1. Survey process and timetable |
1. Definition of survey objective and requirements 1.1. The first work related to the development of the FSS 2016 started in January 2015; 1.2. Consultations with key data users and experts of the agricultural industry were launched regarding the organisation of FSS 2016, as well as the information to be acquired in the survey and the necessary level of detail. The possibility of using information from administrative data sources for FSS 2016 was also analysed. Consultations on the list of FSS indicators and definitions were rounded up in May 2015; 1.3. The target population of the survey was defined in accordance with the survey objective. According to the set precision requirements, the necessary sample size was evaluated; 1.4. In order to promote the survey, informative material was prepared – a brochure providing information on the duration and methods of FSS 2016. The brochure was distributed among respondents in cooperation with MoA, RSS and LRATC. Moreover, rural consultants of LRATC placed information of the survey in the local newspapers. 2. Survey design 2.1. When starting the FSS 2016, the available resources were evaluated, and a detailed work and financial plan was developed. For the most important items of the plan, see Annex 3-1. Timetable FSS 2016 LV. Grant agreement for the action “Contribution of the European Union to farm structure survey 2016 pursuant to Council Regulation (EC) No 1166/2008 of the European Parliament and the Council” was signed 1 December 2015; 2.2. The FSS list of indicators was developed in accordance with Annex III to the Regulation No. 1166/2008 and recommendations of national experts. 2.3. The frame of holdings included in the FSS 2016 was arranged on the basis of information of the Statistical Farm Register. The SFR was developed by the CSB in 1999 and is updated on a regular basis. 2.4. To obtain the data within the framework of FSS 2016, a questionnaire was developed and approved by the Cabinet of Ministers on 1 December 2015. 2.5. For data collection purposes, the unified data collection system ISDAVS CASIS of the CSB was used. 3. Data collection 3.1. Holdings were selected for the FSS 2016 based on their economic size, type of farming and region. In the middle of 2016, SO calculations for the holdings in the SFR were made based on the information obtained from various statistical and administrative data sources: ADC Animal Register, RSS IACS information and also data on sown areas from the last Crop Survey. 3.2. On behalf of the Ministry of Agriculture, LRATC also took part in the FSS. 116 interviewers were engaged in the data collection process. 3.3. It was possible to fill in the FSS 2016 questionnaires on the Internet during the period from 22 August to 11 September. Holdings, the e-mail addresses of which were known, and which had used the RSS electronic area payment application system in 2016, had an opportunity to fill in the FSS 2016 questionnaire on the Internet. Altogether, web questionnaires were completed by 9.5 % of the total number of respondents. In order to reduce the respondent burden and workload of statistical staff, CAPI interviews were carried out simultaneously with the annual Crop Survey. Face-to-face interviews were conducted by LRATC interviewers during the period from 26 September to 5 December 2016. Interviewers were trained beforehand. LRATC interviewers obtained information from about 24.1 thousand or 80.5 % of agricultural holdings. Telephone interviews (CATI) were conducted along with the CAPI interviews during the period from 30 September to 20 December 2016. They took place in the CSB Telephone Interview Centre in Preiļi, where 9 interviewers obtained information on 10.0 % of holdings of the total number of respondents. This surveying method ensured the interviewing of smaller and inactive agricultural holdings, the telephone numbers of which were available. Training of telephone interviewers was conducted on 29 September 2016. 4. Data processing and validation Interviewers performed both data collection and data entry, as well as primary control, because data entry applications contain around 200 logical and mathematical controls. Mathematical and logical controls were developed in compliance with the requirements of the “Data Supplier Manual”. In order to obtain more precise information and facilitate further data processing, they were supplemented with other necessary controls. Once the survey was completed, repeated data control was carried out in the CSB central office, and data were re-verified at the level of holdings. If it was necessary, the employees responsible contacted the holders or managers to update the information. Afterwards, information was integrated from the administrative data sources. Also, these data were analysed at the level of holdings before including them in the database. 5. Data dissemination The preliminary results of FSS 2016 were disseminated in CSB press releases two times: on 27 April and 22 September. The final results were published on 29 December 2016. Microdata have been prepared and sent to Eurostat within the set time. |
2. The bodies involved and the share of responsibilities among bodies |
The CSB of Latvia was the main institution responsible for the organisation of the FSS 2016. Responsibilities of the CSB Agricultural Statistics Section:
- development of the survey methodology and questionnaire form in compliance with the Regulation (EC) No 1166/2008 of the European Parliament and of the Council, Commission Regulation (EU) 715/2014 and needs of the national data users;
- management of the data collection process;
- development of data input programme methodology and logical controls, programme testing;
- interviewer training;
- integration of administrative micro-data into the array data structure;
- control and analysis of FSS 2016 data;
- preparation of the FSS 2016 data set and the National Methodological Report anf the sending thereof to Eurostat;
- development of table layouts, preparation and publishing of press releases and dissemination of the survey results.
Mathematical Support Division:
- design of the FSS 2016 sample;
- determination of extrapolation factors and estimation of sampling errors.
Information Technologies Department:
- development of data input application for the face-to-face interviews (CAPI);
- development of data control application;
- arrangement of summary tables.
- development of data input application for the telephone interviews (CATI) and Internet application (CAWI),
CSB Telephone Interviews Centre:
- organisation and conduction of telephone interviews.
Information and Communication Department:
- responsible for the printing and dissemination of methodological materials of FSS 2016.
Technical Maintenance and Procurement Division
- technical support for the workplaces, communications;
- supply of methodological materials to the regional offices.
Survey staff of CSB
- Agricultural Statistics Section: 10 regular employees for data analysis, processing and publishing;
- Mathematical Support Division – 1 person;
- Information Technologies Department – 2 persons;
- CSB CATI Centre – 9 regular employees.
Survey staff of LRATC (institution responsible for obtaining survey data from 24.1 thsd agricultural holdings):
- 1 manager/coordinator;
- interviewers – 107 persons.
|
3. Serious deviations from the established timetable (if any) |
FSS 2016 was conducted in accordance with the schedule of the Council Regulation (EC) No 1166/2008 of the European Parliament and the Council. |
Annexes: 3-1. Timetable for FSS 2016 |
1. Source of data |
FSS 2016 in Latvia was carried out as a sample survey. The largest share of the information was obtained in the survey; however, in order to reduce the burden on respondents, administrative registers were also used as a source of information:
- ADC Animal register;
- ADC Organic Farming Statistics Information System;
- RSS IACS database.
|
2. (Sampling) frame |
The sample frame is created from agricultural holdings. The list of holdings included in FSS 2016 was created based on the SFR information. The sampling frame is a list frame. To update the SFR various data sources are used: information from regular statistical surveys and censuses, Statistical Business Register, State Land Cadastre, Population Register, ADC Animal Register and RSS IACS database. Before FSS 2016, agricultural holding SO was recalculated using annual Crop and Animal Survey results, IACS database, ADC Animal Register. Such updating gave an opportunity to find new holdings and add them to the SFR. |
3. Sampling design |
3.1 The sampling design |
Sampling design was made as fully probabilistic sampling. The sample of FSS was stratified random sample. |
3.2 The stratification variables |
Since the FSS 2016 sample was formed together with the samples of the Crop Survey and Animal Survey, it was not easy to carry out stratification, as it was necessary to ensure the quality and scope requirements of all three surveys. The sample of FSS 2016 may be theoretically divided into two groups: I. the full coverage strata; II. sampling strata. Group I includes holdings that corresponded to the following characteristics:
- the holding is included in the Crop Survey 2016 and/or Animal Survey 2016;
- the holding carries out organic farming;
- area of UAA is not less than 100 ha;
- total SO >= 4 000 EUR.
In 2016, FSS was conducted simultaneously with the annual Crop Survey 2016 and Animal Survey 2016. Samples of annual surveys were formed based on the FSS 2016 frame, therefore not only criteria and precision requirements of FSS 2016 were used, but also several characteristics of annual surveys, and group I of the FSS sample was stratified in line with the requirements of the annual surveys. In this group stratification has been carried out:
- according to all of the afore mentioned characteristics – the holding is included in the Crop Survey 2016 and Animal Survey 2016, it carries out organic farming, as well as the area of UAA and the SO value correspond with the previously mentioned information;
- by location of holding - 6 regions (NUTS 3 level): 1 – Riga, 2 – Pieriga, 3 - Vidzeme, 4 - Kurzeme, 5 - Zemgale, 6 – Latgale;
- by type of farming – crop, animal, mixed;
- by SO groups:
- 0 EUR;
- 0 < SO < 70 EUR;
- 70 ≤ SO < 100 EUR;
- 100 ≤ SO < 1 000 EUR;
- 1 000 ≤ SO < 1 500 EUR;
- 1500 ≤ SO < 2 000 EUR;
- 2 000 ≤ SO < 4 000 EUR;
- 4 000 ≤ SO < 8 000 EUR;
- 8 000 ≤ SO < 15 000 EUR;
- 15 000 ≤ SO < 25 000 EUR;
- 25 000 ≤ SO < 50 000 EUR;
- 50 000 ≤ SO < 100 000 EUR;
- 100 000 ≤ SO < 250 000 EUR;
- 250 000 ≤ SO < 500 000 EUR;
- 500 000 ≤ SO < 750 000 EUR;
- 750 000 ≤ SO < 1 000 000 EUR;
- SO ≥ 1 000 000 EUR.
- if any of the crop production indicators below exceeds the boundary value:
Indicator |
Boundary value |
Cereals, ha |
89 |
Potatoes, ha |
8 |
Oilseed crops, ha |
30 |
Vegetables, ha |
4 |
Forage crops, ha |
80 |
Meadows and pastures, ha |
200 |
Fruit and berry plantations, ha |
7 |
Dried pulses, ha |
10 |
- if any of the livestock indicators below exceeds the boundary value:
Indicator |
Data source |
Boundary value |
Number of cattle |
Animal Register |
100 |
Number of cows |
Animal Register |
59 |
Number of dairy cows |
Animal Register |
24 |
Number of other cows |
Animal Register |
240 |
Number of other cattle |
Animal Register |
2400 |
Number of sheep |
Animal Register |
100 |
Number of goats |
Animal Register |
18 |
Number of pigs |
Animal Register |
80 |
Number of breeding pigs |
Animal Register |
800 |
Number of other pigs |
Animal Register |
800 |
Number of poultry |
Animal Register |
100 |
Number of laying hens |
Animal Register |
200 |
Number of bees |
Animal Register |
270 |
Number of rabbits |
Animal Register |
170 |
Number of pigs |
Survey data |
80 |
Number of breeding pigs |
Survey data |
80 |
Number of other pigs |
Survey data |
80 |
Number of poultry |
Survey data |
100 |
Group II includes holdings that corresponded to the following characteristics:
- if the holding does not correspond to the characteristics of Group I, and the holding is included in the frame of both the Crop Survey and Animal Survey and does not have previously described FSS stratification, then the FSS stratification is formed according to a special crop farming list, holding specialisation group, location of the holding, as well as crop farming SO group and livestock breeding SO group;
- if the holding does not correspond to the characteristics of Group I, and the holding is included in the frame of the Animal Survey but is not in the frame of the Crop is formed according to a characteristic of the special Animal Survey list, holding specialisation group, location of the holding, as well as crop farming SO group and livestock breeding SO group;
- in other cases, stratification is formed according to biological characteristics, location of the holding, holding specialisation group, crop farming SO and livestock breeding SO, as well as the characteristic of the frame of the Crop Survey and the characteristics of the frame of the Animal Survey.
|
3.3 The full coverage strata |
The sampling ratio is 100 % for:
- all strata with population size of 10 units/holdings or fewer;
- all strata included in Group I. See description in 3.2 The stratification variables;
- all strata where the sample size by Optimal Allocation of Sample was equal to the population size of strata.
|
3.4 The method for the determination of the overall sample size |
The size of sample was decided in accordance with the precision requirements provided in the Regulation (EC) No. 1166/2008 and financial and organizational possibilities. The total sample size was 30 136 agricultural holdings. |
3.5 The method for the allocation of the overall sample size |
See description in Annex 3.1-3.5 The optimal allocation of sample description. |
3.6 Sampling across time |
The sampling over time is not applied. |
3.7 The software tool used in the sample selection |
Procedure for sample selection is self-made using R. |
3.8 Other relevant information, if any |
None. |
4. Use of administrative data sources |
4.1 Name, time reference and updating |
Within the framework of the FSS 2016, three state registers were used:
- ADC Animal Register, including Bovine register – number of livestock on 1 July 2016;
- ADC Organic Farming Statistics Information System – UAA of the holding, on which organic farming production methods are applied in 2016;
- RSS IACS databases – support for rural development during the last 3 years (2014-2016).
The holder of the Animal Register and Organic Farming Statistics Information System is the Agricultural Data Centre, the largest and most important institution under supervision of the Ministry of Agriculture. Information of Animal Register was used as a data source in FSS to provide data on cattle, sheep, goats, horses, rabbits and beehives, as well as the number of organic livestock by species. Legal basis and activities for Animal Register:
- Cabinet Regulation No 393 (Arrangement for the registration of livestock and aquaculture animals, their herds and sheds as well as procedure of animal labelling);
- Supervision programmes in Latvia;
- Directive of European Union 92/102/EEC on the identification and registration of animals;
- Council Regulation (EC) 820/97 establishing a system for the identification and registration of bovine animals and regarding the labelling of beef and beef products;
- Regulation (EC) No 1760/2000 of the European Parliament and of the Council of 17 July 2000 establishing a system for the identification and registration of bovine animals and regarding the labelling of beef and beef products.
In accordance with Section 4 of Cabinet Regulation No 393, the owner is obliged to register and designate livestock, and update information on their designation, transfer, and other events. The herd owner shall within 7-day time submit an application electronically or in paper form, informing about any changes. Data on the area of organically grown crops were obtained from the Organic Farming Statistics Information System. Legal basis and activities for Organic Farming Statistics Information System:
- Cabinet Regulation No 485 “Procedure for the Supervision and Control of Organic Farming”;
- The Ministry of Agriculture shall prepare statistical information in line with the requirements of Regulation No 588/2008 and send this information to Eurostat by 1 July of every year.
In accordance with Cabinet Regulation No 485 “Procedure for the Supervision and Control of Organic Farming” of 26 May 2009, holdings shall submit the following information until 1 February of every year to a control body, in which they have filed a submission regarding inclusion in the control system of organic farming: area, grown crops, production of plant and animal origin, type and amount of production. The control body shall submit individual data to the Agricultural Data Centre to be included in the organic farming statistical information system annually by 30 April. For the provision of information on Rural development support measures during the last three years needed for FSS 2016, RSS IACS information at the customer level was used as the source of information. The database is updated once a year, when the land owner/user submits an application for the single area payment. Legal basis and Activities for RSS:
- Law on Rural Support Service, 04/28/2004;
- Cabinet Regulation No 876 "Rural Support Service Regulations" of 19 October 2004;
- Agriculture and Rural Development Law, 24.04.2004.
|
4.2 Organisational setting on the use of administrative sources |
The Statistics Law of the Republic of Latvia stipulates that in order to implement the State Statistical programme the CSB has the right to receive the necessary information from the state registers or databases free of charge, including individual data on natural persons. In compliance with Article 4 of the European Parliament and Council Regulation No 1166/2008, in FSS surveys Member States shall use information from the Integrated Administration and Control System (IACS) provided for in Regulation (EC) No 782/2003, the System for the Identification and Registration of Bovine Animals provided for in Regulation (EC) No 1760/2000, and the Organic farming register provided for in Regulation (EC) No 834/2007. Bovine Register is part of the Animal Register in Latvia. Representatives of CSB participate in meetings concerning the content of administrative data sources and emphasize needs of statistics, especialy in the context of reducing the administrative burden on the respondents. The CSB has developed constructive cooperation and, as much as possible, takes into account and implements the statistical needs for providing information. |
4.3 The purpose of the use of administrative sources - link to the file |
Please access the information in the file at the link: (link available as soon as possible) |
4.4 Quality assessment of the administrative sources |
|
Method |
Shortcoming detected |
Measure taken |
- coherence of the reporting unit (holding) |
|
Different definitions of reporting units in the Animal Register, the responding unit is the owner of the livestock herd, in IACS – a natural or legal person who receives support payments, in Organic Farming Register – a natural or legal person who has received an organic farming certificate. There may be several livestock herd owners/holders in one statistical holding. In one FSS holding there may be several persons certified for organic farming. |
To link administrative data sources with the SFR information, personal identity number of the herd owner or enterprise registration number in the Business Register is used. For more information, see item 4.6 below. |
- coherence of definitions of characteristics |
Animal Register: Information on characteristics was integrated in the FSS 2016 database directly from the Animal Register. The register contains information on number of livestock by species and category. Breakdown of the cattle into categories meets the FSS definitions. |
|
|
- coverage: |
|
|
|
|
over-coverage |
|
Not found |
|
|
under-coverage |
|
Not found |
|
|
misclassification |
|
Not found |
|
|
multiple listings |
Each unit in administrative data source has its own ID number and multiple listings are not possible |
|
|
- missing data |
|
Not found |
|
- errors in data |
|
Not found |
|
- processing errors |
|
Not found |
|
- comparability |
|
There were no other sources of data on the number of animals, organic farming, as well as support for rural development |
|
- other (if any) |
|
Organic Farming Register: data availability - information on organic farming in 2016 was received on 1 July 2017, which extended the FSS data processing time. |
|
4.5 Management of metadata |
The exchange of data with holders of administrative data is carried out by using the servers of CSB. The responsible employee of the Administrative Data Processing Section of the IT Department places the data received in the data folder “Catalogue of Administrative Resources”, which may be accessed by specialists of the respective field who use the data for statistical purposes. |
4.6 Reporting units and matching procedures |
Reporting units in Animal register, in Organic Farming register and in the RSS IACS database are mutually different as well as different from the FSS reporting unit.
- In Animal register, reporting unit is the livestock herd holder. In one agricultural holding there may be several livestock herd owners/holders. The CSB received individual data from the Animal Register as well as information on the herd owner: name, identity code, and address of residence, address of animal stall, telephone number etc. To combine the data the identity code for physical persons or registration number for legal holdings served as a common identifier. Information on characteristics was integrated in the FSS 2016 database directly from the Animal Register using common identifier.
- In Organic Farming Statistics Information System, responding unit is a natural or legal person who is awarded a certificate of organic farming by any of the certification bodies. In one FSS holding there may be several persons certified for organic farming. The CSB received individual data from the Organic Farming Register as well as information on the certified person: name, identity code and address of residence, number of Organic farming certificate, telephone number etc. To combine the data the identity code for physical persons or registration number for legal holdings served as a common identifier. Information was integrated in the Statistical Farm register. Statistical Farm Register contains the information about agricultural holdings and persons, who lives in each household. There were identified all certified natural and legal persons and added to agricultural holdings.
- The unit registered in the RSS IACS database is client – natural or legal person eligible to apply for the support within the framework of the activities organised by the RSS. Each client has a unique RSS client registration number. The information received by the CSB included identity code, client registration number, address of residence, address of person, telephone number etc. For combining data, personal code for physical persons or registration number for legal holdings is used as a common identifier.
|
4.7 Difficulties using additional administrative sources not currently used |
In order to reduce the burden and costs of AC 2020, possibilities for using new administrative data sources are searched. After summing up the results of FSS 2016, work will be continued on the assessment of the information available in the Tractor Machinery Register of STCA. The initial results of data analysis indicate that the Tractor Machinery Register does not have references to the use of machinery in agriculture. Hereafter, it will be assessed, whether it could be possible to acquire partial data on those employed in agriculture from the SRS databases. |
Annexes: 3.1-3.5 The optimal allocation of sample description |
1. Data collection modes |
The main method used to obtain the data in the FSS 2016 was interviews; however, also other data collection methods were used: in the FSS 2016 three data collection methods were used – face-to-face interviews (CAPI), telephone interviews (CATI) and web survey (CAWI). CAPI interviews: The LRATC interviewers used their laptop computers and data input programme developed by the CSB. During the time period from 26 September until the middle of December 24.1 thousand agricultural holdings or 80.5 % were surveyed. CAWI interviews: Holdings, whose e-mail addresses were known and which in 2016 used RSS electronic area payment application system, had an opportunity to fill in the FSS 2016 on the Internet. Altogether 9.5 % of the total number of respondents completed the web questionnaires. CATI interviews: Telephone interviews were conducted along with the CAPI interviews during the time between 30 September to 20 December 2016. They took place in the CSB Telephone Interviews Centre in Preiļi, where 9 interviewers obtained information on 10.0 % holdings of the total number of respondents. |
2. Data entry modes |
CASIS (Computer Assisted Statistical Information System) was used. There were 3 different types of application:
- application for face-to face interviews – CAPI. The application was installed on the interviewers’ laptops;
- application for telephone interviews – CATI;
- application for fulfilling questionnaire on the Internet – CAWI.
|
3. Measures taken to increase response rates |
A series of activities were introduced to reduce the non-response rate. It was important to inform all the respondents (holdings) in due time on the planned FSS 2016, its aims, the acquired results and the significance thereof. Every respondent of FSS 2016 was sent a letter signed by the Vice President of the CSB, containing information on the survey and the possibilities for providing the required data. The respondents had access to a toll-free phone line over the whole period of the survey. There was an informative brochure prepared, providing information on the FSS, its aims and cooperation partners of the CSB taking part in the realisation of the survey. The brochure was distributed before the survey via RSS and LRATC. When starting the survey, the CSB published a press release containing information on the course of the survey, its target population and methods for submitting data. The respective information was republished by regional press, as well as republican publications linked to agriculture. As the face-to-face interviews began, rural consultants of LRATC informed farmers on the start of the survey through regional press. In order to reduce FSS 2016 non-response, the information was specified by telephone. Mainly respondents not met during data collection process were contacted. Information obtained by telephone was entered into the database. With an aim to specify the information, during the data verification process, respondents received a phone call:
- if questionnaire was filled in only partly;
- if the given data were inaccurate or significantly differed from the information available in other sources.
|
4. Monitoring of response and non-response |
1 |
Number of holdings in the survey frame plus possible (new) holdings added afterwards In case of a census 1=3+4+5 |
71 006 |
2 |
Number of holdings in the gross sample plus possible (new) holdings added to the sample Only for sample survey, in which case 2=3+4+5 |
30 136 |
3 |
Number of ineligible holdings |
244 |
3.1 |
Number of ineligible holdings with ceased activities This item is a subset of 3. |
244 |
4 |
Number of holdings with unknown eligibility status 4>4.1+4.2 |
- |
4.1 |
Number of holdings with unknown eligibility status – re-weighted |
- |
4.2 |
Number of holdings with unknown eligibility status – imputed |
- |
5 |
Number of eligible holdings 5=5.1+5.2 |
29 892 |
5.1 |
Number of eligible non-responding holdings 5.1>=5.1.1+5.1.2 |
1 483 |
5.1.1 |
Number of eligible non-responding holdings – re-weighted |
979 |
5.1.2 |
Number of eligible non-responding holdings – imputed |
504 |
5.2 |
Number of eligible responding holdings |
28 409 |
6 |
Number of the records in the dataset 6=5.2+5.1.2+4.2 |
28 913 |
5. Questionnaire(s) - in annex |
For all types of holdings Lativa has only one type of electronic form. The questionnaire form of FSS 2016 was developed in co-operation with the Ministry of Agriculture and other State institution concerned. The questionnaire form included variables, regarding which it is not possible to attain information from administrative data sources. The questionnaire included the information on agricultural holding: identification number, name of holder and holding, the address of agricultural holding, correspondence address and contact information. The main part of the questionnaire included questions grouped into 8 major sections:
- General characteristics of the agricultural holding;
- Land use;
- Utilisation of arable land;
- Number of livestock (on 1 July 2016);
- Agricultural machinery owned by the holding (on 1 July 2016);
- Farm labour force employed permanently and temporarily (number during the previous 12 months);
- Other gainful activities;
- Agricultural production methods.
|
Annexes: 3.3-5. FSS 2016 questionnaire form |
Methodology for determination of weights (extrapolation factors) |
1. Design weights |
The design weights (basic weights) are calculated according to the sample design. The design weights are calculated as the ratio of the number of holdings in the population to the number of holdings in the sample within each stratum. Unit design weights are calculated according to sampling design and inclusion probabilities of units in the sample: where Nh is population size of stratum h nh is the sample size in stratum h |
2. Adjustment of weights for non-response |
Taking into account the response rate, the holdings are broken down by the ones that have responded and the ones that have not. If information is obtained or imputed on farms and it is in over-coverage, the farms are considered as the ones that have responded; however, in other cases they will be assumed as the ones that have not responded. where nh - sample size in stratum; nRn number of units responded in stratum. Taking into account the response rate, end weights are: w2 = wd * k, wd - design weights. If no farm has responded in a stratum, similar strata are combined and weights are recalculated in stratum. |
3. Adjustment of weights to external data sources |
The GREG estimator is used for estimation of totals. Calibration: Weight calibration was carried out by the number of sheep, pigs, dairy cows, other cattle, poultry, rabbits, bees in each region as of 1 July 2013 and by sown area of crops of IACS – spring rape, winter rape, oats, spring wheat – in each region and total number of farms in each region from frame of FSS. Package “sampling” of software R is used for the calibration, and g-weights are calculated with the help of function “calib” from this package. In turn, calibration is based on the raking method in the function “calib”. Please note that when using the GREG (Generalized Regression) estimator (calibration) for FSS, the weights are not equivalent within one stratum. The GREG estimator is used for the estimation of totals. More on the GREG estimator can found in the following literature:
- “Estimation in Surveys with Nonresponse Carl-Erik Särndal/ SixtenLundström, Wiley”;
- “Estimation in the presence of nonresponse and frame imperfections SixtenLundström/ Carl-Erik Särndal, Statistics Sweden”.
|
4. Any other applied adjustment of weights |
Not applied. |
|