support, means of transport and ease of access to villages. Proportionally more units should be sampled at the upper than the lower layers. In the survey in Oromia Region all zones, approximately 30% of woredas, 17% of P.A.s per woreda and 4% of households per P.A. were sampled. This meant that approximately 1 in 500 households were sampled in Oromia Region. Slide9:  Survey design and sampling frame The Oromia Region covers a third of Ethiopia Slide10:  Survey design and sampling frame (cont.) Stratification by: agro-ecological zone livestock density Stratification by: numbers of livestock types of species Stratification by: agro-ecological zone Slide11:  About 17% of P.A.s in each selected woreda sampled. Thirty households sampled within each selected PA. About 0.2%of all households (5558) sampled in the region. Survey design and sampling frame (cont.) Selected woredas (55 in total) Sampling About 30% of woredas in each zone sampled. Slide14:  Selection of woredas of East Wellega by agro - ecological zone and livestock dens ity Selection of woredas in East Wellaga zone Summary:  Selection of woredas in East Wellaga zone Summary Five woredas selected, one from each of Dega, Dega/Woinadega, Dega/Woinadega/Kolla, Woinadega and Woinadega/Kolla zones. One woreda selected from areas of very high (>180), two from areas of high (70-85) and two from areas of medium livestock density (35-55 livestock per km2). Representative sampling used to ensure balance in selection of woredas across agro-ecological zone and livestock density strata. No element of randomisation. Selection of P.A.s:  Selection of P.A.s In woredas covering different agro-ecological zones ( e.g. Woinadega/Dega) P.A.s were randomly sampled from within each agro-ecological zone. In woredas situated entirely with one agro-ecological zone P.A.s were randomly selected from all P.A.s in the woreda. Selection of households:  Selection of households Households selected as far as possible at random to ensure coverage of households with low, medium and high numbers of livestock. Ten households selected in turn for cattle, sheep and goats as the primary species. All households in each sampled P.A. subsequently categorised into low, medium and high numbers of livestock for each species in turn in order to facilitate estimation of the total numbers of cattle, sheep and goats in the P.A. Sample selection of households by size category for cattle Haro P.A. in Limu Woreda:  Sample selection of households by size category for cattle Haro P.A. in Limu Woreda a Approximately proportional to stratum size but with extra households in more variable ‘high’ group. Methods of sampling (1) :  Methods of sampling (1) Random sampling Samples drawn completely at random, each with an equal chance of being selected. This method generally applied in selection of households in a P.A. This method also used to select P.A.s from selected woredas stratified by agro-ecological zone. Only method that allows unbiased estimation of population numbers. Slide21:  Methods of sampling (2) Representative sampling Samples selected to be representative of population. This method applied at the zone layer. Method makes the estimation of overall population numbers in the zone more difficult unless additional ancillary data are available. Methods of sampling (3) :  Methods of sampling (3) Convenience sampling Samples selected to allow, for example, ease of access to a P.A. or household, or to make best use of available manpower. Occasional use inevitable in such a survey. Document when method used. Provided such cases are few, probably reasonable to assume randomness for the purpose of estimation of population numbers. Slide23:  Methods of sampling (4) Purposive sampling Sampling based, for example, on knowledge of known farming system or known breed unique to a certain location. Document when method used so that suitability of sample for inclusion in calculation of population estimates can be judged. Summary:  Summary Ensure sufficient replication at the upper administrative layers to allow efficient estimates of population numbers and breed distributions to be determined. But match this requirement with knowledge of available manpower, resources (transport, etc), adequacy of infrastructure (administration, roads, etc), quality of field staff and size of budget. A few recommendations for successful survey design and implementation in Ethiopia. Slide25:  Determine what region / zone / woreda / P.A. statistics are already available on numbers of households, people or livestock that may help to improve the efficiency of sample selection and population estimation. Use this information to define appropriate strata for the sampling frame. Ensure that selections of sampling units at the woreda, P.A. and household levels at least are as far as possible at random. Above all, obtain high quality data from a manageable sample of households. Summary (continued)

Add a comment

Related presentations

Related pages

CHAPTER 4 Research Methodology and Design - UnisaIR Home

Chapter 4: Research methodology and design 292 4.2 Research Paradigm According to TerreBlanche and Durrheim (1999), the research process has three
Read more


CHAPTER 4 DESIGN METHODOLOGY AND DATA ANALYSIS This chapter describes the underlying paradigms, methods of the research design, the interpretive
Read more

Design Methodologies

Design Methodologies Instructional, Thinking, Agile, System, or X Problem? Design brings forth what does not come naturally.
Read more

Chapter 5: Design Methodology - Plastics Pipe Institute ...

CHAPTER 5: DESIGN METHODOLOGY Overview of Structural Considerations All pipe, whether flexible or rigid, relies on the backfill structure to transfer loads
Read more

UltraFast Design Methodology Guide for the Vivado Design ...

UltraFast Design Methodology Guide www.xilinx.com 2 UG949 (v2015.3) November 23, 2015 Revision History The following table shows the revision history for ...
Read more

Design and Methodology - Census.gov

Design and Methodology U.S. Department of Commerce Carlos M. Gutierrez, Secretary David A. Sampson, Deputy Secretary Economics and Statistics Administration
Read more

OPUS 4 | Design Methodology for Digital Integrated ...

Design Methodology for Digital Integrated Circuits in OFET Technology Design-Methodik für digitale integrierte Schaltungen in OFET Technologie
Read more

Towards a Framework and a Design Methodology for Autonomic SoC

4. Design Methodology for Autonomic SoC The methodology has to consider fault tolerance as an additional parameter beside, area, performance and power
Read more

Top 4 Design Methodology Team Project profiles | LinkedIn

Here are the top 4 Design Methodology Team Project profiles on LinkedIn. Get all the articles, experts, jobs, and insights you need.
Read more