3. Statistical processing |
Top |
|
1.Survey process and timetable |
The steps of the survey organisation |
|
1. Definition of survey objective and requirements |
|
1.1. Formation of workgroups for survey organisation; discussion and analysis of the experiences of the 2013 sample survey; suggestions for improvements and checks on their practicability |
From April 2015 |
1.2. Consultation of users |
From July 2015 |
1.3. Set-up objectives, target population, statistical units, classifications, precision requirements etc. |
From October 2015 |
1.4. Drafting of the national Regulation in cooperation with the MoA |
October 2015 – April 2016 |
1.5. Survey promotion |
September – October 2016 |
1.6. Publication of the national regulation |
7th September 2016 |
2. Survey design |
|
2.1. Set-up organisation of the survey (e.g. detailed timetable, specification of resources, costs estimation) |
April 2015 |
2.2. Definition of the survey variables |
October 2015 – April 2016 |
2.3. Determining the survey population; checks and release for dispatch |
July – September 2016 |
2.4. Design of the sampling frame and sampling procedures; sampling frame construction |
July – September 2016 |
2.5. Design, further development and testing of the electronic questionnaire |
January – September 2016 |
2.6. Design of the plausibility program by specialists |
October 2015 – March 2017 |
2.7. Compilation of other survey documents (instructions for use etc.) |
January – September 2016 |
3. Data collection |
|
3.1. Checks on the availability of administrative data |
January – July 2016 |
3.2. Obtaining administrative data |
September 2016; April 2017 |
3.3. Printing, addressing and sending the survey documents (external) |
August – October 2016 |
3.4. Sample selection |
October 2016 |
3.5. Recruitment of temporary telefone interviewers |
August – October 2016 |
3.6. Training of telefone interviewers |
October 2016 |
Reference date of the survey last day of the reference period |
31 October 2016 31 December 2016 |
3.7. Fieldwork |
October 2016 – March 2017 |
3.8. Reminder and warning procedures |
December 2016 – April 2017 |
3.9. Evaluation and assessment of fieldwork |
July 2017 |
4. Data processing and validation |
|
4.1.Programming the plausibility program; design of the plausibility application by the IT Dept. and tests of its functions by specialists, using fictitious holdings |
June 2016 – April 2017 |
4.2. Data taken from the electronic questionnaire in tranches for processing by the IT Dept. |
December 2016 – April 2017 |
4.3. Data validation (at record level), imputation, plausibility checks, data correction |
April – October 2017 |
5. Data compilation |
|
5.1. Weight calculation and estimation |
May – December 2017 |
5.2. Calculation of derived variables |
May – December 2017 |
5.3. Calculation of quality indicators (e.g. non-response rates, relative standard errors, coverage errors, bias etc.) |
November – December 2017 |
5.4. Aggregation and tabulation - compilation and programming of the tabulation program (STATcube) and of the Eurostat data files |
May – January 2018 |
5.5. Validation of aggregated data; checks on results; analysis |
September – December 2017 |
6. data analysis |
September – December 2017 |
7. data dissemination |
|
7.1. National dissemination and publication of the results in the form of a press release, rapid report, and publication on the Internet |
May 2017; December 2017; January 2018 |
7.2. Transmission of individual data of the FSS 2016 to Eurostat |
December 2017 |
|
2. The bodies involved and the share of responsibilities among bodies |
The project team Statistics Austria bears ultimate responsibility for implementing the Farm Structure Survey. The Farm Structure Survey is one of many projects of the Agriculture and Forestry Sector of the Directorate for Spatial Statistics (Raumwirtschaft). Its specialist team is supported by EDP (electronic data processing) and statistical experts in methodology. Technical advice concerning the contents is provided by the Working Party of the Advisory Committee on Agricultural Statistics, which comprises leading experts at various relevant institutions/organizations in the agricultural sector. The main tasks of the project team were:
- Compiling the survey program in line with EU requirements and taking national requirements into account;
- Cooperating in drafting the national regulation with the Ministry for Agriculture and Forestry, Environment and Water Management (MoA);
- Placing information articles in relevant media;
- Coordination of tasks between special departments and IT Department;
- Sample design and drawing the sample in cooperation with the statistical experts in methodology;
- Design of a plausibility program in cooperation with the IT Dept;
- Compilation of the questionnaire, instructing printers to print the documents, dispatching documents to the respondents;
- Training the staff in how to execute the survey (hotline agents (permanent staff) and telephone interviewers (temporary staff)); training the staff of Statistics Austria in how to process the survey;
- The hotline agents (permanent staff) providing respondents with information during the survey phase;
- Obtaining, processing and combining data from the various sources;
- Initiating reminder procedures in case of belated respondents;
- Reminding respondents who had not replied and reporting non-respondents to the competent authorities;
- Checks on the completeness and plausibility of the data;
- Specifications for the creation of the database (STATcube - data cube), tabulation, publication and dissemination of results;
- Processing individual data in line with EU rules, extrapolation and transmission to Eurostat.
Advisory Committee on Agricultural Statistics The Federal Statistics Act 2000 (BGBl. I No 163/1999, as last amended by BGBl. I No 136/2001, BGBl. I No 71/2003, BGBl. I No 92/2007, BGBl. I No 125/2009, BGBl. I No. 111/2010 and BGBl. I No 40/2014) provides for the creation of Advisory Committees for the various relevant areas of activity. The Farm Structure Survey comes under the aegis of the Advisory Committee on Agricultural Statistics, which comprises experts from various Austrian institutions (representatives of the MoA, the Governments of the Länder, Chambers of Agriculture at Land level, the Austrian Chamber of Agriculture, LBG Wirtschaftstreuhand- and Beratungsgesellschaft (a limited company), the University of Agriculture (Universität für Bodenkultur) etc.). This body is tasked with providing Statistics Austria, which bears ultimate responsibility for the survey as such, with mainly technical advice and support in the planning and implementation of the survey. |
3. Serious deviations from the established timetable (if any) |
No significant deviations from the schedule. |
|
3.1. Source data |
1. Source of data |
The FSS 2016 was conducted in form of a sample survey. The administrative data (IACS, System for identification and registration of bovine animals (cattle register)), Veterinary Information System (VIS) were integrated into the electronic questionnaire and checked by respondents while completing the questionnaire. The funding data (measures for rural development) were consolidated after the survey with FSS data using the unique identification number of the holding. Details on individual purpose(s) of the use of administrative sources see concept 3.1. Source Data, item 4.3. |
2. (Sampling) frame |
The frame of the FSS 2016 essentially comprised the active holding units in the Farm Register (or Agricultural and Forestry Register, AFR) meeting the national thresholds applied (178832 units). The type of frame is a list frame. The Farm Register is continuously updated in the light of various primary agricultural surveys and by comparison with various types of administrative data (applications for subsidies, social insurance information, necrologies etc.). |
3. Sampling design |
3.1 The sampling design |
The sample was designed as a one-stage stratified random sample of holdings and therefore is a probability design. There is no sub-sampling for some of the variables. The sampling method is systematic sampling with equal probabilities. |
3.2 The stratification variables |
Selected variables from the Agricultural Census 2010 were basically used for stratification purposes, e.g. total land area, orchards or vineyards, livestock numbers (the most recent data on cattle, pigs, sheep, goats and poultry) and the number of workers. The holdings in the sampling frame were divided into between 6 and 15 strata depending on NUTS2 (=federal state/province/Land). In Burgenland, Carinthia, Lower Austria, Upper Austria and Styria the strata 1 to 9 were formed by combining the size categories for the variables 'total area in ha (TA)' and 'arable area in ha (AA)'. In Salzburg, Tyrol, Vorarlberg and Vienna the strata 1 to 3 were formed by combining the size categories for the variables 'total area in ha (TA)'. The other strata comprise holdings with a high livestock population, a large labour force or significant fruit/vine cultivation. For stratification purposes, the conditions shown in Annexes 3.1-3.2. Detailed-stratification-and-delimitations-per-NUTS2 and 3.1-3.2. Stratification-conditions-NUTS 2 had to be met in each Federal State (Land). |
3.3 The full coverage strata |
Full coverage strata can be found highlighted in grey in the Annex 3.1-3.2. Detailed-stratification-and-delimitations-per-NUTS2, which shows the detailed stratification and delimitations per NUTS2 (=federal state/province/Land). |
3.4 The method for the determination of the overall sample size |
The sample survey covered about 30000 holdings. This size of sample guarantees sufficient accuracy. Calculations have shown that the required mandatory specifications regarding simple relative standard errors are satisfied with this sample size. Thus, 17 % of all holdings had to be surveyed. |
3.5 The method for the allocation of the overall sample size |
The total survey size of 30000 holdings was allocated to the 9 NUTS2-regions (federal states) proportionally to the square root of total area plus number of cows plus number of pigs. Subsequently, within each federal state the survey size calculated in the first step was allocated to the different strata according to Neyman-Pearson. |
3.6 Sampling across time |
A new sample is drawn on each occasion. |
3.7 The software tool used in the sample selection |
SAS procedure PROC OPTMODEL |
3.8 Other relevant information, if any |
Not available |
4. Use of administrative data sources |
4.1 Name, time reference and updating |
Integrated administration and control system (IACS) including the Austrian program for the funding of environmentally sustainable extensive agriculture that conserves the natural world (ÖPUL)
- Regulation (EU) No 1307/2013 of the European Parliament and of the Council of 17 December 2013 establishing rules for direct payments to farmers under support schemes within the framework of the common agricultural policy and repealing Council Regulation (EC) No 637/2008 and Council Regulation (EC) No 73/2009, as amended by Regulation (EU) No 1310/2013, and the Commission Delegated Regulations (EU) No 639/2014, No 994/2014, No 1001/2014, No 1378/2014, and No 2015/851.
- data are updated every year and used according to the reference dates and periods of the survey (see concept 2.8 Time coverage).
System for identification and registration of bovine animals (Cattle register)
- 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 and repealing Council Regulation (EC) No 820/97, as amended by Council Regulations (EC) No 1791/2006 and No 517/2013 and Regulation (EU) No 653/2014 of the European Parliament and of the Council,
- data are continuously updated and used according to the reference dates and periods of the survey (see concept 2.8 Time coverage).
Register of organic farms (comment: in a strict sense, there is no register of organic farms in Austria, information is taken from IACS/ÖPUL) (information see above). Measures for rural development
- Regulation (EU) 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.
- data are updated every year and used according to the reference dates and periods of the survey (see concept 2.8 Time coverage).
Based on Article 4 (2), Austria requested further the use of following administrative data: Veterinary Information System (VIS)
- Directive 2000/16/EC of the European Parliament and the Council of 10 April 2000 amending Council Directive 79/373/EEC on the marketing of compound feeding stuffs and Council Directive 96/25/EC on the circulation of feed materials; Council Regulation (EC) No 21/2004 of 17 December 2003 establishing a system for the identification and registration of ovine and caprine animals and amending Regulation (EC) No 1782/2003 and Directives 92/102/EEC and 64/432/EEC; Animal Diseases Act – TSG, Act of 6 August 1909, RGBl. No 177, concerning the defense and eradication of animal diseases (as amended in 2008 BGBl. I 2008/36); Animal identification and registration act 2009 BGBl. II No 291/2009: Regulation of the Minister of Health on the identification of pigs, sheep, goats and equidae and the registration of animal husbandry (Animal identification and Registration ordinance 2009; TKZVO 2009); Approved or registered ABP-plants according to Reg. (EC) No 1069/2009; Lists of approved food establishments according to Reg. (EC) No 853/2004 and Reg. (EC) No 852/2004.
- data are updated every year and used according to the reference dates and periods of the survey (see concept 2.8 Time coverage).
|
4.2 Organisational setting on the use of administrative sources |
Under the Federal Statistics Act 2000, BGBl. I No 163/1999, as amended by BGBl. I No 136/2001, BGBl. I No 71/2003, BGBl. I No 92/2007, BGBl. I No 125/2009, BGBl. I No. 111/2010 and BGBl. I No 40/2014), Statistics Austria is required to use the available administrative data instead of information obtained using its own questionnaires, so as to minimize the respondents' workload. On the other hand there is an obligation on the holders of administrative data to cooperate. According to the Regulation (BGBl. II No 243/2016) regarding the preparation of statistics concerning the structure of the agricultural holdings in the year 2016 Agrarmarkt Austria (AMA) is required to transmit the administrative data necessary for collecting the survey variables at the request of Statistics Austria free of charge on an electronic data carrier. Furthermore according to the Federal Statistics Act the position of Statistics Austria has/would have to be taken into consideration in the conceptual design and subsequent related revisions of the administrative sources. In the light of growing efforts to cut costs and streamline the requests on the administrative side the practical coordination with the statistical requirements might not allways be optimally implemented and has to be compensated by recoding or by directly asking the farmers in individual issues. Most effective would be the precise coordination of administrative and statistical concepts and requirements already at European level. |
4.3 The purpose of the use of administrative sources - link to the file |
Please access the information in the file at the following 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) |
The definition of the reporting unit in FSS is the same as the one used for the subsidy system and the related administrative data. In this context quality is ensured by the routine maintenance of the Agricultural and Forestry Register, which is continuously updated in the light of various primary agricultural surveys and by comparison with various types of administrative data (applications for subsidies, IACS, VIS, social insurance information etc.). |
--- |
--- |
- coherence of definitions of characteristics |
IACS including ÖPUL: Correlations between the variables were assessed. Area data has to be provided by 15 May of the current harvest year. The record day of the area data correlates with the record day of the IACS data (15 May 2016). Information on ecological farming has to be provided by 15 November. |
IACS including ÖPUL: Essentially, there is a good correlation between the variables. In the AMA multiple application, area data are often recorded more detailed in Schlagnutzungsarten (land parcels). In a very small minority of cases variables may not be covered sufficiently by administrative data |
IACS including ÖPUL: Transition tables were used for aggregating land parcels from IACS to the required FSS categories. In isolated cases Statistics Austria had to assign the IACS data to relevant items in the FSS program. Not sufficiently by administrative data covered variables, were entered into the electronic questionnaire in the frame of the FSS 2016 by respondents. |
Cattle register: Correlations between the variables were assessed. Record day of cattle data (FSS) correlates with the record day of the cattle data base (1 April 2016). |
Cattle register: The cattle register contains the complete bovine categories, with exception of dairy cows/other cows, for the breakdown of individual items in accordance with the guidelines of the FSS. |
Cattle register: The number of cows and the total number of cattle for the reference date 1 April 2016 was integrated into the electronic questionnaire. The respondents had to divide the number of cows into dairy cows and other cows in questionnaire. The detailed information (categories) of the rest of the cattle was consolidated after the survey with FSS data using the unique identification number of the holding. |
- coverage: |
|
|
|
|
over-coverage |
IACS including ÖPUL, Cattle register, Measures for rural development, VIS: Over-coverage is excluded. |
--- |
IACS including ÖPUL, Cattle register, Measures for rural development, VIS: Out-of scope units were excluded. |
|
under-coverage |
IACS including ÖPUL: Unreported events of units, which can be considered by themselves as covered by administrative data, are excluded due to the control- and sanctioning mechanisms (see also comparability below). |
IACS including ÖPUL: Data are only available for holdings which submit a multiple application to Agrarmarkt Austria. |
IACS including ÖPUL: Holdings which do not submit a claim in a particular year are required to make the data available in the course of the FSS. |
Measures for rural development: 100% coverage by administrative data. |
--- |
--- |
Cattle register, VIS: Unreported events of units, which can be considered by themselves as covered by administrative data, are excluded due to the control- and sanctioning mechanisms (see also comparability below). |
Cattle register, Measures for rural development, VIS: Data are only available for agricultural holdings keeping livestock which require to be reported by the VIS full survey or submit an IACS funding request. |
Cattle register, Measures for rural development, VIS: Agricultural holdings keeping livestock which neither require to be reported by the VIS full survey nor submit an IACS funding request have to provide the information about livestock as part of the Farm Structure Survey. This applies only to very few holdings keeping exclusively poultry, horses or farmed game. |
|
misclassification |
IACS including ÖPUL, Cattle register, Measures for rural development, VIS: Misclassification is excluded. |
--- |
--- |
|
multiple listings |
IACS including ÖPUL, Cattle register, Measures for rural development, VIS: There are no multiple listing errors due to the use of a unique registration code for each holding. |
--- |
--- |
- missing data |
IACS including ÖPUL: The occurrence of variables were assessed. |
IACS including ÖPUL: There are missing data on B_1_7 Fresh vegetables, melons, strawberries, B_1_8 Flowers and B_2 Kitchen gardens. |
IACS including ÖPUL: The respondents are required to make the data available in the course of the FSS. |
Cattle register, VIS: Missing data are excluded due to the control- and sanctioning mechanisms. |
--- |
--- |
- errors in data |
IACS including ÖPUL: In situ inspections are carried out at farmers’ premises by AMA as part of the IACS and false declarations punished by reductions in premiums. Statistics Austria is using already checked data (September). Therefore it can be assumed that the administrative data are highly accurate. The use of a digital cadaster map (DCM) is also expected to yield accurate information on areas. |
--- |
--- |
Cattle register: A series of checks are performed in the course of the administrative process, this allows that the use of data from the Cattle Register makes for a greater accuracy and reliability of the results for a cattle population. |
--- |
--- |
VIS: The annually on 1 April surveyed data were integrated into the VIS database. As these administrative data underlie continuous controls, a great accuracy and reliability of the results can be assumed. |
--- |
--- |
- processing errors |
IACS including ÖPUL, Cattle register, Measures for rural development, VIS: In principle there are no processing errors in the context with the use of administrative data due to the use of unique identifiers on both sides (administration and statistics). |
--- |
IACS including ÖPUL, Cattle register, Measures for rural development, VIS: Nevertheless, by prefilling the administrative data (IACS and VIS) concerning areas and livestock into the electronic questionnaire the respondents would have the opportunity to check and correct the data while completing the electronic questionnaire. |
- comparability |
IACS including ÖPUL: In 1995 the information in the IACS was compared with the results of agricultural surveys carried out in the traditional manner (Farm Structure Surveys and livestock censuses). |
IACS including ÖPUL: Comparison of the items revealed a very good match with the agricultural surveys. |
IACS including ÖPUL: Nevertheless, by prefilling the administrative data concerning areas into the electronic questionnaire the respondents would have the opportunity to check and correct the data while completing the electronic questionnaire. In case of large discrepancies the farmers would have been recalled. |
Cattle register: Comparisons were made between the analyses of the Cattle Register and the results of the livestock censuses. The data were checked as part of the application to use administrative data from the Cattle Register instead of statistical cattle surveys. |
Cattle register: There was a close correlation between the results of this source and the livestock censuses. |
--- |
VIS: Comparisons of VIS data were made with results from statistical surveys. |
VIS: Comparisons suggest a very good correlation between this source and the results from statistical surveys. |
VIS: Nevertheless, by prefilling the administrative data concerning livestock into the electronic questionnaire the respondents would have the opportunity to check and correct the data while completing the electronic questionnaire. In case of large discrepancies the farmers would have been recalled. |
--- |
--- |
Austria runs the FSS late in the year for the sake of optimising the use of administrative data. To an earlier date the administrative data would not be available in the same (verified and highly consistent) quality to pre-fill answers in the questionnaires which are then checked by farmers during the survey. |
- other (if any) |
--- |
--- |
IACS including ÖPUL, Cattle register, Measures for rural development, VIS: In isolated cases data had to be aggregated. |
4.5 Management of metadata |
IACS area data and lifestock data are stored in the Farm Register (or Agricultural and Forestry Register, AFR). If available, Metadata of IACS, ÖPUL, Cattle register, VIS and Measures for rural development are stored and maintained in electronic format. |
4.6 Reporting units and matching procedures |
Integrated administration and control system (IACS) including ÖPUL, System for identification and registration of bovine animals (cattle register), Measures for rural development, Electronic Veterinary Register pursuant to §8 Animal Diseases Act (TSG) – Veterinary Information System (VIS): The reporting unit (statistical unit) used is the agricultural and forestry holding. Around 14% of holdings includes two or more sub-holdings (in many cases, these are alpine pasture units). The definition of the reporting unit meets the definitions used for the subsidy system and the related administrative data. Every agricultural and forestry holding (resp. subholding) has a unique identification number, that is used in subsidies as well as in the frame of statistical surveys. The assignment between various data sources is done with this unique identification number (exact matching procedure). There were no mismatching cases. |
4.7 Difficulties using additional administrative sources not currently used |
Not existing. |
Annexes: 3.1-3.2. detailed-stratification-and-delimitations-per-NUTS2 3.1-3.2. stratification-conditions-NUTS2 |
3.2. Frequency of data collection |
Frequency of data collection |
The frequency of the surveys/data collection is determined by the European legislation and each survey is dealt with under a separate national regulation. |
|
3.3. Data collection |
1. Data collection modes |
The Farm Structure Survey was held solely using an electronic questionnaire (e-Quest). The farmers were able to submit their return either directly at the computer after entering their user ID and password (direct respondents, CAWI) or during a personal interview by telephone using the same electronic questionnaire (CATI). About 60% of the questionnaires were returned by direct respondents. About 40 % of the respondents provided information by means of telephone interviews. Those farmers, who took the help of the telephone interviewers to complete and submit their questionnaire, either called directly the free hotline or arranged an interview by sending a prepaid reply card back to Statistics Austria, filled in with their telephone number and availability (workday and time window). The competent interviewers opened the farmer’s survey form using the farmer’s access data and filled out the electronic questionnaire with the necessary information. The survey took the form of a personalised electronic questionnaire, in which the name and the address of the holdings were already entered and only had to be checked and, if necessary, corrected. Detailed information material on how to use the electronic questionnaire and administer the Farm Structure Survey was sent directly to the respondents by post in October 2016. These consisted of an accompanying letter, a survey manual with a full description of the electronic questionnaire and a reply card (to arrange an interview) with a prepaid envelope. A dedicated free hotline was set up by Statistics Austria to answer any questions that arose during the survey phase. In addition, queries could be sent by e-mail to Agrarstrukturerhebung@statistik.gv.at. |
2. Data entry modes |
Due to an exclusive use of the electronic questionnaire, no separate data collection was necessary. The data were entered by the respondents (CAWI) or by the telefone interviewers (Electronic data capture during telephone interview, see above). The data of only very few holdings, who submitted their questionnaire by mail or fax to Statistics Austria, had to be entered manually into the electronic questionnaire. The IT department took the data directly from the questionnaire in tranches and imported it into a database. |
3. Measures taken to increase response rates |
The following measures were taken to increase response rates.
Maintaining up-to-date information in the Farm Register (or Agricultural and Forestry Register, AFR)
The Farm Register (or Agricultural and Forestry Register, AFR) on which the FSS is based is constantly being enhanced in terms of technical aspects and content maintenance as a result of increased updating options (administrative data, other registers etc.). In the run-up to the survey additional measures were taken to improve the up-to-dateness (e.g. adjustments with necrologies etc.). Awareness campaign Concerted articles were published in the relevant newspapers, journals and web pages to inform on the survey, its purposes and the importance of cooperation. Trainings
- Training of all hotline agents (permanent staff of Statistics Austria) engaged in the FSS.
- Training of all telephone interviewers (temporary staff) engaged in the FSS.
Hotline-strategy For the hotline staff and telephone interviewers at Statistics Austria, a list of frequently asked questions with regard to the motivation of respondents was compiled as a means of preparing them for questions from "difficult" respondents. Hotline agents were trained to convince respondents, who called and signalised to boycott the FSS, to give the information via direct telephone interview. Regular meetings of the hotline agents and the issue of newsletters facilitated information flows. Telephone Interviews Although the relatively low return quota of the reply cards (to arrange an interview) requires further improvement (only about 12,1% reply cards were returned), the farmers very willingly accepted the possibility to provide the information via telephone interview in the end. The low return quota of the reply cards had to be offset by intensified research for telephone contact information in the Farm Register (or Agricultural and Forestry Register, AFR), phone book and internet. Reminders for overdue questionnaires A graded series of reminders for overdue questionnaires was applied.
- In former surveys the local authorities had to ensure that all questionnaires were returned and it was their responsibility to ask the holdings (either in person, over the telephone, in writing or by registered mail) to complete the questionnaire. Since the FSS 2013 the Austrian local authorities have not been directly involved. Now the holdings were targeted by telephone-interview procedures by staff at Statistics Austria.
- About 3457 farmers, who had failed to complete the questionnaire on time or who could not be contacted by phone, were reminded and notified of the legal consequences via registered letter.
- 93 farmers insisted on their refusal and were ultimately reported by Statistics Austria to the relevant authorities in April 2017 to initiate administrative penal proceedings. Since Statistics Austria has no executive power to pursue administrative penal proceedings, information about these holdings had to be given to the relevant administrative districts that are responsible in Austria for conducting prosecutions. Normally a fine is imposed and a deadline is set for supplying the required information, i.e. payment of the fine does not release the farmer from the duty of supplying information; he/she must still provide the data in all cases.
- After prosecution 24 farmers submitted their data properly – if very late in some cases – to Statistics Austria. 69 farmers did not cooperate at all. Their data had to be imputed.
|
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 |
155314 |
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 |
27728 |
3 |
Number of ineligible holdings |
1542 |
3.1 |
Number of ineligible holdings with ceased activities This item is a subset of 3. |
498 |
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 |
26186 |
5.1 |
Number of eligible non-responding holdings 5.1>=5.1.1+5.1.2 |
69 |
5.1.1 |
Number of eligible non-responding holdings – re-weighted |
- |
5.1.2 |
Number of eligible non-responding holdings – imputed |
69 |
5.2 |
Number of eligible responding holdings |
26117 |
6 |
Number of the records in the dataset 6=5.2+5.1.2+4.2 |
26186 |
5. Questionnaire(s) - in annex |
See annex. |
Annexes: 3.3-5. Questionnaire_2016_AT |
3.4. Data validation |
Data validation |
Edit rules/checks Micro-level processing was carried out by means of extensive plausibility checks. The formal checks on the data involved a plausibility program containing about 196 plausibility rules, which again included all the controls of the Data Suppliers Manual. The types of checks performed were completeness checks, data validation, valid value checks, range checks, relational checks, arithmetic checks, ratio edits. The plausibility rules made distinctions between the following types of error:
- Automatic errors
These were errors that could be automatically corrected using programmed instructions.
- Information errors (100)
This mainly involved identifying input errors. Limit values were incorporated into the program for certain items in particular, e.g. to prevent entries being made in the wrong units of measurement (for example m²) in the case of specialised crops. If these limits were exceeded, this fact was reported. Processing staff then had to investigate or use their specialised knowledge to confirm that the data were correct or make the necessary corrections.
- Other errors (96)
Processing staff had to correct these, either by recalling/consulting the respondents or on the basis of their specialised knowledge.
Moreover, the nil returns were examined. If, for example, administrative information on the holding was available, the holding was surveyed again. This was done in close collaboration with staff dealing with the Farm Register (or Agricultural and Forestry Register, AFR), as the information from the nil returns (business closure, leasing, etc.) were used for updating the registers. Tools used for data validation Application for Plausibility checks The data sets were checked for missing, incorrect or implausible information, using an extensive plausibility application. The program was developed in cooperation with the IT Dept. Care was taken to ensure that missing, incorrect and implausible entries were detected by the program and either highlighted or immediately corrected. The functionality of the plausibility program was first checked using fictitious holdings. The correction applications contained a number of deliberate errors in order to check whether the program would recognise and report them. For the FSS 2016 about 81 % of the questionnaires needed further checking due to information errors or "real" errors. This share was higher than in 2013 due to the variables concerning soil and manure management practices. For each holding, all errors and information errors were listed and categorised. The errors detected (incorrect entries, missing or implausible data) had to be investigated and rectified by the processing team. Errors were eliminated and plausibility checks carried out directly via the application. Holdings for which errors still remain are flagged as incorrect and had to be processed once again. This process was repeated until the program detected no more errors or inconsistencies. The staff themselves could correct logical obvious errors. Frequent meetings of the staff facilitated information flows. Discussing the main issues arising from the work made it easier to standardise the criteria to solve similar situations. Missing or incorrect entries were completed from other data sources wherever available (e.g. administrative data from IACS or ÖPUL, “total area information” from the social insurance for farmers) to avoid burdening the respondents. The forestry yearbook, containing the areas of Austria’s largest forestry holdings, was another means of checking data. If these sources were not exhaustive, individual items from the 2010 or 2013 Farm Structure Survey were used, wherever possible, to supplement and/or check the data. Where this did not provide clarity, individual holdings had to be contacted by telephone. Figure: Plausibility application - search function Level of data validation At the level of the Electronic questionnaire (respondent/telefone interviewer) The questionnaire was designed in such a way that certain data items were checked for plausibility and for completeness respectively while being entered or before the questionnaire could be returned, with the result that serious errors did not go undetected and were not accepted. In order not to overload the questionnaire application and make it unnecessarily difficult for the respondents to use, this immediate plausibility check had to be limited to the most important content. Preventive measures were also taken to avoid instances of individual questionnaire sheets inadvertently being "skipped over": the marker was required to enter "The entries on this page are complete" on every page of the questionnaire. The checks in the electronic questionnaire included completeness checks, valid value checks, relational checks. At the level of the Application for Plausibility checks (processing team) Formal checks on the data imported into the database involved the plausibility program mentioned above. |
|
3.5. Data compilation |
Methodology for determination of weights (extrapolation factors) |
1. Design weights |
Design weights were obtained by taking the inverse of selection probabilities of holdings. |
2. Adjustment of weights for non-response |
The response rate was ultimately 99.77%. The data of the 69 units which refused to fill in the questionnaire, could be imputed by using administrative or other data-sources (Internet etc.). Therefore no re-weighting for non-response was necessary. Non-response because of non-existence of holdings at the time of the data collection: No adaption of sample weights necessary. |
3. Adjustment of weights to external data sources |
The weights were not adjusted to external sources. |
4. Any other applied adjustment of weights |
No other adjustment of weights. |
|
3.6. Adjustment |
[Not requested] |