{"id":28012,"date":"2023-05-24T10:07:19","date_gmt":"2023-05-24T13:07:19","guid":{"rendered":"https:\/\/mindthegraph.com\/blog\/?p=28012"},"modified":"2023-05-24T10:07:21","modified_gmt":"2023-05-24T13:07:21","slug":"sampling-bias","status":"publish","type":"post","link":"https:\/\/mindthegraph.com\/blog\/cs\/vyberove-zkresleni\/","title":{"rendered":"Probl\u00e9m zvan\u00fd zkreslen\u00ed v\u00fdb\u011bru vzorku"},"content":{"rendered":"<p>Bez ohledu na pou\u017eitou metodiku nebo studovan\u00fd obor mus\u00ed v\u00fdzkumn\u00ed pracovn\u00edci zajistit, aby pou\u017e\u00edvali reprezentativn\u00ed vzorky, kter\u00e9 odr\u00e1\u017eej\u00ed charakteristiky studovan\u00e9 populace. Tento \u010dl\u00e1nek se bude zab\u00fdvat konceptem v\u00fdb\u011brov\u00e9ho zkreslen\u00ed, jeho r\u016fzn\u00fdmi typy a zp\u016fsoby uplatn\u011bn\u00ed a osv\u011bd\u010den\u00fdmi postupy pro zm\u00edrn\u011bn\u00ed jeho \u00fa\u010dink\u016f.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Co je to v\u00fdb\u011brov\u00e9 zkreslen\u00ed?<\/h2>\n\n\n\n<p>V\u00fdb\u011brov\u00e9 zkreslen\u00ed ozna\u010duje situaci, kdy jsou n\u011bkte\u0159\u00ed jedinci nebo skupiny v populaci za\u0159azeni do vzorku s v\u011bt\u0161\u00ed pravd\u011bpodobnost\u00ed ne\u017e ostatn\u00ed, co\u017e vede ke zkreslen\u00ed nebo nereprezentativnosti vzorku. K tomu m\u016f\u017ee doj\u00edt z r\u016fzn\u00fdch d\u016fvod\u016f, nap\u0159. v d\u016fsledku nen\u00e1hodn\u00fdch metod v\u00fdb\u011bru vzorku, zkreslen\u00ed vlastn\u00edho v\u00fdb\u011bru nebo zkreslen\u00ed v\u00fdzkumn\u00edka.<\/p>\n\n\n\n<p>Jin\u00fdmi slovy, v\u00fdb\u011brov\u00e9 zkreslen\u00ed m\u016f\u017ee ohrozit platnost a zobecnitelnost v\u00fdsledk\u016f v\u00fdzkumu t\u00edm, \u017ee vzorek je zkreslen ve prosp\u011bch ur\u010dit\u00fdch charakteristik nebo hledisek, kter\u00e9 nemus\u00ed b\u00fdt reprezentativn\u00ed pro \u0161ir\u0161\u00ed populaci.&nbsp;<\/p>\n\n\n\n<p>V ide\u00e1ln\u00edm p\u0159\u00edpad\u011b mus\u00edte v\u0161echny \u00fa\u010dastn\u00edky pr\u016fzkumu vybrat n\u00e1hodn\u00fdm zp\u016fsobem. V praxi v\u0161ak m\u016f\u017ee b\u00fdt obt\u00ed\u017en\u00e9 prov\u00e9st n\u00e1hodn\u00fd v\u00fdb\u011br \u00fa\u010dastn\u00edk\u016f kv\u016fli omezen\u00edm, jako jsou n\u00e1klady a dostupnost respondent\u016f. I v p\u0159\u00edpad\u011b, \u017ee neprovedete n\u00e1hodn\u00fd sb\u011br dat, je nezbytn\u00e9 si uv\u011bdomit potenci\u00e1ln\u00ed zkreslen\u00ed, kter\u00e9 by se mohlo v datech vyskytnout.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Mezi p\u0159\u00edklady zkreslen\u00ed v\u00fdb\u011bru pat\u0159\u00ed:<\/h3>\n\n\n\n<ol>\n<li><strong>Dobrovolnick\u00e1 zaujatost<\/strong>: \u00da\u010dastn\u00edci, kte\u0159\u00ed se dobrovoln\u011b z\u00fa\u010dastn\u00ed studie, mohou m\u00edt jin\u00e9 charakteristiky ne\u017e ti, kte\u0159\u00ed se dobrovoln\u011b nez\u00fa\u010dastn\u00ed, co\u017e vede k nereprezentativn\u00edmu vzorku.<\/li>\n\n\n\n<li><strong>V\u00fdb\u011br vzork\u016f bez n\u00e1hodn\u00e9ho v\u00fdb\u011bru<\/strong>: Pokud v\u00fdzkumn\u00edk vyb\u00edr\u00e1 \u00fa\u010dastn\u00edky pouze z ur\u010dit\u00fdch m\u00edst nebo pouze \u00fa\u010dastn\u00edky s ur\u010dit\u00fdmi charakteristikami, m\u016f\u017ee to v\u00e9st ke zkreslen\u00ed vzorku.<\/li>\n\n\n\n<li><strong>Zkreslen\u00ed p\u0159e\u017eit\u00ed<\/strong>: K tomu doch\u00e1z\u00ed, kdy\u017e vzorek zahrnuje pouze jedince, kte\u0159\u00ed p\u0159e\u017eili nebo usp\u011bli v ur\u010dit\u00e9 situaci, a vynech\u00e1v\u00e1 ty, kte\u0159\u00ed nep\u0159e\u017eili nebo neusp\u011bli.<\/li>\n\n\n\n<li><strong>V\u00fdhodn\u00fd odb\u011br vzork\u016f<\/strong>: Tento typ v\u00fdb\u011bru zahrnuje v\u00fdb\u011br snadno dostupn\u00fdch \u00fa\u010dastn\u00edk\u016f, nap\u0159\u00edklad t\u011bch, kte\u0159\u00ed jsou n\u00e1hodou pobl\u00ed\u017e, nebo t\u011bch, kte\u0159\u00ed odpov\u00eddaj\u00ed na online pr\u016fzkum, co\u017e nemus\u00ed reprezentovat \u0161ir\u0161\u00ed populaci.<\/li>\n\n\n\n<li><strong>Potvrzovac\u00ed zkreslen\u00ed<\/strong>: V\u00fdzkumn\u00edci mohou nev\u011bdomky nebo z\u00e1m\u011brn\u011b vyb\u00edrat \u00fa\u010dastn\u00edky, kte\u0159\u00ed podporuj\u00ed jejich hypot\u00e9zu nebo v\u00fdzkumnou ot\u00e1zku, co\u017e vede ke zkreslen\u00ed v\u00fdsledk\u016f.<\/li>\n\n\n\n<li><strong>Hawthornsk\u00fd efekt<\/strong>: \u00da\u010dastn\u00edci mohou zm\u011bnit sv\u00e9 chov\u00e1n\u00ed nebo odpov\u011bdi, kdy\u017e v\u011bd\u00ed, \u017ee jsou zkoum\u00e1ni nebo pozorov\u00e1ni, co\u017e vede k nereprezentativn\u00edm v\u00fdsledk\u016fm.<\/li>\n<\/ol>\n\n\n\n<p>&nbsp;Pokud jste si t\u011bchto zkreslen\u00ed v\u011bdomi, m\u016f\u017eete je p\u0159i anal\u00fdze zohlednit a prov\u00e9st korekci zkreslen\u00ed a l\u00e9pe porozum\u011bt populaci, kterou va\u0161e data reprezentuj\u00ed.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Typy zkreslen\u00ed v\u00fdb\u011bru vzorku<\/h2>\n\n\n\n<ul>\n<li><strong>V\u00fdb\u011brov\u00e9 zkreslen\u00ed<\/strong>: nastane, kdy\u017e vzorek nen\u00ed reprezentativn\u00ed pro populaci.<\/li>\n\n\n\n<li><strong>Zkreslen\u00ed m\u011b\u0159en\u00ed<\/strong>: nastane, kdy\u017e jsou shrom\u00e1\u017ed\u011bn\u00e9 \u00fadaje nep\u0159esn\u00e9 nebo ne\u00fapln\u00e9.<\/li>\n\n\n\n<li><strong>Zkreslen\u00ed p\u0159i pod\u00e1v\u00e1n\u00ed zpr\u00e1v<\/strong>: nastane, kdy\u017e respondenti poskytnou nep\u0159esn\u00e9 nebo ne\u00fapln\u00e9 informace.<\/li>\n\n\n\n<li><strong>Neodpov\u00eddaj\u00edc\u00ed zkreslen\u00ed<\/strong>: nastane, kdy\u017e n\u011bkte\u0159\u00ed \u010dlenov\u00e9 populace na pr\u016fzkum neodpov\u00ed, co\u017e vede k nereprezentativn\u00edmu vzorku.<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\">P\u0159\u00ed\u010diny zkreslen\u00ed v\u00fdb\u011bru vzorku<\/h2>\n\n\n\n<ol>\n<li><strong>V\u00fdhodn\u00fd odb\u011br vzork\u016f<\/strong>: v\u00fdb\u011br vzorku na z\u00e1klad\u011b v\u00fdhodnosti, nikoli na z\u00e1klad\u011b v\u011bdeck\u00e9 metody.<\/li>\n\n\n\n<li><strong>Zkreslen\u00ed vlastn\u00edho v\u00fdb\u011bru<\/strong>: do pr\u016fzkumu jsou zahrnuti pouze ti, kte\u0159\u00ed se ho dobrovoln\u011b z\u00fa\u010dastn\u00ed, co\u017e nemus\u00ed b\u00fdt reprezentativn\u00ed pro celou populaci.<\/li>\n\n\n\n<li><strong>Zkreslen\u00ed v\u00fdb\u011brov\u00e9ho souboru<\/strong>: kdy\u017e v\u00fdb\u011brov\u00fd soubor pou\u017eit\u00fd pro v\u00fdb\u011br vzorku nen\u00ed reprezentativn\u00ed pro populaci.<\/li>\n\n\n\n<li><strong>Zkreslen\u00ed p\u0159e\u017eit\u00ed<\/strong>: kdy\u017e se v\u00fdzkumu \u00fa\u010dastn\u00ed pouze n\u011bkte\u0159\u00ed \u010dlenov\u00e9 populace, co\u017e vede k nereprezentativn\u00edmu vzorku. Nap\u0159\u00edklad pokud v\u00fdzkumn\u00edci prov\u00e1d\u011bj\u00ed pr\u016fzkum pouze u \u017eij\u00edc\u00edch osob, nemus\u00ed z\u00edskat informace od osob, kter\u00e9 zem\u0159ely p\u0159ed proveden\u00edm studie.<\/li>\n\n\n\n<li><strong>Zkreslen\u00ed v\u00fdb\u011bru vzorku v d\u016fsledku nedostate\u010dn\u00fdch znalost\u00ed<\/strong>: nerozpozn\u00e1n\u00ed zdroj\u016f variability, kter\u00e9 mohou v\u00e9st ke zkreslen\u00fdm odhad\u016fm.<\/li>\n\n\n\n<li><strong>Zkreslen\u00ed v\u00fdb\u011bru vzorku zp\u016fsoben\u00e9 chybami p\u0159i spr\u00e1v\u011b vzorku<\/strong>: nepou\u017eit\u00ed vhodn\u00e9ho nebo dob\u0159e funguj\u00edc\u00edho v\u00fdb\u011brov\u00e9ho souboru nebo odm\u00edtnut\u00ed \u00fa\u010dasti ve studii, co\u017e vede k neobjektivn\u00edmu v\u00fdb\u011bru vzorku.<\/li>\n<\/ol>\n\n\n\n<h2 class=\"wp-block-heading\">Zkreslen\u00ed v\u00fdb\u011bru vzork\u016f v klinick\u00fdch studi\u00edch<\/h2>\n\n\n\n<p>Klinick\u00e9 studie slou\u017e\u00ed k testov\u00e1n\u00ed \u00fa\u010dinnosti nov\u00e9 l\u00e9\u010dby nebo l\u00e9k\u016f na ur\u010dit\u00e9 populaci. Jsou nezbytnou sou\u010d\u00e1st\u00ed procesu v\u00fdvoje l\u00e9k\u016f a zji\u0161\u0165uj\u00ed, zda je l\u00e9\u010dba bezpe\u010dn\u00e1 a \u00fa\u010dinn\u00e1 p\u0159ed jej\u00edm uvoln\u011bn\u00edm pro \u0161irokou ve\u0159ejnost. Klinick\u00e9 studie jsou v\u0161ak tak\u00e9 n\u00e1chyln\u00e9 k v\u00fdb\u011brov\u00e9mu zkreslen\u00ed.<\/p>\n\n\n\n<p>K v\u00fdb\u011brov\u00e9mu zkreslen\u00ed doch\u00e1z\u00ed, kdy\u017e vzorek pou\u017eit\u00fd pro studii nen\u00ed reprezentativn\u00ed pro populaci, kterou m\u00e1 reprezentovat. V p\u0159\u00edpad\u011b klinick\u00fdch studi\u00ed m\u016f\u017ee k v\u00fdb\u011brov\u00e9mu zkreslen\u00ed doj\u00edt, pokud jsou \u00fa\u010dastn\u00edci k \u00fa\u010dasti vybr\u00e1ni selektivn\u011b nebo jsou vybr\u00e1ni sami.<\/p>\n\n\n\n<p>\u0158ekn\u011bme, \u017ee farmaceutick\u00e1 spole\u010dnost prov\u00e1d\u00ed klinickou studii s c\u00edlem ov\u011b\u0159it \u00fa\u010dinnost nov\u00e9ho l\u00e9ku proti rakovin\u011b. Rozhodne se z\u00edskat \u00fa\u010dastn\u00edky studie prost\u0159ednictv\u00edm inzer\u00e1t\u016f v nemocnic\u00edch, na klinik\u00e1ch a v podp\u016frn\u00fdch skupin\u00e1ch pro l\u00e9\u010dbu rakoviny a tak\u00e9 prost\u0159ednictv\u00edm online p\u0159ihl\u00e1\u0161ek. Vzorek, kter\u00fd shrom\u00e1\u017ed\u00ed, v\u0161ak m\u016f\u017ee b\u00fdt zkreslen\u00fd sm\u011brem k t\u011bm, kte\u0159\u00ed jsou v\u00edce motivov\u00e1ni k \u00fa\u010dasti ve studii nebo kte\u0159\u00ed maj\u00ed ur\u010dit\u00fd typ rakoviny. To m\u016f\u017ee zt\u00ed\u017eit zobecn\u011bn\u00ed v\u00fdsledk\u016f studie na \u0161ir\u0161\u00ed populaci.<\/p>\n\n\n\n<p>Aby se minimalizovalo zkreslen\u00ed v\u00fdb\u011bru v klinick\u00fdch studi\u00edch, mus\u00ed v\u00fdzkumn\u00ed pracovn\u00edci uplat\u0148ovat p\u0159\u00edsn\u00e1 krit\u00e9ria pro za\u0159azen\u00ed a vylou\u010den\u00ed a postupy n\u00e1hodn\u00e9ho v\u00fdb\u011bru. T\u00edm se zajist\u00ed, \u017ee vzorek \u00fa\u010dastn\u00edk\u016f vybran\u00fdch pro studii je reprezentativn\u00ed pro \u0161ir\u0161\u00ed populaci, co\u017e minimalizuje jak\u00e9koli zkreslen\u00ed shrom\u00e1\u017ed\u011bn\u00fdch \u00fadaj\u016f.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Probl\u00e9my zp\u016fsoben\u00e9 zkreslen\u00edm v\u00fdb\u011bru vzorku<\/h2>\n\n\n\n<p>V\u00fdb\u011brov\u00e9 zkreslen\u00ed je problematick\u00e9, proto\u017ee je mo\u017en\u00e9, \u017ee statistika vypo\u010dten\u00e1 z v\u00fdb\u011brov\u00e9ho souboru je systematicky chybn\u00e1. To m\u016f\u017ee v\u00e9st k systematick\u00e9mu nadhodnocen\u00ed nebo podhodnocen\u00ed p\u0159\u00edslu\u0161n\u00e9ho parametru v populaci. Vyskytuje se v praxi, proto\u017ee prakticky nen\u00ed mo\u017en\u00e9 zajistit dokonalou n\u00e1hodnost p\u0159i v\u00fdb\u011bru vzorku.<\/p>\n\n\n\n<p>Pokud je m\u00edra zkreslen\u00ed mal\u00e1, lze vzorek pova\u017eovat za p\u0159im\u011b\u0159enou aproximaci n\u00e1hodn\u00e9ho vzorku. Pokud se nav\u00edc vzorek v\u00fdrazn\u011b neli\u0161\u00ed v m\u011b\u0159en\u00e9 veli\u010din\u011b, m\u016f\u017ee b\u00fdt zkreslen\u00fd vzorek st\u00e1le rozumn\u00fdm odhadem.<\/p>\n\n\n\n<p>Zat\u00edmco n\u011bkte\u0159\u00ed jedinci mohou z\u00e1m\u011brn\u011b pou\u017e\u00edt neobjektivn\u00ed vzorek, aby dos\u00e1hli zav\u00e1d\u011bj\u00edc\u00edch v\u00fdsledk\u016f, \u010dast\u011bji je neobjektivn\u00ed vzorek pouze odrazem obt\u00ed\u017e\u00ed p\u0159i z\u00edsk\u00e1v\u00e1n\u00ed skute\u010dn\u011b reprezentativn\u00edho vzorku nebo neznalosti zkreslen\u00ed v jejich procesu m\u011b\u0159en\u00ed nebo anal\u00fdzy.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Extrapolace: mimo rozsah<\/h2>\n\n\n\n<p>Vyvozov\u00e1n\u00ed z\u00e1v\u011br\u016f o n\u011b\u010dem, co p\u0159esahuje rozsah dat, se ve statistice naz\u00fdv\u00e1 extrapolace. Jednou z forem extrapolace je vyvozov\u00e1n\u00ed z\u00e1v\u011br\u016f z neobjektivn\u00edho vzorku: proto\u017ee metoda v\u00fdb\u011bru systematicky vylu\u010duje ur\u010dit\u00e9 \u010d\u00e1sti zkouman\u00e9 populace, z\u00e1v\u011bry se vztahuj\u00ed pouze na vybranou subpopulaci.<\/p>\n\n\n\n<p>K extrapolaci doch\u00e1z\u00ed tak\u00e9 tehdy, kdy\u017e se nap\u0159\u00edklad z\u00e1v\u011br zalo\u017een\u00fd na vzorku vysoko\u0161kol\u00e1k\u016f aplikuje na star\u0161\u00ed dosp\u011bl\u00e9 nebo na dosp\u011bl\u00e9 s pouze osmilet\u00fdm vzd\u011bl\u00e1n\u00edm. Extrapolace je \u010dastou chybou p\u0159i pou\u017e\u00edv\u00e1n\u00ed nebo interpretaci statistiky. N\u011bkdy je extrapolace z d\u016fvodu obt\u00ed\u017enosti nebo nemo\u017enosti z\u00edskat kvalitn\u00ed \u00fadaje to nejlep\u0161\u00ed, co m\u016f\u017eeme ud\u011blat, ale v\u017edy je t\u0159eba ji br\u00e1t p\u0159inejmen\u0161\u00edm s rezervou - a \u010dasto s velkou d\u00e1vkou nejistoty.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Od v\u011bdy k pseudov\u011bd\u011b<\/h2>\n\n\n\n<p><a href=\"https:\/\/en.wikipedia.org\/wiki\/Sampling_bias\">Jak je uvedeno na Wikipedii<\/a>, p\u0159\u00edkladem toho, jak m\u016f\u017ee existovat neznalost zkreslen\u00ed, je roz\u0161\u00ed\u0159en\u00e9 pou\u017e\u00edv\u00e1n\u00ed pom\u011bru (tzv. fold change) jako m\u00edry rozd\u00edlu v biologii. Proto\u017ee je snaz\u0161\u00ed dos\u00e1hnout velk\u00e9ho pom\u011bru u dvou mal\u00fdch \u010d\u00edsel s dan\u00fdm rozd\u00edlem a relativn\u011b obt\u00ed\u017en\u011bj\u0161\u00ed dos\u00e1hnout velk\u00e9ho pom\u011bru u dvou velk\u00fdch \u010d\u00edsel s v\u011bt\u0161\u00edm rozd\u00edlem, mohou b\u00fdt p\u0159i porovn\u00e1v\u00e1n\u00ed relativn\u011b velk\u00fdch \u010d\u00edseln\u00fdch m\u011b\u0159en\u00ed p\u0159ehl\u00e9dnuty velk\u00e9 v\u00fdznamn\u00e9 rozd\u00edly.&nbsp;<\/p>\n\n\n\n<p>N\u011bkte\u0159\u00ed to naz\u00fdvaj\u00ed \"demarka\u010dn\u00edm zkreslen\u00edm\", proto\u017ee pou\u017eit\u00ed pom\u011bru (d\u011blen\u00ed) nam\u00edsto rozd\u00edlu (ode\u010d\u00edt\u00e1n\u00ed) posouv\u00e1 v\u00fdsledky anal\u00fdzy z v\u011bdy do pseudov\u011bdy.<\/p>\n\n\n\n<p>N\u011bkter\u00e9 vzorky pou\u017e\u00edvaj\u00ed zkreslen\u00fd statistick\u00fd design, kter\u00fd v\u0161ak umo\u017e\u0148uje odhadnout parametry. Nap\u0159\u00edklad americk\u00e9 N\u00e1rodn\u00ed centrum pro zdravotn\u00ed statistiku v mnoha sv\u00fdch celost\u00e1tn\u00edch pr\u016fzkumech z\u00e1m\u011brn\u011b prov\u00e1d\u00ed nadm\u011brn\u00e9 v\u00fdb\u011bry u men\u0161in, aby dos\u00e1hlo dostate\u010dn\u00e9 p\u0159esnosti odhad\u016f v r\u00e1mci t\u011bchto skupin.<\/p>\n\n\n\n<p>Tato \u0161et\u0159en\u00ed vy\u017eaduj\u00ed pou\u017eit\u00ed v\u00fdb\u011brov\u00fdch vah, aby bylo mo\u017en\u00e9 z\u00edskat spr\u00e1vn\u00e9 odhady pro v\u0161echny etnick\u00e9 skupiny. Pokud jsou spln\u011bny ur\u010dit\u00e9 podm\u00ednky (p\u0159edev\u0161\u00edm spr\u00e1vn\u00fd v\u00fdpo\u010det a pou\u017eit\u00ed vah), umo\u017e\u0148uj\u00ed tyto v\u00fdb\u011bry p\u0159esn\u00fd odhad popula\u010dn\u00edch parametr\u016f.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Osv\u011bd\u010den\u00e9 postupy pro zm\u00edrn\u011bn\u00ed zkreslen\u00ed p\u0159i v\u00fdb\u011bru vzorku<\/h2>\n\n\n\n<p>Je nezbytn\u00e9 zvolit vhodnou metodu v\u00fdb\u011bru vzorku, aby v\u00fdsledn\u00e9 \u00fadaje p\u0159esn\u011b odr\u00e1\u017eely zkoumanou populaci.<\/p>\n\n\n\n<ol>\n<li><strong>Techniky n\u00e1hodn\u00e9ho v\u00fdb\u011bru vzork\u016f<\/strong>: Pou\u017eit\u00ed technik n\u00e1hodn\u00e9ho v\u00fdb\u011bru zvy\u0161uje pravd\u011bpodobnost, \u017ee vzorek je reprezentativn\u00ed pro populaci. Tato technika pom\u00e1h\u00e1 zajistit, aby byl vzorek co nejreprezentativn\u011bj\u0161\u00ed pro danou populaci, a tud\u00ed\u017e m\u00e9n\u011b pravd\u011bpodobn\u00e9, \u017ee bude obsahovat zkreslen\u00ed.<\/li>\n\n\n\n<li><strong>V\u00fdpo\u010det velikosti vzorku<\/strong>: V\u00fdpo\u010det velikosti vzorku by m\u011bl b\u00fdt proveden tak, aby byla k dispozici dostate\u010dn\u00e1 s\u00edla pro testov\u00e1n\u00ed statisticky v\u00fdznamn\u00fdch hypot\u00e9z. \u010c\u00edm v\u011bt\u0161\u00ed je velikost vzorku, t\u00edm lep\u0161\u00ed je reprezentace populace.<\/li>\n\n\n\n<li><strong>Anal\u00fdza trend\u016f<\/strong>: Hled\u00e1n\u00ed alternativn\u00edch zdroj\u016f dat a anal\u00fdza v\u0161ech pozorovan\u00fdch trend\u016f v datech, kter\u00e9 mohou b\u00fdt nevybran\u00e9.<\/li>\n\n\n\n<li><strong>Kontrola zaujatosti<\/strong>: V\u00fdskyt zkreslen\u00ed by m\u011bl b\u00fdt monitorov\u00e1n, aby se zjistilo systematick\u00e9 vylu\u010dov\u00e1n\u00ed nebo nadm\u011brn\u00e9 zahrnut\u00ed konkr\u00e9tn\u00edch datov\u00fdch bod\u016f.<\/li>\n<\/ol>\n\n\n\n<p><strong>Pozor na vzorky<\/strong><\/p>\n\n\n\n<p>Zkreslen\u00ed v\u00fdb\u011bru vzorku je p\u0159i prov\u00e1d\u011bn\u00ed v\u00fdzkumu v\u00fdznamn\u00fdm faktorem. Bez ohledu na pou\u017eitou metodiku nebo studovan\u00fd obor mus\u00ed v\u00fdzkumn\u00ed pracovn\u00edci zajistit, aby pou\u017e\u00edvali reprezentativn\u00ed vzorky, kter\u00e9 odr\u00e1\u017eej\u00ed charakteristiky zkouman\u00e9 populace.<\/p>\n\n\n\n<p>P\u0159i vytv\u00e1\u0159en\u00ed v\u00fdzkumn\u00fdch studi\u00ed je nezbytn\u00e9 v\u011bnovat velkou pozornost procesu v\u00fdb\u011bru vzorku a tak\u00e9 metodice pou\u017eit\u00e9 ke sb\u011bru dat ze vzorku. M\u011bly by se pou\u017e\u00edvat osv\u011bd\u010den\u00e9 postupy, jako jsou techniky n\u00e1hodn\u00e9ho v\u00fdb\u011bru vzork\u016f, v\u00fdpo\u010det velikosti vzorku, anal\u00fdza trend\u016f a kontrola zkreslen\u00ed, aby se zajistilo, \u017ee v\u00fdsledky v\u00fdzkumu budou platn\u00e9 a spolehliv\u00e9, a t\u00edm se zv\u00fd\u0161\u00ed pravd\u011bpodobnost, \u017ee ovlivn\u00ed politiku a praxi.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Poutav\u00e9 v\u011bdeck\u00e9 infografiky b\u011bhem n\u011bkolika minut<\/h2>\n\n\n\n<p><a href=\"http:\/\/mindthegraph.com\/\">Mind the Graph<\/a> je v\u00fdkonn\u00fd online n\u00e1stroj pro v\u011bdce, kte\u0159\u00ed pot\u0159ebuj\u00ed vytv\u00e1\u0159et vysoce kvalitn\u00ed v\u011bdeckou grafiku a ilustrace. Platforma je u\u017eivatelsky p\u0159\u00edv\u011btiv\u00e1 a p\u0159\u00edstupn\u00e1 v\u011bdc\u016fm s r\u016fznou \u00farovn\u00ed technick\u00fdch znalost\u00ed, tak\u017ee je ide\u00e1ln\u00edm \u0159e\u0161en\u00edm pro v\u011bdce, kte\u0159\u00ed pot\u0159ebuj\u00ed vytv\u00e1\u0159et grafiku pro sv\u00e9 publikace, prezentace a dal\u0161\u00ed v\u011bdeck\u00e9 komunika\u010dn\u00ed materi\u00e1ly.<\/p>\n\n\n\n<p>A\u0165 u\u017e jste v\u00fdzkumn\u00fd pracovn\u00edk v oblasti p\u0159\u00edrodn\u00edch, fyzik\u00e1ln\u00edch nebo technick\u00fdch v\u011bd, Mind the Graph nab\u00edz\u00ed \u0161irokou \u0161k\u00e1lu zdroj\u016f, kter\u00e9 v\u00e1m pomohou srozumiteln\u011b a vizu\u00e1ln\u011b p\u0159esv\u011bd\u010div\u011b sd\u011blit v\u00fdsledky va\u0161eho v\u00fdzkumu.<\/p>\n\n\n\n<div style=\"height:21px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img decoding=\"async\" loading=\"lazy\" width=\"600\" height=\"338\" src=\"https:\/\/mindthegraph.com\/blog\/wp-content\/uploads\/2022\/10\/r3qiu0qenda-3.gif\" alt=\"\" class=\"wp-image-25130\"\/><\/figure><\/div>\n\n\n<div style=\"height:21px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"is-layout-flex wp-block-buttons\">\n<div class=\"wp-block-button aligncenter\"><a class=\"wp-block-button__link has-background wp-element-button\" href=\"https:\/\/mindthegraph.com\/app\/offer-trial\" style=\"border-radius:50px;background-color:#dc1866\" target=\"_blank\" rel=\"noreferrer noopener\">Za\u010dn\u011bte vytv\u00e1\u0159et infografiky zdarma<\/a><\/div>\n<\/div>\n\n\n\n<div style=\"height:44px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>","protected":false},"excerpt":{"rendered":"<p>P\u0159i prov\u00e1d\u011bn\u00ed v\u00fdzkumu v oborech, jako je statistika, soci\u00e1ln\u00ed v\u011bdy a epidemiologie, je z\u00e1sadn\u00edm faktorem zkreslen\u00ed v\u00fdb\u011bru vzorku. <\/p>","protected":false},"author":38,"featured_media":28013,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[959,28],"tags":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v19.9 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>A problem called Sampling bias - Mind the Graph Blog<\/title>\n<meta name=\"description\" content=\"Sampling bias is a critical consideration when conducting research within disciplines such as statistics, social science, and epidemiology.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/mindthegraph.com\/blog\/cs\/vyberove-zkresleni\/\" \/>\n<meta property=\"og:locale\" content=\"cs_CZ\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"A problem called Sampling bias\" \/>\n<meta property=\"og:description\" content=\"Sampling bias is a critical consideration when conducting research within disciplines such as statistics, social science, and epidemiology.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/mindthegraph.com\/blog\/cs\/vyberove-zkresleni\/\" \/>\n<meta property=\"og:site_name\" content=\"Mind the Graph Blog\" \/>\n<meta property=\"article:published_time\" content=\"2023-05-24T13:07:19+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2023-05-24T13:07:21+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/mindthegraph.com\/blog\/wp-content\/uploads\/2023\/05\/sampling-bias-blog.jpg\" \/>\n\t<meta property=\"og:image:width\" content=\"1123\" \/>\n\t<meta property=\"og:image:height\" content=\"612\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/jpeg\" \/>\n<meta name=\"author\" content=\"Gilberto de Abreu\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:title\" content=\"A problem called Sampling bias\" \/>\n<meta name=\"twitter:description\" content=\"Sampling bias is a critical consideration when conducting research within disciplines such as statistics, social science, and epidemiology.\" \/>\n<meta name=\"twitter:image\" content=\"https:\/\/mindthegraph.com\/blog\/wp-content\/uploads\/2023\/05\/sampling-bias-blog.jpg\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Gilberto de Abreu\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"8 minutes\" \/>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"A problem called Sampling bias - Mind the Graph Blog","description":"Sampling bias is a critical consideration when conducting research within disciplines such as statistics, social science, and epidemiology.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/mindthegraph.com\/blog\/cs\/vyberove-zkresleni\/","og_locale":"cs_CZ","og_type":"article","og_title":"A problem called Sampling bias","og_description":"Sampling bias is a critical consideration when conducting research within disciplines such as statistics, social science, and epidemiology.","og_url":"https:\/\/mindthegraph.com\/blog\/cs\/vyberove-zkresleni\/","og_site_name":"Mind the Graph Blog","article_published_time":"2023-05-24T13:07:19+00:00","article_modified_time":"2023-05-24T13:07:21+00:00","og_image":[{"width":1123,"height":612,"url":"https:\/\/mindthegraph.com\/blog\/wp-content\/uploads\/2023\/05\/sampling-bias-blog.jpg","type":"image\/jpeg"}],"author":"Gilberto de Abreu","twitter_card":"summary_large_image","twitter_title":"A problem called Sampling bias","twitter_description":"Sampling bias is a critical consideration when conducting research within disciplines such as statistics, social science, and epidemiology.","twitter_image":"https:\/\/mindthegraph.com\/blog\/wp-content\/uploads\/2023\/05\/sampling-bias-blog.jpg","twitter_misc":{"Written by":"Gilberto de Abreu","Est. reading time":"8 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/mindthegraph.com\/blog\/sampling-bias\/","url":"https:\/\/mindthegraph.com\/blog\/sampling-bias\/","name":"A problem called Sampling bias - Mind the Graph Blog","isPartOf":{"@id":"https:\/\/mindthegraph.com\/blog\/#website"},"datePublished":"2023-05-24T13:07:19+00:00","dateModified":"2023-05-24T13:07:21+00:00","author":{"@id":"https:\/\/mindthegraph.com\/blog\/#\/schema\/person\/b28781b05825270dad9cba59503a9321"},"description":"Sampling bias is a critical consideration when conducting research within disciplines such as statistics, social science, and epidemiology.","breadcrumb":{"@id":"https:\/\/mindthegraph.com\/blog\/sampling-bias\/#breadcrumb"},"inLanguage":"cs-CZ","potentialAction":[{"@type":"ReadAction","target":["https:\/\/mindthegraph.com\/blog\/sampling-bias\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/mindthegraph.com\/blog\/sampling-bias\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/mindthegraph.com\/blog\/"},{"@type":"ListItem","position":2,"name":"A problem called Sampling bias"}]},{"@type":"WebSite","@id":"https:\/\/mindthegraph.com\/blog\/#website","url":"https:\/\/mindthegraph.com\/blog\/","name":"Mind the Graph Blog","description":"Your science can be beautiful!","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/mindthegraph.com\/blog\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"cs-CZ"},{"@type":"Person","@id":"https:\/\/mindthegraph.com\/blog\/#\/schema\/person\/b28781b05825270dad9cba59503a9321","name":"Gilberto de Abreu","image":{"@type":"ImageObject","inLanguage":"cs-CZ","@id":"https:\/\/mindthegraph.com\/blog\/#\/schema\/person\/image\/","url":"https:\/\/secure.gravatar.com\/avatar\/cc861028818e8a4fffa388f920fbdae9?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/cc861028818e8a4fffa388f920fbdae9?s=96&d=mm&r=g","caption":"Gilberto de Abreu"},"url":"https:\/\/mindthegraph.com\/blog\/cs\/author\/giba\/"}]}},"_links":{"self":[{"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/posts\/28012"}],"collection":[{"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/users\/38"}],"replies":[{"embeddable":true,"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/comments?post=28012"}],"version-history":[{"count":3,"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/posts\/28012\/revisions"}],"predecessor-version":[{"id":28023,"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/posts\/28012\/revisions\/28023"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/media\/28013"}],"wp:attachment":[{"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/media?parent=28012"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/categories?post=28012"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/tags?post=28012"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}