OFFICIAL STATISTICS PRINCIPLES
In February 2005, Statistical Programme Committee of the European Commission accepted the European Statistics Code of Practice and declared to be bound by its principles. The principles are based on existing international standards. In this context 15 principles and 82 indicators have been generated. Necessary legal arrangements in the Statistics Law of Turkey No. 5429 have been made according to these principles. European Statistics Code of Practice have been revised in 2011 and 2017. The most current version is released on 16 November 2017. In this release, a new principle, "1.bis: Coordination and Cooperation", was added under Institutional Environment section, hence the number of principles increased from 15 to 16 in Code of Practice. New principle is emphasizing the importance of having a programme, coordination and cooperation in statistics production. The five year Programme also is based on these principles considering the production and dissemination of official statistics.
Intended matters in production of official statistics are similar for providers, users and producers who constitute the cornerstones of the production process. Therefore, guidelines for compilation, verification, analysis and dissemination of statistics must exist. The main purpose of these principles is to provide and increase the public trust in statistics which are produced to serve the formation of policies and to monitor the implementation of policies. This trust is built upon the quality of data and the transparency of statistical methods. However, the commitments made to the quality of applications should be open to public observation.
All institutions / organizations within the scope of the OSP may have to apply to the private sector contractors the field implementation and research phases in order to provide this information on the official statistical production in cases where they can not obtain the information they need to maintain their managerial activities. In this case, institutions / organizations are responsible for TURKSTAT within the following principles. However, private sector contractors are also obliged to implement these principles.
In order to increase the confidence in the official statistical production and the acceptable standards in production of official statistical researches, basic principles to be taken into account in production and dissemination of official statistics are as follows:
1. Professional independence: The reliability of official statistics is provided by the independence the private sector as well as the public institutions and other political, regulatory or administrative agencies.
The institutions responsible for the production of official statistics are responsible for producing and distributing the statistics independently.
2. Coordination and Cooperation: TurkStat ensures coordination and effective cooperation between the institutions / organizations included in the Program for the production, improvement and dissemination of of official statistics.
3. Mandate for data collection: The institutions and organizations responsible for the production of official statistics, in the context of the purposes of Turkish Statistical System, should have an explicit legal authority to collect data.
4. Adequacy of resources: The resources allocated for the production and distribution of official statistics should be sufficient to meet the requirements of the Turkish Statistical System.
5. Commitment to quality: The quality of statistical data, statistical techniques must be checked regularly and should be reported.
All institutions / organizations within the scope of the OSP are obliged to comply with the principles of reliability, consistency, confidentiality and transparency in the statistics they produce.
Institutions / organizations should follow the quality of compilation, processing and distribution processes and the quality components of produced statistics in official statistics they produce. Institutions / organizations are obliged to provide quality commitment even if private statistics firms are to be produced by private sector contractors.
6. Statistical confidentiality: The institutions and organizations producing official statistics have to take every precaution against unlawful access to or use of confidential data. The confidential data compiled, processed and preserved for the production of official statistics cannot be delivered to any administrative, judiciary or military authority or person, can not be used for purposes other than statistics or as an instrument of proof.
Data confidentiality and security procedures and guidelines are regulated by the “Regulation of Procedure and Principles of Data Confidentiality and Confident Data Security in Official Statistics" published in the Official Gazette No. 26204 in 20/06/2006.
7. Impartiality and objectivity: The institutions and organizations producing official statistics should produce and disseminate these, under the principles of scientific independence, equal access of all users and impartiality, and professionally in a transparent manner. Data providers must notice, which data to be collected, the reasons and the purpose of the data collection, the statistical methods and the essence of the data to be published.
8. Sound methodology: European and other international standards in collection, processing and distribution of official statistics must be exactly minded.
9. Appropriate statistical procedures: All applied statistical processes from data compilation to verification, must form the basis of the production of qualified statistics.
10. Non-excessive burden on respondents: In production and distribution of statistics, resources should be used effectively. The respondent's burden on households, firms or individuals must be observed and the appropriate targets must be identified accordingly. Monitoring the user's needs and expectations, appropriate dissemination forms for different user groups must be developed.
11. Cost effectiveness: Resources must be used efficiently.
12. Relevance: Official statistics must meet the user needs at the national level, must be enough to explain its topic, and must be disseminated along with its definitions and classifications.
13. Accuracy and reliability: Official statistics must reflect the current situation and the changes accurately and reliably according to their topic and scope.
14. Timeliness and punctuality: Official statistics must be produced timely within the existing data infrastructure. Time of release must be determined beforehand and official statistics must be released in accordance with the National Data Release Calendar.
15. Coherence and comparability: Official statistics must be consistent within and in time, comparable across regions and countries, available for combination with related data from different sources and available for shared use.
16. Accessibility and clarity: Official statistics must be presented in a clear and understandable manner, and be published in a convenient an accessible way, together with supporting metadata and guidance. Distribution must be with tools easily accessed by each user (printed media, media, internet, etc.). The basic aim should be to distribute the data along with the relevant metadata via the web free of charge.
Provision of official statistics mirroring the facts, responding to the needs of users, provided to users impartially and simultaneously, being in compliance with the data confidentiality, respecting the public right to get information are basic principles. In order to assess the data quality and appropriateness of the principles of official statistics, all the necessary information and methods used in the production of official statistics are announced to the public by appropriate means in accordance with standards set by TURKSTAT.
All institutions and organizations within the Programme are responsible for carrying out their duties within the framework of these principles. The Programme seeks to ensure the implementation of these principles and prepared accordingly. Detailed description of the topic is included in the "Handbook of a Quality Study" published by TURKSTAT. (www.tuik.gov.tr)
1. The scope of the research depending on the target size of the study and the target group (depending on the subject, the purpose, the place, the time, and the time the research will be carried out) should be clearly indicated,
2. The variables to be measured during the planning of the research should be determined.
3. The framework of the study (a list of all target group units and sample selection allowed) should be identified; where it should be obtained or how it should be created.
4. The method of sampling should be determined within the rules of probability, depending on the nature of the research and the existing framework.
5. Calculating sample size of the researcher; the permissible level of error, the sensitivity level, the loss data rate, the structure of the sampling design and the number of layers to be produced, and the stratification should be carried out for the purposes of the research.
7. Institutions / organizations should clearly identify the concepts, definitions and classifications used in the official statistics they will produce. Should take into account the standard concept, definitions and classifications determined by TURKSTAT and regularly follow the European Union (EU) and other international standards and document all definitions, concepts and classifications used in official statistics.
8. "Area Implementation Guidelines" should be prepared in which explanations and concepts related to the research, explanations about the questions on the questionnaire of the researcher and explanations about the application processes defined by the personnel who are to be involved in the field application are defined.
9. Pre-test and pilot studies of the questionnaire should be performed before the application of the field, the functioning of the questionnaire, the suitability of the data collection method, The information should be reported and the questionnaire and field application guide should be revised in accordance with the results obtained.
10. Selection possibilities should be calculated by making sample selection according to design criteria. In the selection phase, firstly check whether the frame is completely taken and then check the calculation of the selection possibilities.
11. The qualifications of the personnel (interviewer, controller, etc.) to be assigned during field application should be defined and training should be given to these personnel in the direction of training materials no later than 15 days before the start of field application.
12. The edit / code instructions to be used in the data entry process must be created.
13. Micro and macro level data analysis processes should be defined and determined by whom.
14. Unit / substance non-response rates should be calculated, provided that the non-response rate for non-rare characteristics is not exceeded 30%, depending on the variable to be measured. If imputation is used instead of non-response, the same rates should be maintained.
15. In the field application process, it is necessary to define in advance what error level will be returned to the field and the field application process should be controlled. It must be determined in advance by whom and by which method these checks will be carried out.
16. Design inaccuracies should be documented by taking into account non-responsiveness, initial selection possibilities and external sources of data.
17. Institutions should calculate the coefficient of variation for the basic variables in the studies conducted under official statistics. They should report quality indicators such as non-response rates, edit rates, imputation rates, and coefficient of variation values.
18. Institutions should provide explanatory information, which is defined as the metadata of the research, together with the final results they will present to the user regarding the official statistics they produce.
The most basic principles to be considered in the research process are defined above. For the administrative records, which may be source data in the production of official statistics, the publication "The Role of Administrative Records in Statistics Production" should be taken into consideration.