3. Statistical processing |
Top |
|
1.Survey process and timetable |
The main activities are shown in the table:
Key activities |
Timetable |
Formation of the Working Group responsible for the organisation and implementation of the action |
August 2015 |
Analysis of the Eurostat’s methodological requirements |
August 2015 |
Creation of a methodology of the Farm Structure Survey 2016 |
October 2015 |
Drafting a survey questionnaire for agricultural companies and enterprises and for farmers’ and family farms and instructions for filling them in |
December 2015 |
Drafting instructions for filling in the survey questionnaire |
January 2016 |
Preparation of specifications and other documentation for the public procurement procedure (task for subcontractors – data collection) |
February 2016 |
Approval of the questionnaire and instructions for filling it in and publication in the Official Gazette |
February 2016 |
Preparation of the task for primary data entry and creation of a logical control program |
March 2016 |
Drawing of a survey sample |
March 2016 |
Public procurement procedure (selection of subcontractors) |
May 2016 |
Preparation of a program for the entry of primary data on farmers’ and family farms with arithmetical and logical control implemented (using ORBEON software) |
May 2016 |
Training of the interviewers |
June 2016 |
E-survey for farmers’ and family farms |
June 2016 |
Preparation of a program for the entry of primary data on agricultural companies and enterprises with arithmetical and logical control implemented (using ABBYY eFormFiller 2.5 software) |
July 2016 |
Taking data from administrative data sources and prefilling questionnaires with them |
August 2016 |
Preparation of a program for data processing, as well as for the implementation of arithmetical and logical control (using ORACLE software) |
August 2016 |
Interviewing farmers’ and family farms – face-to-face interview |
August 2016 |
Interviewing farmers’ and family farms – telephone interviews |
October 2016 |
Collection and checking of data from agricultural companies and enterprises |
November 2016 |
Creation of a survey database |
December 2016 |
Data processing, analysis and editing |
August 2017 |
Performing calculations and evaluation of the survey results |
October 2017 |
Preparation of a primary data transfer program to Eurostat |
October 2017 |
Data preparation for transmission to Eurostat |
November 2017 |
Transmission of survey data to Eurostat, preparation of the National Methodological Report and its delivery to Eurostat |
December 2017 |
Preparation of a special publication containing the final results of the survey |
May 2018 |
|
2. The bodies involved and the share of responsibilities among bodies |
Statistics Lithuania was responsible for the FSS 2016 preparation, organisation, data checking, data analysis, calculations, preparation of the survey results, and transmission to Eurostat and publication thereof. These works as well as all other methodological works were carried out by the employees of Statistics Lithuania. The market research company selected during the public procurement procedure was responsible for the data collection from farmers’ and family farms. The interviewers were trained by the employees of Statistics Lithuania responsible for the FSS 2016. |
3. Serious deviations from the established timetable (if any) |
There were no serious deviations from the established calendar in Lithuania during the FSS 2016. |
|
3.1. Source data |
1. Source of data |
The data is collected by means of a sample survey (with an exhaustive coverage of agricultural companies). Administrative sources are used for some characteristics, for the units in the sample, either by taking directly the data or by prefilling the questionnaires. |
2. (Sampling) frame |
In order to have accurate information about all agricultural holdings, such administrative sources were used: IACS Crop Declaration Database, Animal Register, Address Register, Population Register. Also, statistical registers were used: Statistical Business Register and Statistical Farm Register. All population of agricultural holdings was divided into two lists – list with farmers' and family farms and list with agricultural companies and enterprises. The sampling frame was a list frame of farmers' and family farms. The quality of this list was checked and some corrections have been done after comparison with the Population Register and Address Register. |
3. Sampling design |
3.1 The sampling design |
One-stage stratified random sampling design of holdings has been used for farmers‘ and family farms. A census has been used for agricultural companies and enterprises. |
3.2 The stratification variables |
Agricultural holdings in the sample had been stratified by standard output of farm and municipalities. The full scale survey was carried out for agricultural companies and enterprises. There were 6 strata of standard output and 52 strata of municipalities for farmers' and family farms. The total number of strata was 310. Farmers' and family farms were stratified into 309 strata (including the strata with probability equal 1), for all agricultural companies and enterprises one strata (strata 310) was composed (whereas all companies and enterprises were selected, strata 310 was with probability 1). |
3.3 The full coverage strata |
The following agricultural holdings had been selected with probability equal 1:
- agricultural companies and enterprises,
- organic farms,
- holdings with special crops (e.g. nut trees, nurseries, perennial plants for twining, weaving, energy purposes, flax and etc.),
- holdings growing ostriches,
- holdings with standard output of more than EUR 10 000.
|
3.4 The method for the determination of the overall sample size |
The size of the sample was determined in accordance with the precision requirements provided in Regulation 1166/2008. The number of agricultural holdings in the population and in the sample by strata is provided in the following table:
|
Strata |
Frame population |
Sample |
Lithuania |
|
185 967 |
45 928 |
Farmers’ and family farms |
|
185 047 |
45 008 |
|
1 |
73 196 |
1 710 |
|
2 |
43 014 |
4 104 |
|
3 |
25 626 |
5 441 |
|
4 |
16 016 |
6 558 |
organic farms |
5 |
2 470 |
2 470 |
farms with standard output of more than EUR 10000 and specific farms |
6 |
24 725 |
24 725 |
Agricultural companies and enterprises |
|
920 |
920 |
|
3.5 The method for the allocation of the overall sample size |
Neyman allocation method had been used to allocate units across strata of standard output and proportional allocation method had been used to allocate units across strata of municipalities. |
3.6 Sampling across time |
A new sample is drawn in each occasion. |
3.7 The software tool used in the sample selection |
The SAS software was used to select the sample. |
3.8 Other relevant information, if any |
Please find attached information on sampling methods in annex 3.1.-3.8. FSS 2016 sampling methods. |
4. Use of administrative data sources |
4.1 Name, time reference and updating |
Organic Farming Register: During the FSS 2016, organic farming characteristics were taken from the Organic Farming Register directly (without questioning agricultural holdings). A public institution Ekoagros performs certification of organic farms. The Organic Farming Register is updated each year. Reference day – 31 December 2016. The legal basis of the Organic Farming Register is as follows:
- Council Regulation (EC) No 834/2007 on organic production and labelling of organic products;
- Commission Regulation (EC) No 889/2008 of 5 September 2008 laying down detailed rules for the implementation of Council Regulation (EC) No 834/2007 on organic production and labelling of organic products with regard to organic production, labelling and control;
Integrated Administration and Control System (IACS) Crop Declaration Database: IACS Crop Declaration Database consists of agricultural holdings (both natural and legal persons) which declare their UAA. This database is maintained by a State Enterprise Agricultural Information and Rural Business Centre. IACS Crop Declaration Database is updated each year. During the FSS 2016, data from IACS Crop Declaration Database were used for prefilling the FSS 2016 questionnaires for agricultural companies and enterprises and for direct entering the data for farmers’ and family farms (without direct questioning), which have declared their UAA in 2016. Reference year – 2016. The legal basis of the IACS is as follows:
- Regulation (EU) No 1306/2013 of the European Parliament and of the Council of 17 December 2013 on the financing, management and monitoring of the common agricultural policy and repealing Council Regulations (EEC) No 352/78, (EC) No 165/94, (EC) No 2799/98, (EC) No 814/2000, (EC) No 1290/2005 and (EC) No 485/2008;
- Commission delegated regulation (EU) No 640/2014 of 11 March 2014 supplementing Regulation (EU) No 1306/2013 of the European Parliament and of the Council with regard to the integrated administration and control system and conditions for refusal or withdrawal of payments and administrative penalties applicable to direct payments, rural development support and cross compliance;
- Commission implementing regulation (EU) No 809/2014 of 17 July 2014 laying down rules for the application of Regulation (EU) No 1306/2013 of the European Parliament and of the Council with regard to the integrated administration and control system, rural development measures and cross compliance.
Animal Register (Bovine Register): This Register also is part of IACS system and is created for identification and registration of farm animals. Legal basis for Animal Register is the same as for IACS Crop Declaration Database. Order of registration of animals in this Register depends on animal species. Animal Register is updated every day. Reference day – 1 June 2016. National Paying Agency Database (Rural Development Measures database): The National Paying Agency is the only accredited institution managing the measures of support for agriculture, rural development and fisheries. Each year, farm holders willing to receive a support according to Rural Development Program and modernise their agricultural holdings submit applications, which are carefully checked and approved. Legal basis for Rural Development Measures is Regulation No 1305/2013 of the European Parliament and of the Council of 17 December 2013 on support for rural development by the European Agricultural Fund for Rural Development (EAFRD) and repealing Council Regulation (EC) No 1698/2005 A list of applicants for the FSS 2016 needs from the National Paying Agency was received following special request. Reference period – 2014-2016. |
4.2 Organisational setting on the use of administrative sources |
Data from administrative sources are obtained on the basis of data provision agreements or following special requests. Statistics Lithuania has sign the agreements with managers of administrative data sources indicating what data will be available, for what purpose they will be used, in which frequency, in what way (direct connection to register data or another way), data confidentiality assurance. In case of special request, Statistics Lithuania applies to the institution which is manager of certain data source in an official letter. Statistics Lithuania does not participate to the conceptual design (characteristics, definitions, classifications, formats, etc.) and subsequent related revisions of the administrative sources. |
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) |
In the IACS Crop Declaration Database, Animal Register and National Paying Agency Database, all holdings have the unique holding identifier (code of the holding). The unique holdings identifiers are set up in the Agriculture and Rural Business Register (holdings register). Therefore, data of all above mentioned administrative data sources can be linked to each other using this identifier. Statistics Lithuania also uses this identifier for linking data. Only those agricultural holdings, which are not registered in the Agriculture and Rural Business Register, do not have unique holding identifier. In the Organic Farming Register, all holdings also have the unique holding identifier (code of the holding). Therefore, holdings from this register can be linked to other administrative data sources as well as to statistical data. |
However, the Animal Register could be mentioned as a small exception. In this register, farm animals are registered to their owner (personal code of the owner is used), instead of agricultural holding. However, personal codes of owners of farm animals can be linked to the agricultural holding (to the certain unique holding identifier) and the agricultural holding to which certain owner of farm animals belongs could be determined. |
An additional step in the preparation of Animal Register data for the FSS 2016 needs was done. The owners of farm animals were attributed to certain agricultural holding using their personal codes and unique holding identifier. |
- coherence of definitions of characteristics |
|
Organic Farming Register: The reference period for crops in the Organic Farming Register is the same as it was required in the FSS 2016. In the Organic Farming Register, animals of the holdings are certified during the period from May to August. The exact date is different for different holdings and depends on the date when the certification body staff has visited the farm for certification. According to the methodology of the FSS 2016, farm animals had to be counted as they were on 1 June 2016. |
Organic Farming Register: Comparisons of the data and adjustments were done using another administrative source – the Animal Register. |
|
Organic Farming Register: As direct insert from this source to the survey database is done, there should be precise classification of required characteristics. Sometimes it is a problem to classify detailed distribution of the organic crops into variables required in the FSS. |
Organic Farming Register: Variables which have the same definition were used. |
|
IACS Crop Declaration Database: Areas of outdoor vegetables, strawberries, potatoes, flowers and ornamental plants, greenhouses (vegetables and flowers and ornamental plants), mushrooms, pome and stone fruits, berries, walnuts, nurseries (especially small) can be declared together as one area "other crops". |
IACS Crop Declaration Database: These areas were collected from agricultural holdings during the data collection. |
|
Animal Register: According to the legislation, farmers have an obligation to render information about farm animals which are registered by herds (pigs, poultry, rabbits, etc.) to the Animal Register at least 2 times per year. Exception is only for pigs. Herds of pigs should be registered at least once per quarter. Therefore, on 1 June 2016 exact number of these farm animals was not known. |
Animal Register: Questionnaires were prefilled with available information from the Animal Register and farmers had the possibility to correct figures if prefilled figures were incorrect. |
- coverage: |
|
|
|
|
over-coverage |
|
IACS Crop Declaration Database: Some farms declare crops, but in fact land is leased and this land is cultivated by another farm. |
IACS Crop Declaration Database: If during the FSS 2016 it was determined that there were out-of-scope units, they were excluded. |
|
under-coverage |
|
Animal Register: It covers fewer units than it should. Farms with small number of pigs, poultry, rabbits do not register their farm animals. |
Animal Register: During the FSS 2016 missing information is being collected. |
|
IACS Crop Declaration Database: Only farms which are willing to receive support declare their crop areas. Therefore this data source covers fewer units than it should. Farms which have UAA, but do not declare it, are not included in this database. |
IACS Crop Declaration Database: Direct questioning of farmers' and family farms which do not declare their UAA was carried out. |
|
misclassification |
|
There were no misclassification errors in administrative data sources used for the FSS 2016 needs. |
|
|
multiple listings |
|
Duplicates were not observed in administrative data sources used for the FSS 2016 needs. |
|
- missing data |
|
Not found. |
|
- errors in data |
|
Animal Register: Some holders who according to the Register had farm animals, actually never had. |
Animal Register: Holders had the possibility to correct their data or data were corrected by interviewers. |
- processing errors |
|
Not found. |
|
- comparability |
|
There were no other sources of data on organic farming, declared UAA, support for rural development. |
|
- other (if any) |
|
Not found. |
|
4.5 Management of metadata |
Administrative metadata are stored and maintained in dedicated databases. |
4.6 Reporting units and matching procedures |
Organic Farming Register: Reporting unit - agricultural holding (which is natural or legal person), undertaking organic production. The identification of holders and linkage to the appropriate holder in the database were done using the unique holding identifier (code of the holding). Organic production is an overall system of farm management and food production that combines best environmental practices, a high level of biodiversity, the preservation of natural resources, the application of high animal welfare standards and a production method in line with the preference of certain consumers for products produced using natural substances and processes. The organic production method thus plays a dual societal role, where it on the one hand provides for a specific market responding to a consumer demand for organic products, and on the other hand delivers public goods contributing to the protection of the environment and animal welfare, as well as to rural development. Integrated Administration and Control System (IACS) Crop Declaration Database: Reporting unit is beneficiary, i.e. applicant who meets all requirements for benefits and for which the support is awarded. The applicant - agricultural entity (agricultural holding) laid down in regulations applying for the current year to receive direct payments for utilised agricultural land, coupled support for livestock and support for area under the Rural Development Program measures. Actually, reporting unit is a farm (holding which is natural or legal person) which produce crop production.The identification of holders and linkage to the appropriate holder in the FSS 2016 database were done using the unique holding identifier (code of the holding). Animal Register (Bovine Register): Reporting unit - the owner of farm animals (which is natural or legal person), who keeps different species of farm animals. The identification of holders and linkage to the appropriate holder in the FSS 2016 database were done using the unique holding identifier (code of the holding). National Paying Agency Database (Rural Development Measures database): Reporting unit - agricultural holding (which is natural or legal person) willing to receive a support according to Rural Development Program and submitting an application for receiving this support. The identification of holders and linkage to the appropriate holder in the FSS 2016 database were done using the unique holding identifier (code of the holding). |
4.7 Difficulties using additional administrative sources not currently used |
Not applicable. |
Annexes: 3.1.-3.8. FSS 2016 sampling methods |
3.2. Frequency of data collection |
Frequency of data collection |
FSS data are available for the following years: 2003, 2005, 2007, 2010, 2013 and 2016. |
|
3.3. Data collection |
1. Data collection modes |
During the FSS 2016 in Lithuania the following data collection modes were used:
- Self-completed e-questionnaires;
- Face-to-face interview using portable computers;
- Telephone interview.
Agricultural companies and enterprises filled in the e-questionnaires themselves and sent them directly to Statistics Lithuania for further processing. Farmers' and family farms also had a possibility to fill in questionnaires and to transmit the filled-in questionnaires to Statistics Lithuania by themselves via the Internet using the new electronic statistical data preparation and transfer system e-Statistics for the Population. Thus, an e-survey was conducted. The e-survey was started on 1 June 2016 and was finished on 26 June 2016. After e-survey, all farmers' and family farms selected for the FSS 2016 were divided into two groups. One group covered farms with more than 10 hectares of UAA (about 30 000 farms); these farms were surveyed face-to-face. The remaining farms (about 15 000 farms with UAA of 10 hectares or less) were included in the other group; these farms were surveyed by phone. Interviewers started to work on 27 June 2016. First of all, face-to-face interviews were conducted. Face-to-face interviews were started on 27 June 2016 and finished on 9 September 2016 (lasted more than two months). The telephone interview was started on 12 September 2016 and lasted until 14 October 2016. |
2. Data entry modes |
Agricultural companies and enterprises filled in the questionnaires themselves and sent them directly to Statistics Lithuania for further processing via the electronic statistical data preparation and transfer system e-Statistics. ABBYY eFormFiller 2.5 software was used. The questionnaires were prefilled with available data from administrative data sources. The FSS 2016 questionnaire for farmers’ and family farms was prepared using ORBEON software. The market research company collected data from those farmers who did not fill in the questionnaires themselves. The market research company used portable computers (this requirement was included in the public procurement specification). Statistical data were entered into the same electronic statistical data preparation and transfer system e-Statistics for the Population (corresponding rights were granted to interviewers for this task). The data entered were sent to Statistics Lithuania (online) for further processing and checking. A special program created by ORACLE was used for statistical data processing and creation of the FSS 2016 database at Statistics Lithuania. When preparing for the FSS 2016, it was planned to take some data from administrative data sources and to put them into survey questionnaires without questioning agricultural holdings. This plan has been implemented. Therefore, after the data collection, data from the Crop Declaration Database, Animal Register, National Paying agency Database were taken and imputed directly into survey database. For those farms which declare their UAA, all crop areas (except outdoor vegetables, strawberries, potatoes, flowers and ornamental plants, greenhouses (vegetables and flowers and ornamental plants), mushrooms) and other UAA (except pome and stone fruits, berries, walnuts, nurseries) were imputed. Also, data on cattle, sheep, goats and horses as well as data on support for rural development and ecological focus area were imputed. Data were imputed into database created by ORACLE. SAS statistical package was used for linking statistical data from several sources according to the selected criteria and for the calculation of derived statistical indicators. All computer programs needed for the survey data collection, primary data entry, processing, analysis and calculations were created by the employees of Statistics Lithuania. |
3. Measures taken to increase response rates |
A big promotion campaign was conducted. Before the survey, as well as during the survey, there were many articles in the most popular newspapers about the FSS 2016 in general, the progress, the number of the farms surveyed, difficulties. The press release about beginning of the FSS 2016 was prepared and published in the Official Statistics Portal (https://osp.stat.gov.lt/) on 1 June 2016. Also, during the survey fielworks the information about the FSS 2016 was posted on the web site of Statistics Lithuania (in the section "News"). As data collection process (survey fieldwork) was subcontracted, i.e. services of a market research company were purchased, it was tried to inform about this fact farmers‘ and family farms as much as possible. Website of Statistics Lithuania was also used as much as possible. Before the survey, each respondent received a personal informational letter, explaining the main purpose of the survey, when it would be carried out, which data would be collected, the use and protection of data, contact information of the Agriculture and Environment Statistics Division of Statistics Lithuania. Also, in this letter the market research company which collected data from farmers’ and family farms was indicated. The interviewers received the special instructions for filling in the questionnaires. A special manual for them was composed. Also, trainings for interviewers from market research company was conducted. During these trainings it was explained that data of each holding in the survey sample is very important, especially of large farms. In case when there was no one who could answer the questionnaire at the moment when the interviewer was visiting a holding, the interviewer had to leave a letter with his contact information and to visit the holding once more or to conduct interview by phone, if farmer agreed to provide statistical data by phone. According to terms of contract between Statistics Lithuania and the market research company, which provided data collection service, the farm could be considered as non-response unit only if farm was contacted twice and no reaction from farmer by phone was received. There were only a few telephone reminders for agricultural companies and enterprises because they filled in the FSS 2016 questionnaire themselves. For farmers’ and family farms, reminders were unnecessary because they were questioned by interviewers. At the last stage of data collection, all agricultural holdings who marked that the holding does not perform agricultural activity or holding’s land is temporarily uncultivated, but the holder means to continue agricultural activity in the future, were selected by employees of Statistics Lithuania. Using personal ID, these farms were checked in the Crop Declaration Database or Animal Register and if they were found in the mentioned administrative data sources, they were investigated by interviewers once more or questioned by phone. Thus, all possible actions for increasing response rates were accomplished. |
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 |
185 967 |
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 |
45 928 |
3 |
Number of ineligible holdings |
2 114 |
3.1 |
Number of ineligible holdings with ceased activities This item is a subset of 3. |
1 605 |
4 |
Number of holdings with unknown eligibility status 4>4.1+4.2 |
910 |
4.1 |
Number of holdings with unknown eligibility status – re-weighted |
0 |
4.2 |
Number of holdings with unknown eligibility status – imputed |
0 |
5 |
Number of eligible holdings 5=5.1+5.2 |
42 904 |
5.1 |
Number of eligible non-responding holdings 5.1>=5.1.1+5.1.2 |
6 063 |
5.1.1 |
Number of eligible non-responding holdings – re-weighted |
1 028 |
5.1.2 |
Number of eligible non-responding holdings – imputed |
5 035 |
5.2 |
Number of eligible responding holdings |
36 841 |
6 |
Number of the records in the dataset 6=5.2+5.1.2+4.2 |
41 876 |
5. Questionnaire(s) - in annex |
Two questionnaires were created for the FSS 2016: 1) for farmers‘ and family farms; 2) for agricultural companies and enterprises. It was decided to create two questionnaires due to need to adjust these questionnaires to two different data collection and entry systems (e-Statistics and e-Statistics for the Population). These two questionnaires were approved by two different orders of the Director General of Statistics Lithuania. In both questionnaires the characteristics are divided into several parts (I–VI):
- Data on the agricultural holding;
- Land of the holding;
- Number of livestock;
- Farm labour force;
- Other gainful activities related to the holding;
- Soil and manure management practices.
Both questionnaires are provided in the Annex. |
Annexes: 3.3-5. FSS 2016 questionnaire (for agricultural companies and enterprises) 3.3-5. FSS 2016 questionnaire (for farmers' and family farms) |
3.4. Data validation |
Data validation |
The respondents filling in the questionnaire themselves as well as interviewers had to observe whether the answers provided were not contradicting to each other and complied with the logical and arithmetical connections given in the questionnaires. There were 93 different logical and arithmetic controls for the FSS 2016 questionnaire for agricultural companies and enterprises (both to the programs created using ORACLE software and ABBYY Form Filler). Also, there were 115 different logical and arithmetic controls for the FSS 2016 questionnaire for farmers' and family farms. Logical and arithmetical controls were consistent with Eurostat validation rules. There were mandatory and ignored errors. Mandatory errors had necessarily to be corrected. Ignored errors were designed to draw attention to the fact that there may be an error. After filling in the questionnaire, respondents or interviewers could see an error protocol (if there were any errors). This protocol appeared after checking the questionnaire due to logical and arithmetical controls. Then the respondents or interviewers had to correct all the errors before sending the questionnaire. If they tried to transmit a questionnaire with errors, incorrect questionnaires were returned to them for correction. Incorrect questionnaires had not been loaded to the database.
- Agricultural companies and enterprises filled in the electronic questionnaire and sent it directly to Statistics Lithuania using the electronic statistical reporting portal (e-Statistics) for further processing. If any uncertainties were obtained, specialists in the regional data preparation divisions checked if all the data were filled in, and mistakes were corrected and unclear items were cleared out by question to the company or enterprise by phone.
- Interviewers (as well as farmers), who used ORBEON software for entering data, gave the feedback on the errors in each questionnaire and had to correct them.
- When statistical data received by Statistics Lithuania were uploaded into the program for data processing, they were checked once again – whether they comply with the conditions of control. If non-conformity is found, its origin is determined and it is eliminated.
Information on the technological process of primary statistical data processing, its stages and description were provided in the technical requirements for market research company. The requirements for statistical data control were provided as well. The conditions, when such control should be carried out, are as follows:
- The description of control consists of error classification, error table and correction audit table;
- The error classification consists of the following: an error code (ID), which is a digital serial number; an error text, defining relations or deviation of statistical indicators; an error attribute marked by letter I – “May be ignored”, P – “Must be corrected”;
- Errors may be logical and arithmetic. They may be made by a respondent or may occur in the process of entry or processing;
- Errors were corrected manually, discussed with interviewers or by contacting the farm.
Comparisons of the survey data both at micro and macro level were made in order to ensure data quality by detecting outliers and discrepancies. Using the holder’s ID and holding’s ID, data obtained directly from the holding were compared to the data taken from administrative sources. Micro data comparisons were made. If any outliers were obtained, the specialists responsible for the FSS 2016 in the central statistical office contacted the holding in order to clarify those outliers. The following administrative sources were used:
- IACS Crop Declaration Database,
- Animal Register,
- National Paying agency Database.
One more useful tool for the data validation was the standalone validation tool created by Eurostat. Statistics Lithuania used this tool and it helped to clarify all the errors and to have the data which is eligible for the validation rules. |
|
3.5. Data compilation |
Methodology for determination of weights (extrapolation factors) |
1. Design weights |
The extrapolation factors were obtained by using SAS procedure surveyselect. Sampling weights were defined as the inverse of the units‘ selection probabilities. |
2. Adjustment of weights for non-response |
The re-weighting method was applied and reweighted Horvitz-Thompson estimator was used to estimate the indicators of survey. |
3. Adjustment of weights to external data sources |
Weights were not adjusted to external data sources. |
4. Any other applied adjustment of weights |
There was no other applied adjustment of weights. |
|
3.6. Adjustment |
[Not requested] |