Statistical data are the results of statistical work and the objects of statistical methods and theories. Statistical data, also known as statistical information or quantitative information, usually refers to the data obtained by counting and measuring a social phenomenon or natural phenomenon under a specific time and space condition according to the characteristics (attributes and quantity) of individuals in the population. Therefore, statistical data have three elements: time, space and data, which are indispensable. Statistical data can be classified as follows.
(1) According to different measurement methods, it can be divided into point data and measurement data. The digital data collected by counting the number of points without measuring tools are called point counting data. When counting the number of points, we need to classify things according to certain attributes first, and then count the number of points, so it is also called attribute data or discrete data. The data obtained by measuring with certain measuring tools (weights and measures, monetary scales, working hours, etc.) are called measurement data, such as length, weight, time, value, etc.
(2) According to whether the data is directly obtained, it can be divided into original data and secondary data. The statistical data to be processed, which are obtained by the investigators through observation, counting, measurement, experiment or registration directly from the data source, are called original data. The ready-made data that have been processed, transformed from individual to overall, and can explain the quantitative characteristics of the overall phenomenon to a certain extent, are called secondary data (indirect data, secondary data). Such information is usually taken from government agencies, various yearbooks and various public data.
(3) According to the time attribute of statistical data, it can be divided into static data and dynamic data. The data representing the static state of phenomenon in a specific time and space phase I is called static data or cross-sectional data, such as the classification data of urban residents' living expenses in a certain year, and the purchase, sales, and storage data of each counter group in a certain month of a shopping mall. The data representing the evolution of phenomena in a specific period are called dynamic data or time series data, which are the results of long-term registration of static data arranged in chronological order. For example, the income and expenditure data of living expenses of urban residents in a city over the years, and the statistical data of purchase, sales and storage of a mall over the years.
(4) According to the scope of statistical data, it can be divided into comprehensive data and sample data. Comprehensive data refers to the data obtained by investigating all individuals in the study population, such as census data and comprehensive statistical statement data. Sampling data refers to the data obtained by spot checking individuals within the research population, such as sampling inspection data of industrial product quality, sampling survey data of agricultural product output, etc. [2]
The characteristics of statistical data are the characteristics of statistical information statistical method The data collected and described about the quantitative characteristics of the overall phenomenon are statistical information 。 Statistical data or information have the following significant characteristics. (1) Objectivity. Statistical data must be concrete facts obtained from observation, investigation, experiment or registration, not fabricated data, so statistical information is objective.
(2) Overall. Statistical data is a description of the overall quantitative performance of social or natural phenomena, rather than the quantitative characteristics of individuals, so statistical information is overall.
(3) Quantification. Statistics are generally quantitative information, which can show the overall quantitative performance of the study under certain time and space conditions, including quantity, quantitative relationship and quantitative boundaries.
(4) Scalability. Any statistical data or information can be expanded in terms of time, space, structure and relevance, so that the statistical information can be enriched, systematic and complete. [2]
In general, statistical data sorting should go through the following four steps:
1. Review the original data. Some errors are inevitable in the process of statistical investigation due to various reasons. In order to ensure the original quantity of statistical data, the survey data must be reviewed before entering the data processing process. The content of the audit is mainly the integrity, timeliness and accuracy of the data.
2. Data classification and summary. Through the keyboard Scanner 、 Tape drive And record the data into the Magnetic medium And then use it again Statistical database Software. 3. Making Statistical table Or draw Statistical chart 。 This is the main form to show the results of statistical arrangement. The drawing of statistical charts is to express the results of statistical arrangement in the form of histogram, line chart, curve chart, fan chart and other intuitive forms, which is widely used in the propaganda and education of the masses. 4. Storage and release of data. Network technology enables the storage and release of statistical data from traditional paper pages to disk information and then to web information, realizing paperless storage and online access of statistical data. Because of the strong interactivity of the network, the public can not only obtain and transmit the statistical information of the region, the country and even the world, but also obtain and transmit personalized statistical data to each other at any time, anywhere, with a mouse click. [3]
In order to ensure the quality of statistical data and prevent errors, statistical data must be reviewed repeatedly. The review of survey data before collection is one of the essential contents of statistical sorting. The main content of the audit is to check the integrity, timeliness and accuracy of the data.
The review of the integrity of the investigation data refers to the inspection of the investigation data for any omission, including whether the items in the questionnaire have been answered, whether the investigation unit is complete, and whether there is any omission or missing report.
The review of the timeliness of the investigation data refers to the inspection of whether the investigated units have submitted the data according to the specified date, and the inspection of the reasons for failing to submit the data as required.
The review of the accuracy of the survey data is to check whether the answers in the questionnaire meet the requirements and whether there are errors. When auditing, you can generally check from two aspects: logic check and calculation check. Logic check is to check whether the contents of the investigation data are reasonable, whether there are contradictions between the items, and check the balance between the figures against the relevant data to reveal whether there are contradictions in logic. The calculation check is to check whether there is any error in the indicator caliber, calculation method and calculation result of each number in the questionnaire, and whether the calculation unit is inconsistent with the provisions.
Through inspection, in case of any missing report, missing part or missing item, it is necessary to immediately urge and supplement the report. In case of any inaccuracy or doubt, it is necessary to deal with it according to different circumstances.
(1) For general errors that can be confirmed, correct them on behalf and notify the filling unit to check them.
(2) If there is something suspicious or cannot be corrected on behalf of the applicant, the applicant shall inquire the original reporting unit.
(3) The investigation materials with serious errors shall be returned and reported again. [3]
