STATISTICS AGENCY

This process involves studying the current and future needs of users of official statistics of all categories (including foreign users) in statistical data, as well as determining when, in what form and for what purposes they are needed.

Needs are also identified by studying the practice of statistical departments of foreign countries, national and international organizations involved in the production of similar statistics, as well as the methods and rules they use.

As part of the process, producers of official statistics are planning activities for the development and (or) preparation of statistical data (including statistical publications) that will be published in the future, as well as systems and tools used at the stage of their dissemination. This includes planning work on the development and (or) improvement of samples of tables, graphs, charts and statistical publications for further publication, layouts and samples of statistical materials posted on websites, information systems and special software for publishing and publishing statistical materials, training seminars, guidelines for the design of website pages and filling in statistical publications.

Sources used in the process of planning published statistical publications and other data include metadata obtained from the previous collection of statistical data.

As part of this process, work is underway to create new or improve existing data collection mechanisms used at the “Primary Data Collection” stage. Mechanisms for data collection will be formed or improved in accordance with the plans developed during the “Planning of production processes” stage.

Data collection can be organized using one or more methods (personal or telephone surveys, paper, electronic or online observation forms or questionnaires). This process also includes preparing the content and structure of the relevant tools (observation forms and questionnaires), testing how they work in practice, such as testing questions in a statistical report or observation form.

As part of this process, the design of the frame and sample is carried out to organize and conduct data collection processes, as provided in the “Planning the frame and sample” process.

This process involves coordinating survey process samples in one direction (to detect identical similarities or substitution of units) or sampling between survey processes in different directions using a common frame and single register data (to identify identical similarities or optimize response burden). As part of this process, quality assurance work is carried out, as well as the approval of the general population and sample. Indicators from administrative and other sources can be used as an auxiliary indicator when designing the sample.

In the process of data integration, the results of the processes of the “Collecting primary data” stage are summarized, i.e. aggregation of the collected data. The aggregated data may come from various external and internal sources and are the result of various data collection methods. Electronic databases and summary tables of primary and (or) administrative data are used as data sources. The integration result is an interconnected set of data.

Data integration can be carried out not only at the “Data Processing” stage, but also at all stages and processes of statistical production. After the integration, depending on the requirements for their protection, the data can be depersonalized (identifiers such as name, address are removed) in order to ensure confidentiality.

During the preparation of preliminary statistical materials, the data obtained from the results of processes 5.7 “Calculation of aggregated data” and 5.8 “Completion of data generation” are converted into the form of preliminary statistical materials. This process involves the creation of additional indicators, such as indices, trends or seasonally adjusted time series, as well as the recording of qualitative characteristics, such as coefficients of variation.

In this process, work is carried out to review the results of the output tables prepared at the “Data Processing” stage, as well as the necessary information obtained from the database or the database of ready-made documents, to build diagrams, cartograms, graphs and new tables based on the generated data.

Also, when analyzing the quality of statistical data, statistical indicators are considered for reliability and accuracy (validity, absence of errors, calculation of coefficients of variation, etc.), integrity (completeness of coverage, compliance with statistical methodology), consistency, comparability (territory, range of objects covered, unit of measurement etc.), homogeneity (the presence or absence of cases of exclusion of data, a sharp change in them, omissions).

As part of the process of upgrading statistical production systems, the systems (e g databases) that store statistical data and metadata ready for dissemination are upgraded. Specifically, the system upgrade process includes: formatting of data and metadata ready for input into systems (databases) of output data; loading data and metadata into systems (databases) of output statistical data; ensuring linkage (consistency) of statistical data with the corresponding metadata.

These measures should in most cases be carried out in earlier processes, but the process of updating statistical production systems will finally check that all necessary data and metadata are available and ready for dissemination.

Work on the formation and collection of data for evaluation can be carried out as part of any other stage or process.

This process can be carried out in various forms, including the generation and collection of user comments and suggestions, process metadata, system metrics, and employee suggestions. Also, inputs to the evaluation process can be reports on the progress of the work defined in the agreed action plan for the next period based on the results of the evaluation of the previous production period. The data collection process for the evaluation of the production cycle includes data collection activities for: evaluation of the quality of statistical data; monitoring the burden on respondents; assessing user satisfaction and trust in official statistics; monitoring staff satisfaction with the stages and processes of production.

However, specific activities may be required to evaluate certain processes, including small observations (post-enumeration controls, follow-up interviews, research on the effectiveness of data dissemination).

As part of this process, Producers of Official Statistics consult with stakeholders in order to accurately define their needs for statistics, including the form and timing of their submission, as well as the purposes for their further use, taking into account previously identified needs and confirming their needs.

At the same time, it is necessary to achieve a deep understanding of the needs of users, i.e. determine not only what kind of statistical data are required, but also in what context, when, in what form and, most importantly, for what reasons their formation is necessary.

In this process, work is planned to prepare the composition of collected and generated indicators in order to prepare official statistical data and their metadata, as well as statistical classifiers and reference books.

At the same time, due to the interdependence of the processes for determining the indicators necessary for collecting data and choosing the mechanisms for their collection, with the wide use of existing domestic and foreign experience, it is advisable to carry out this process in parallel with process 2.3 “Planning for the collection of primary data”.

The process of forming or improving the components of processing and analysis involves work on the reuse of existing or the formation of new components identified at the stage “Planning production processes” and necessary for the stages “Data processing” and “Data analysis”.

In this process, work is carried out to create new information systems or special software, make changes and (or) additions to existing information systems or special software for data processing and analysis. The process ends after the completion of the development and testing of information systems or special software for data processing and analysis, the development of related technical and operational documentation.

This process ensures the readiness of personnel, production mechanisms, systems and technologies to collect data and metadata in all planned modes, and includes the following activities: preparing a data collection strategy; training of data collection personnel; ensuring that the necessary tools for data collection are available; completing the systems necessary to request and receive data; ensuring the protection of the collected data; preparation of data collection tools; notification of respondents, work with them; translation of materials into the required languages.

For non-observation sources of data collection, this process ensures that the necessary confidentiality processes, systems and procedures are in place to obtain and extract the necessary information from these sources.

As part of this process, classification and coding of primary data are carried out. When classifying data, statistical and other classifiers, nomenclatures and reference books used in practice in the field of official statistics are used.

The process of classification and coding of primary data can be carried out automatically (or manually) by assigning numerical codes to text answers (data) in accordance with a predefined statistical classification system.

Also, some questions in the observation forms or in the administrative data source may have variants of coded response categories. Some data may be encoded using an automated process (for example, machine-reading prepared responses) or using an interactive manual process.

In the process of validating statistical materials, work is carried out to determine the compliance of data collection and generation activities with previously set quality requirements and expected results. In this process, all data is collected to obtain in-depth knowledge of the particular observed area of statistics, and this knowledge is used to identify any deviations from the expected results in specific conditions and to analyze the information.Validation work may include:

verification of compliance with the established requirements of the primary data coverage indicator and response rates; comparison of statistical data with previous cycles (if applicable); checking the availability of metadata for the prepared material and the compliance of this material with the requirements of metadata and quality indicators; comparison of statistical data with other relevant (external and internal) data sources; study of discrepancies and differences in statistical data; documented evidence that the statistics produced for the observed industry are in line with the expected results.

As part of the process of preparing official statistics for dissemination, official statistics are produced in the forms planned earlier in process 2.1. “Planning statistical output”.

Official statistics produced to meet the needs of users can be presented in various forms. In particular, they may include printed publications, press releases, newsletters, tables, interactive graphics, cartograms, public open datasets, downloadable files, videos, information on official websites, etc. The standard steps for producing data for dissemination are: preparation of components of the statistical material (filling in the layout of the statistical publication, comments, tables, graphs, etc.); bringing the components into the form of a finished product (combining all the data, generalizing in the form of a finished statistical publication); review, approval, editing, checking for errors and compliance with the publication criteria of the final statistical material; production of statistical material (publication, preparation for placement in dissemination media). .

The evaluation process is carried out by summarizing, processing and analyzing the data obtained. In this process, the data is compared to expected or target results (if any) and summarized in an assessment report or checklist.

The quality of statistical data is assessed on a five-point scale based on filling out a special questionnaire using a rating method that characterizes the degree of compliance of statistical observations and data in them with the main quality criteria.

Work on monitoring the burden on respondents is carried out by summarizing data on the time spent by respondents on filling out forms of statistical observations. Based on the data obtained, the reasons for the high labor intensity of filling out the forms of statistical observations are clarified. In assessing the level of user satisfaction and trust in official statistics, the responses to surveys conducted among users are summarized and analyzed. Based on this, an index of user confidence in official statistics is calculated. Monitoring of staff satisfaction with the stages and processes of production is carried out by conducting a survey among employees of state statistics bodies. Monitoring determines the level of staff satisfaction with the working environment, production stages and processes, the image of the organization, the moral and psychological climate in the team, prospects for career growth and training.

The final evaluation report should note any quality assurance issues, any deviations from expected results, and recommendations for appropriate changes to be made to each process or step based on these deficiencies.

As part of this process, following consultations with users, a preliminary list of statistical indicators containing the information required by them is developed. Also, these statistics and their qualitative characteristics will be agreed with users to ensure that they meet their needs.

When determining the goals for generating statistical data, consideration should be given to the confidentiality of primary data and the availability of resources for their generation.

In this process, methods and tools are specified that are exactly acceptable for collecting primary data. Work within this process is planned depending on the type of tools or data collection methods (electronic reports, electronic surveys, paper questionnaires, administrative databases, web scraping technologies, technologies for geospatial data, etc.).

This process includes planning methods and tools for data collection, questions and answers on observation forms, planning any formal agreements related to the provision of data, including memorandums of cooperation, and determining methods for collecting primary data.

Activities for the formation or improvement of data dissemination components are carried out in accordance with those planned in process 2.1. “Planning of output statistical data” works on the formation of new and (or) improvement of existing components and services for the dissemination of official statistical data.

This process includes the formation of new and (or) further strengthening of all existing components and services for the dissemination of official statistical data, in particular, starting from those used in the production of traditional paper publications, ending with those that provide the establishment of web services, the production of open data.

In this process, primary data is collected. Data collection is organized by obtaining primary data from respondents in paper or electronic form, conducting surveys of respondents by specialists conducting observation. As part of this process, primary data are entered into information systems and special software, primary data are entered by specialists using mobile devices at the place of direct observation, and the submission of data by respondents is taken into account.

The collection of primary data includes the following steps: collection of primary data in electronic and (or) paper form; collection of primary data in other ways; collecting administrative data from administrative data providers; verification of completed and submitted forms of statistical observation; keeping records of the submission of primary data by respondents. This process may also include monitoring the data collection process and making any necessary changes to the data collection process.

The error checking and data validation process examines data to identify potential problems, errors and inconsistencies, including outliers or outliers, unanswered questions, and coding errors. Audits document that the established data production processes, including data collection and processing, can be carried out efficiently and effectively.

Error checking and data validation can be applied to data from any type of source (primary data, administrative data, etc.), including before or after their integration, as well as to imputed data in the "Data Editing and Imputation" process. While this process is about identifying and correcting actual or potential errors, any corrections that make significant changes to the data are made as part of the Data Editing and Imputation process.

In the process of preparing interpretations and explanations for statistical materials, their deep understanding is achieved and the basis for the correct interpretation of official statistics is created.

Interpretation, clarification and correct interpretation of statistical materials are provided through a comprehensive review of statistical data using various mechanisms and approaches, an assessment of how well they reflect pre-planned and expected results, as well as in-depth statistical analysis. Advanced statistical analysis includes: time series analysis; analysis of compatibility and comparability of data; analysis of cases of data revision (analysis of differences between the data of the initial assessment and the revised data); analysis of asymmetries (differences with data from other countries) of data, etc.

As part of managing the dissemination of official statistical data, all conditions are ensured for organizing the publication and dissemination of official statistical data, including compliance with the terms, forms and methods of their publication. This process includes holding press conferences, briefings for specific user groups such as the media or relevant stakeholders, and the imposition of any prohibitions on the disclosure of official statistics before they are released.

Also in this process is the following: delivery of official statistical data to users (in paper and electronic form) based on approved lists; publication of open statistics on the websites of Producers of Official Statistics; publication of statistical data in the media; providing users with access to an open database available on the websites of producers of official statistics in the "on-line" mode; provision of official statistical data in accordance with bilateral agreements and treaties, at the request of users; creating opportunities for the use of confidential statistical data for research purposes; exchange of statistical data at the international level.

In the process of determining the future action plan, work is carried out to summarize all the necessary instructions (instructions) for the development, coordination and approval of future action plans by analyzing the existing gaps in the system, based on the report generated by the results of the production cycle assessment.

In this process: analyzes the final information on the assessments and monitoring carried out; proposals are prepared for further improvement of statistical activities, taking into account the opinions and suggestions of users and respondents; a plan of future actions is developed, agreed with stakeholders and approved based on analysis and proposals.

This process may also include the creation of opportunities to monitor the impact of the approved future action plan on the stages and processes of statistical production. In turn, the results of this monitoring can be a source of information necessary for its evaluation when repeating this stage in the future.

In this process, a list of required data is determined, corresponding to the satisfaction of user needs, i.e. a list of statistical data to be provided, from their point of view.

At the same time, taking into account the identified needs of users, consultations and agreements with them, a draft statistical program for the next year is formed, containing the required statistical indicators, as well as draft forms of statistical observations, which are its main source of information.

Population and sampling planning apply only to processes carried out on the basis of a specially formed sample, including those related to the collection of data through statistical observations. This process is used to create a sampling plan suitable for conducting statistical surveys in a particular area or area and, if necessary, to create a register that is the basis for sampling, determines the most appropriate criteria and methods (complete or sample observations) of sampling.

Sampling sources are administrative and statistical registers, census data and information obtained in the course of other sample surveys. Where necessary, the Producer of Official Statistics combines these sources.

In the process of layout (location) of production processes, the layout of production systems and technologies used within the production cycle is ensured, from the collection of primary data to the dissemination of official statistics. In this process, in particular, work is being carried out to transform existing processes, systems and technologies for statistical observations, collection, processing, analysis, dissemination of data or any of them, combining several processes of various stages, setting up systems.

This process ensures the systematic functioning of the practical production processes created in process 2.6 “Planning production systems and processes” for the specified purposes.

The process of completing the collection of primary data implies the completion of the collection of primary data in electronic and (or) paper form for the corresponding reporting period according to a certain form of statistical observation, control of the completeness of data coverage, as well as input (loading) of the collected primary data into the appropriate electronic software for processing.

Entering (loading) data into electronic software can be done manually or automatically.

The formation of data reflecting the completeness of their coverage for a specific form of statistical observation for the corresponding reporting period, as well as the primary database, electronic administrative databases and (or) summary tables indicates the completion of this process. When controlling the completeness of the submission of statistical observation forms, it is established whether data have been received from all planned statistical units. In the implementation of such control, it is possible to return to process 4.3 “Conducting the collection of primary data”.

If the data is found to be inaccurate, missing, unreliable or out of date, then new values (indicators) can be added using various methods as part of this process, or obsolete data can be removed.

When it is necessary to make changes to primary data and (or) administrative data, as a rule, respondents and providers of administrative data are contacted and the changed data are re-obtained from them. The specific actions implemented in this process are: determination of the need to add new data or change them; selection of the editing method to be used; adding new or changing existing values (indicators); entering new values (indicators) into the data set and marking them as changed; updating existing or developing new metadata based on the results of editing and imputation.

The primary data identification countermeasures process ensures that confidentiality rules are respected in the dissemination of statistical data and metadata in accordance with certain guidelines and work organization rules or methodologies developed in process 2.5. “Planning processing and analysis processes”.

This process includes work to verify that the output data (tables, files and other statistical materials) comply with confidentiality requirements, that the prepared data does not directly or indirectly identify individuals or legal entities and does not disclose individual data about them, the exclusion of such data in if they exist, creating any barriers to the disclosure of data. The degree and method of counteracting identification varies depending on the type or category of statistical indicators.

Advertising for the dissemination of official statistics is carried out with the aim of maximizing their distribution to the general public, expanding the circle of users, increasing user satisfaction and confidence in official statistics.

This process includes the use of various means of communication with potential users for more targeted work, as well as the use of websites, media, social networks, e-mail, etc. to facilitate the process of informing users about statistics. In particular, to strengthen relations with users, measures should be taken to inform them about the availability of new statistical publications and materials, to create favorable conditions for users to receive statistical publications and materials.

Also, the obligatory tools for advertising official statistical data are the placement of news, advertisements and individual headings on websites on the Internet, the creation of a catalog of statistical publications, the establishment of public relations, including the organization of meetings with the media, interested departments, holding presentations, exhibitions, seminars and round tables.

In order to determine the correspondence of the primary sources used to the needs, clarify the conditions and methods for obtaining them, taking into account any restrictions on the use of data, as well as cases of duplication of indicators, the availability of data offered by users is checked by producers of official statistics.

This process also assesses the legal framework governing the collection and use of primary data. Based on the results of the assessment, if necessary, proposals will be considered for changing existing and adopting new legal acts, establishing cooperation with providers of administrative data, and making changes to the forms of statistical observations.

When planning the processing and analysis of the collected data, a methodology for processing statistical data is developed, which will be necessary for use at the stages “Data processing” and “Data analysis”. At the same time, based on the methods and sources of primary data collection, the methods and rules for coding, editing, imputation, and calculation of the obtained data are specified.

This process also includes the development of methods for combining data from different sources, checking, validating, evaluating and finalizing them in a generalized form.

The process of testing production systems provides for testing the suitability of tools, information systems and special software related to the collection, processing, analysis and dissemination of data generated and (or) improved in processes 3.1, 3.2 and 3.3 of the stage “Formation of production processes”, checking their interaction and well-coordinated work in general.

During testing, the causes of errors and failures in systems and software are identified, the problems that have arisen in them are studied, analyzed and eliminated, decisions are made to improve information systems or special software and increase the efficiency of the system as a whole, and its complex debugging is carried out. This process ensures the harmonious operation of all decisions taken on the organization of production.

As part of the process of forming new indicators and statistical units, a calculation is made for indicators and statistical units that are not clearly distinguished at the stage of collecting primary data, but are necessary for the production of the required (necessary) data.

In this process, new indicators are formed by applying arithmetic formulas in relation to one or more indicators that exist in the data set, or using various model hypotheses. Also in this process, new statistical units may be formed by aggregating or splitting data related to statistical units or by using various other estimation methods. In particular, if the units of observation are individuals, then by summarizing the data relating to them, it will be possible to generate data for individual households, and if the units of observation are legal entities, then it will be possible to obtain data for individual enterprises.

As part of this process, the compliance of the generated statistical data with the intended purpose, quality level requirements and their readiness for use is checked. The process of completing the production of official statistics includes: verification of data compliance with the intended purpose; determining at what level (section) the data will be published and additional explanations will be applied to them; сompilation of supporting information, including interpretations, comments, technical notes, information notes and other necessary metadata; preparation of supporting internal documents; discussing previously published data with relevant industry experts; translation of statistical materials distributed in other languages; approval of the final content (content) of statistical material (collection, table, press release, etc.) for publication (distribution).

As part of the user support process, registration of requests and orders of individuals and legal entities for the provision of statistical data is ensured and responses are sent to them in a timely manner. Similar requests and orders should be reviewed regularly to inform quality management processes as they reflect new and changing user needs.

The answers provided can also be posted in the Questions and Answers section of the website, as the availability of this section to all users reduces the need to answer similar and repetitive questions from other users. In this process are carried out: receiving and registering user requests; consideration of user requests and decision on the provision (or non-provision) of data; preparing and sending responses to user requests. This process also includes support for any partner organizations involved in the dissemination of official statistics.

The process of developing a statistical data production model includes documenting the results of all processes of the “Statistical data needs study” stage in the form of a model for the approval of the statistical data production process in a new or modified form.

The following steps are taken in this process: determination, coordination and approval of forms of statistical observations, determination and coordination of other administrative sources, which are the main source in the formation of official statistical data; organizing the development, coordination and approval of the draft annual statistical program, which determines the implementation of the processes of production of statistical data and its application for execution.

This process includes the planning of all processes required for the full production cycle of official statistics, in particular from the collection of primary data to the dissemination of official statistics. The correct organization of this process ensures that all processes within the production cycle are carried out efficiently without any drawbacks or excessive efforts.

When planning, the priority should be the systematic application of existing systems and technologies in a specific production process, the possibility of making new decisions when deficiencies are identified. Also, within the framework of this process, the work of personnel with production systems is analyzed, the performers who are responsible for the implementation of the planned tasks for the organization of production are clearly defined.

Testing of statistical processes includes work on practical testing, pilot use of production processes. As a rule, in order to ensure that statistical processes function in accordance with the expected goals, this process involves the collection of a small amount of data, as well as their processing and analysis to test data collection mechanisms.

After practical tests have been carried out and (or) used as pilots, it may be necessary to return to the adjustment, adjustment and improvement of these processes. This process allows you to make adjustments to the mechanisms, systems and components used in the production process.

In the process of calculating data weights, weights are calculated for individual data arrays according to the statistical methodology developed in process 2.5 “Planning processing and analysis processes”. In other words, cost indicators are formed that characterize the significance of each individual element of the observed phenomenon in their total amount.

Weighting factors are used to recalculate individual data calculated against an aggregate indicator in order to ensure their representativeness in the target population (for example, in the case of sample observations) or to correct totals in case of incomplete reporting of necessary data. In other cases, to manage mutually inconsistent indicators, a method of weighting against these indicators may be required.

In this process, measures are being taken to put into operation regulated and systematized production mechanisms, systems and technologies, including improved and newly created services. These activities include: preparation of technical and operational documentation for the components of production systems, creation of manuals for system users; training of users on the functioning of the production process; commissioning of production mechanisms, information systems and software.

The aggregated data calculation process calculates aggregated data from microdata or other individual data and final aggregated data for the target population. This process involves aggregation, i.e. summarizing data from accounts (primary data, administrative data) that have certain common characteristics (for example, by demographic or geographical classifications), as well as determining averages and dispersion indicators, weighting.

Data extrapolation work may also be included in this process. Various methods are used to assess the accuracy of indicators for the general population (distribution of sample observation data to the general population or the use of correction factors to refine the continuous observation).

As part of the data generation completion process, the results of other processes in the Data Processing phase are brought together into a single data set and used as the primary source in the Data Analysis phase.

At times, this process can be more intermediate than final, especially in processes that “suffer” from a significant time pressure for data production, as well as in production processes where the production of preliminary and final estimates is agreed in advance. The end result of this process is aggregated data generated and validated in the Data Processing phase.