Why do we use statistics in business

statistics

term

Statistics as a field of knowledge and science offers and develops a comprehensive methodological-quantitative instrument for the description and evaluation of empirical findings with universal application possibilities in politics, economy and society and all humanities, social and natural sciences including medicine and technology, in which with numbers or Reviews is being worked on. The results of statistical studies are also referred to as statistics. Statistics is also used as a synonym for Estimator.

Objectives and objects

Statistical methods are aimed at describing repeatedly observable processes under the same general conditions with the help of measures or evaluations. Typically, statistical investigations are aimed at describing (describing) or exploring (exploring) a suitably delimited population of units (e.g. the population of a country, the daily production of a manufacturing plant) or, in particular, drawing conclusions about the structures in a population (induction; inference ) to enable. If the population is completely empirically recorded (full survey), the statistical methods concentrate on the description of the findings, i.e. the transfer of the empirically obtained individual information into globally characteristic statements (descriptive statistics). For this purpose, frequency distributions, mean values, proportional values, measures of dispersion, correlation coefficients and tabular and graphic display methods are used. The assessment of a population is successful even if it was only partially empirically recorded, provided a random sample (representative sample), i.e. a subpopulation obtained through a random selection process, is available. This assessment by means of random sampling is done by point and interval estimation of parameters of the population and by testing hypotheses about such variables (inferential statistics). The inferential statistics are based on the calculation of probability and therefore provide probability-based statements. Exploratory data analysis occupies a position between descriptive and inductive statistics. It is designed as a search process for structures and special features in the data, which is intended to uncover new questions or hypotheses, and is usually not possible without intensive use of a computer.

Statistical methods that are primarily used to answer substance-scientific questions (marking the development of a price level, the fertility of a population, determining the gross national product of a country) are placed alongside methodological statistics as material statistics. Within these sub-areas, e.g. population, economic, company statistics or special price, income, productivity or university statistics are differentiated.

Econometrics is closely linked to statistics. It focuses on the modeling and statistical characterization of economic phenomena. The comprehensive use of point and interval estimation as well as hypothesis testing is essential in econometrics because only random sample results can be available for economic processes. Statistics is therefore the most important component of econometrics alongside economic theory, the material part of which is referred to as empirical economic research or applied econometrics.

In decision theory, the probability-based methods of statistics are of particular importance for the information situation uncertainty and risk, for example in insurance, in the analysis of financial markets or in creditworthiness assessments in banking.

Steps in statistical investigations

The starting point is an investigation goal, e.g. identification of the employee structure of a company, which must first be statistically specified by defining the investigation variables to be taken into account, such as age, level of training, length of membership and the relationships to be investigated. The population, e.g. the workforce of a company, is to be delimited objectively, spatially and temporally. In the survey, i.e. the empirical determination of the values ​​of the examination variables, questions of costs are directly linked to the quality of the data to be achieved. When evaluating documents that actually serve other purposes (secondary statistics), e.g. personnel files, greater profitability is often bought with an imperfect focus on the research objective. Even in the case of written or oral surveys or observations specifically for the purposes of a statistical investigation (primary statistics), the correct recording of the target population (problems of non-recording and multiple recording) and the correct determination of the variable values ​​(entry errors, refusal to provide information) are prerequisites for usable results. The determined data are presented in a clear form using tables and graphics. Suitable computer programs are essential for this preparation. The processed results are finally fed into the analysis, whereby in each case methods of description (determination of parameters including correlation coefficients) are at the beginning. The procedure is analogous if, for example, a statistical quality control or a marketing study is to be carried out.

Representative statistics

If a partial survey (sample survey) is carried out, for example from 400 people who belong to a workforce of 12,000 people in a company, then the selection of the samples must be random, the sample results should be transferred as estimates to the population or hypotheses about the population can be tested . Numerous sampling methods are available for the randomness of the selection, each of which enables an adequate transfer of the sample results to the population.

The problem areas of determining the examination variables in view of the objective of the examination, the correct delimitation of the population, the correct collection of the data, the error-free processing and the analysis methods that meet the objective of the examination as well as, in the case of representative examinations, the random selection of the sample units are considered adequacy problems of the statistics (adequation). summarized. Your solution is directly linked to the quality of the results achieved.

Subareas of methodological statistics

Measures for the characterization of frequency distributions, especially mean values, measures of dispersion and correlation coefficients, which can be viewed from a descriptive and inferential perspective, are of universal application importance. The basis of inferential statistics is the calculation of probability. In sample theory, models for the optimal random selection of subpopulations with their essential inferential and organizational implications are considered. The theory of the estimator functions (estimator function) deals with the suitability of measures obtained from random samples for parameter estimation. In the theory of hypothesis testing (statistical test procedures), hypotheses developed based on substance science about parameters or distribution functions of basic populations are statistically checked, accepting a small probability of error (level of significance). The statistical decision theory forms the theoretical superstructure of the estimation and hypothesis testing and can also be understood as part of the (more general) decision theory. In this context, decisions under uncertainty with full or partial knowledge of probability distributions over the set of environmental conditions are of particular importance. The analysis of the connection between several statistical variables takes place in the regression analysis, in which the functional form of randomly disturbed causal relationships is in the foreground, as well as in the correlation analysis, in which the intensity of the statistical connection between variables is recorded. The methods of time series analysis, which can be used for forecasts, are of particular importance in economics. The dynamic analysis of numerous economic and demographic processes is based on the theory of stochastic processes. Numerous methods of multivariate statistics (cluster analysis, factor analysis, discriminant analysis, variance analysis) are used in marketing.

The subject of population statistics are the methods and results of recording and forecasting population movements through births, marriages, mortality and migration.

Of primary importance in economic statistics are the national accounts, which also include the statistical problems of macroeconomic aggregate variables, the input-output analysis, the theory of index numbers (index number), e.g. price index numbers, as well as the statistical analysis of economic concentration.

In operational statistics, the focus is on methods of progress statistics, sampling procedures in accounting, work process analysis (multi-moment procedures; multi-moment time studies) and ongoing control and monitoring of the quality of manufacturing processes.

Institutions of official statistics

The principles of technical centralization and regional decentralization apply to official statistics in Germany. In addition to the statistics that are independent from the subject (triggered statistics), there are also departmental statistics by administrations if statistical documents arise during administrative processes or if their processing cannot be separated from the administrative process.

The statistic, which is independent from the subject, is carried out by the Federal Statistical Office (StBA), the State Statistical Offices and the statistical offices and offices of the municipalities, especially the large cities. The Federal Office in Wiesbaden, an independent federal authority under the Ministry of the Interior, is primarily responsible for preparing and further developing statistics for federal purposes, working towards the uniform implementation of the programs by the federal states, compiling and publishing the results, and presenting statistics from the EU and international organizations and conduct national accounts. The main part of the implementation work is incumbent on the state offices; they influence the federal statistics via the statistical advisory board at the Federal Statistical Office.

The legal basis of official statistical activity is derived from Art. 73 No. 11 and Art. 83 GG. They result from the law on statistics for federal purposes (BStatG) of January 22, 1987, from the Federal Data Protection Act (BDSG) of January 14, 2003 and from state laws such as state data protection laws. The implementation of federal statistics must usually be ordered by a federal law.

Sources of official statistical data

These are published by the Federal Statistical Office, the statistical offices of the federal states, the statistical offices of other countries and international organizations (e.g. the Statistical Office of the European Community). The Federal Statistical Office mainly publishes the Statistical Yearbook as well as Economy and Statistics (published monthly with a text and a number part). Much information is also made available online. In addition to these summarizing publications, i.e. covering all fields of activity of official statistics, publications within specialist series that are characterized by thematic specialization take place in a less regular sequence. Numerous data can also be called up on the homepage of the Federal Statistical Office.

training

Statistics is the subject of or part of numerous courses of study at universities, technical colleges and vocational academies.

courses

Statistics can be studied in the Bachelor's and Master's area at several German universities (e.g. at TU Dortmund).

Statistics is a core component of almost all economic and social science courses, such as the Bachelor and Master courses in economics, business administration, business education, psychology and sociology, as well as various scientific, technical and medical courses. In addition, statistics (occasionally including, occasionally in addition to econometrics), for example in the bachelor's and master's degree programs in economics, can be selected as a specialization area at some locations. Analogous regulations apply to various natural science courses. In the mathematics course in particular, statistics is one of the various possible focal points, especially in the last phase of the bachelor’s degree and in the master’s degree as mathematical statistics (stochastics).

In the business administration departments of the universities of applied sciences and the vocational academies, compulsory basic training in statistics has also been introduced as a rule.

Course content

Particularly in the economics courses at universities, the basic training includes parts of descriptive statistics, business statistics, population statistics as well as probability calculations and the essential basics of inferential statistics.

Building on the basic training, the areas of economic statistics, history and population statistics, sampling procedures, regression and correlation analysis, estimation and test methods, econometrics and applied economic research, in particular econometrics of the financial markets, time series analysis, forecasting methods, statistical decision theory, multivariate analysis methods, non-distribution test methods and represent stochastic processes.