Ekonomska istraživanja se najčešće moraju oslanjati na raspoložive podatke koji se dobijaju iz različitih izvora: statistički popisi, knjigovodstveni izveštaji, ankete, zakoni, tehničke norme. Često se dešava u praksi da su neki podaci nedostupni i ne mogu se obezbediti dodatnim statističkim istraživanjima. Tada se po posebnom postupku dolazi do približnih vrednosti ovih podataka.
Sa stanovišta jedinica osmatranja podaci se dele na:
- makroekonomske – odnose se na nacionalnu privredu ili grupu nacionalnih privreda, regionalne jedinice, delatnosti ili grane privrede,
- mikroekonomske – odnose se na privredne jedinice, a to su obično preduzeća, proizvodi, domaćinstva lil pojedinci.
Sa stanovišta vremenskih dimenzija podaci mogu da budu:
- Vremenske serije su podaci koji se odnose na uzastopne vremenske jedinice jednakih razmaka. Dele se na intervalne (dnevno, mesečno…) i momentne (pokazuju stanje neke veličine određenog trenutka).
- Podaci uporednih preseka sadrže uzorke pojedinaca, domaćinstava, firmi, gradova, regija, država, ili nekih drugih jedinica, u jednom vremenskom trenutku.
- Panel podaci sadrže vremenske serije za sve podatke uporednih preseka u datom skupu podataka. Na primer, skup podataka o platama, obrazovanju i zaposlenju jedne skupine pojedinaca u periodu od 10 godina.
Economic research usually has to rely on available data obtained from various sources: statistical censuses, accounting reports, surveys, laws, technical norms. It often happens in practice that some data are unavailable and cannot be provided by additional statistical research. Then, following a special procedure, the approximate values of these data are arrived at.
From the point of view of observation units, the data are divided into:
- macroeconomic – refer to the national economy or a group of national economies, regional units, activities or branches of the economy,
- microeconomic – refer to economic units, which are usually companies, products, households and individuals.
From the point of view of time dimensions, the data can be:
- Time series are data that refer to consecutive time units at equal intervals. They are divided into interval (daily, monthly…) and instantaneous (showing the state of some quantity at a certain moment).
- Cross-sectional data contain samples of individuals, households, firms, cities, regions, countries, or some other units at one point in time.
- Panel data contain time series for all cross-sectional data in a given data set. For example, a set of data on wages, education and employment of one group of individuals over a period of 10 years.