{"id":55803,"date":"2024-12-12T09:00:00","date_gmt":"2024-12-12T12:00:00","guid":{"rendered":"https:\/\/mindthegraph.com\/blog\/?p=55803"},"modified":"2024-12-09T14:05:01","modified_gmt":"2024-12-09T17:05:01","slug":"chi-square-test","status":"publish","type":"post","link":"https:\/\/mindthegraph.com\/blog\/cs\/chi-square-test\/","title":{"rendered":"Test ch\u00ed-kvadr\u00e1t: Pochopen\u00ed a pou\u017eit\u00ed tohoto statistick\u00e9ho n\u00e1stroje"},"content":{"rendered":"<p>Ch\u00ed-kvadr\u00e1t test je mocn\u00fdm n\u00e1strojem ve statistice, zejm\u00e9na pro anal\u00fdzu kategori\u00e1ln\u00edch dat v r\u016fzn\u00fdch form\u00e1ch a oborech. V n\u011bkter\u00fdch souborech dat p\u0159edstavuj\u00ed spojit\u00e1 \u010d\u00edsla, zat\u00edmco v jin\u00fdch kategori\u00e1ln\u00ed data p\u0159edstavuj\u00ed data seskupen\u00e1 podle pohlav\u00ed, preferenc\u00ed nebo \u00farovn\u011b vzd\u011bl\u00e1n\u00ed. P\u0159i anal\u00fdze kategori\u00e1ln\u00edch dat je ch\u00ed-kvadr\u00e1t test \u0161iroce pou\u017e\u00edvan\u00fdm statistick\u00fdm n\u00e1strojem pro zkoum\u00e1n\u00ed vztah\u016f a vyvozov\u00e1n\u00ed smyslupln\u00fdch poznatk\u016f. Tento \u010dl\u00e1nek se zab\u00fdv\u00e1 t\u00edm, jak ch\u00ed-kvadr\u00e1t test funguje, jeho aplikacemi a pro\u010d je pro v\u00fdzkumn\u00edky a datov\u00e9 analytiky nezbytn\u00fd.<\/p>\n\n\n\n<p>V tomto blogu se budeme zab\u00fdvat t\u00edm, jak ch\u00ed-kvadr\u00e1t test funguje, jak se prov\u00e1d\u00ed a jak jej lze interpretovat. Ch\u00ed-kvadr\u00e1t test m\u016f\u017eete pou\u017e\u00edt k lep\u0161\u00edmu pochopen\u00ed anal\u00fdzy dat, a\u0165 u\u017e jste student, v\u00fdzkumn\u00fd pracovn\u00edk nebo se zaj\u00edm\u00e1te o anal\u00fdzu dat obecn\u011b.<\/p>\n\n\n\n<h2>Pochopen\u00ed v\u00fdznamu testu ch\u00ed-kvadr\u00e1t<\/h2>\n\n\n\n<p>Ch\u00ed-kvadr\u00e1t test je z\u00e1kladn\u00ed statistick\u00e1 metoda pou\u017e\u00edvan\u00e1 ke zkoum\u00e1n\u00ed vztah\u016f mezi kategori\u00e1ln\u00edmi prom\u011bnn\u00fdmi a k testov\u00e1n\u00ed hypot\u00e9z v r\u016fzn\u00fdch oblastech. Pochopen\u00ed pou\u017eit\u00ed ch\u00ed-kvadr\u00e1t testu m\u016f\u017ee v\u00fdzkumn\u00fdm pracovn\u00edk\u016fm pomoci identifikovat v\u00fdznamn\u00e9 vzorce a asociace v jejich datech. V r\u00e1mci nulov\u00e9 hypot\u00e9zy porovn\u00e1v\u00e1 pozorovan\u00e9 \u00fadaje s t\u00edm, co bychom o\u010dek\u00e1vali, kdyby mezi prom\u011bnn\u00fdmi neexistoval \u017e\u00e1dn\u00fd vztah. V oborech, jako je biologie, marketing a spole\u010densk\u00e9 v\u011bdy, je tento test u\u017eite\u010dn\u00fd zejm\u00e9na p\u0159i testov\u00e1n\u00ed hypot\u00e9z o rozd\u011blen\u00ed populace.<\/p>\n\n\n\n<p>Podstatou ch\u00ed-kvadr\u00e1t testu je m\u011b\u0159en\u00ed rozd\u00edlu mezi pozorovan\u00fdmi a o\u010dek\u00e1van\u00fdmi \u010detnostmi v kategori\u00e1ln\u00edch datech. Pomoc\u00ed n\u011bj m\u016f\u017eeme odpov\u011bd\u011bt na ot\u00e1zky jako nap\u0159: \"Li\u0161\u00ed se pozorovan\u00e9 vzorce dat od toho, co by se dalo o\u010dek\u00e1vat n\u00e1hodou?\" nebo \"Jsou dv\u011b kategori\u00e1ln\u00ed prom\u011bnn\u00e9 na sob\u011b nez\u00e1visl\u00e9?\".<\/p>\n\n\n\n<h3>Typy ch\u00ed-kvadr\u00e1t test\u016f<\/h3>\n\n\n\n<p>Ch\u00ed-kvadr\u00e1t test existuje ve dvou z\u00e1kladn\u00edch form\u00e1ch - test shody a test nez\u00e1vislosti - ka\u017ed\u00e1 z nich je uzp\u016fsobena pro konkr\u00e9tn\u00ed statistick\u00e1 \u0161et\u0159en\u00ed.<\/p>\n\n\n\n<p><strong>1. Ch\u00ed-kvadr\u00e1t test shody<\/strong><\/p>\n\n\n\n<p>U jednotliv\u00fdch kategori\u00e1ln\u00edch prom\u011bnn\u00fdch se testuje, zda se \u0159\u00edd\u00ed ur\u010dit\u00fdm rozd\u011blen\u00edm. K ov\u011b\u0159en\u00ed, zda pozorovan\u00e1 data odpov\u00eddaj\u00ed o\u010dek\u00e1van\u00e9mu rozd\u011blen\u00ed, se \u010dasto pou\u017e\u00edv\u00e1 model nebo historick\u00e1 data.<\/p>\n\n\n\n<figure class=\"wp-block-image alignwide size-full\"><a href=\"https:\/\/mindthegraph.com\/science-figures\/?utm_source=blog&amp;utm_medium=cta-final&amp;utm_campaign=conversion\"><img decoding=\"async\" loading=\"lazy\" width=\"651\" height=\"174\" src=\"https:\/\/mindthegraph.com\/blog\/wp-content\/uploads\/2024\/06\/mind-the-graph-1.png\" alt=\"Logo Mind the Graph, platformy pro tvorbu v\u011bdeck\u00fdch ilustrac\u00ed a vizualizac\u00ed pro v\u00fdzkumn\u00e9 pracovn\u00edky a pedagogy.\" class=\"wp-image-54660\" srcset=\"https:\/\/mindthegraph.com\/blog\/wp-content\/uploads\/2024\/06\/mind-the-graph-1.png 651w, https:\/\/mindthegraph.com\/blog\/wp-content\/uploads\/2024\/06\/mind-the-graph-1-300x80.png 300w, https:\/\/mindthegraph.com\/blog\/wp-content\/uploads\/2024\/06\/mind-the-graph-1-18x5.png 18w, https:\/\/mindthegraph.com\/blog\/wp-content\/uploads\/2024\/06\/mind-the-graph-1-100x27.png 100w\" sizes=\"(max-width: 651px) 100vw, 651px\" \/><\/a><figcaption class=\"wp-element-caption\">Mind the Graph - <a href=\"https:\/\/mindthegraph.com\/science-figures\/?utm_source=blog&amp;utm_medium=cta-final&amp;utm_campaign=conversion\">Vytv\u00e1\u0159ejte poutav\u00e9 v\u011bdeck\u00e9 ilustrace.<\/a><\/figcaption><\/figure>\n\n\n\n<p>P\u0159em\u00fd\u0161lejte o tom, \u017ee byste 60kr\u00e1t hodili kostkou. Proto\u017ee je kostka spravedliv\u00e1, o\u010dek\u00e1vali byste, \u017ee se ka\u017ed\u00e1 strana objev\u00ed desetkr\u00e1t, ale skute\u010dn\u00e9 v\u00fdsledky se m\u00edrn\u011b li\u0161\u00ed. Chcete-li zjistit, zda je tato odchylka v\u00fdznamn\u00e1, nebo zda je pouze v\u00fdsledkem n\u00e1hody, m\u016f\u017eete prov\u00e9st test dobr\u00e9 shody.<\/p>\n\n\n\n<p><strong>P\u0159\u00edslu\u0161n\u00e9 kroky:<\/strong><\/p>\n\n\n\n<ol>\n<li>Na z\u00e1klad\u011b teoretick\u00e9ho rozd\u011blen\u00ed ur\u010dete o\u010dek\u00e1van\u00e9 \u010detnosti.<\/li>\n\n\n\n<li>Pot\u00e9 je porovnejte s pozorovan\u00fdmi frekvencemi.<\/li>\n\n\n\n<li>Vypo\u010d\u00edtejte ch\u00ed-kvadr\u00e1t statistiku pro kvantifikaci odchylky.<\/li>\n<\/ol>\n\n\n\n<p>V\u00fdzkumn\u00edci tento test \u010dasto pou\u017e\u00edvaj\u00ed p\u0159i kontrole kvality, v genetice a dal\u0161\u00edch oborech, kde cht\u011bj\u00ed porovnat pozorovan\u00e1 data s teoretick\u00fdm rozd\u011blen\u00edm.<\/p>\n\n\n\n<p><strong>2. Ch\u00ed-kvadr\u00e1t test nez\u00e1vislosti<\/strong><\/p>\n\n\n\n<p>V tomto testu se hodnot\u00ed nez\u00e1vislost dvou kategori\u00e1ln\u00edch prom\u011bnn\u00fdch. T\u00edmto testem se zkoum\u00e1, zda se rozd\u011blen\u00ed jedn\u00e9 prom\u011bnn\u00e9 li\u0161\u00ed nap\u0159\u00ed\u010d \u00farovn\u011bmi druh\u00e9 prom\u011bnn\u00e9. Kontingen\u010dn\u00ed tabulky, kter\u00e9 zobrazuj\u00ed rozd\u011blen\u00ed \u010detnost\u00ed prom\u011bnn\u00fdch, se obvykle testuj\u00ed na nez\u00e1vislost pomoc\u00ed ch\u00ed-kvadr\u00e1t testu.<\/p>\n\n\n\n<p>P\u0159edpokl\u00e1dejte, \u017ee provedete pr\u016fzkum, ve kter\u00e9m se \u00fa\u010dastn\u00edk\u016f zept\u00e1te na jejich pohlav\u00ed a preferovan\u00fd typ filmu (ak\u010dn\u00ed, drama, komedie). Ch\u00ed-kvadr\u00e1t test nez\u00e1vislosti lze pou\u017e\u00edt ke zji\u0161t\u011bn\u00ed, zda pohlav\u00ed ovliv\u0148uje filmov\u00e9 preference, nebo zda jsou nez\u00e1visl\u00e9.<\/p>\n\n\n\n<p><strong>P\u0159\u00edslu\u0161n\u00e9 kroky:<\/strong><\/p>\n\n\n\n<ol>\n<li>Vytvo\u0159te kontingen\u010dn\u00ed tabulku pro tyto dv\u011b prom\u011bnn\u00e9.<\/li>\n\n\n\n<li>Na z\u00e1klad\u011b p\u0159edpokladu, \u017ee prom\u011bnn\u00e9 jsou nez\u00e1visl\u00e9, vypo\u010d\u00edtejte o\u010dek\u00e1van\u00e9 \u010detnosti.<\/li>\n\n\n\n<li>Pomoc\u00ed ch\u00ed-kvadr\u00e1t statistiky porovnejte zji\u0161t\u011bn\u00e9 \u010detnosti s o\u010dek\u00e1van\u00fdmi \u010detnostmi.<\/li>\n<\/ol>\n\n\n\n<p>Ve v\u00fdzkumu trhu, zdravotnictv\u00ed a vzd\u011bl\u00e1v\u00e1n\u00ed se tento test hojn\u011b pou\u017e\u00edv\u00e1 ke studiu vztahu mezi demografick\u00fdmi prom\u011bnn\u00fdmi a v\u00fdsledky, nap\u0159\u00edklad vztahu mezi \u00farovn\u00ed vzd\u011bl\u00e1n\u00ed a volebn\u00edmi preferencemi.<\/p>\n\n\n\n<h2>Pou\u017eit\u00ed ch\u00ed-kvadr\u00e1t testu v re\u00e1ln\u00fdch situac\u00edch<\/h2>\n\n\n\n<p>Ch\u00ed-kvadr\u00e1t test je obzvl\u00e1\u0161t\u011b u\u017eite\u010dn\u00fd p\u0159i pr\u00e1ci s kategori\u00e1ln\u00edmi daty, jako je pohlav\u00ed, preference nebo politick\u00e1 p\u0159\u00edslu\u0161nost, a slou\u017e\u00ed k testov\u00e1n\u00ed vztah\u016f a vzorc\u016f. Testy nez\u00e1vislosti a vhodnosti se pou\u017e\u00edvaj\u00ed k ur\u010den\u00ed, zda existuje v\u00fdznamn\u00fd vztah mezi dv\u011bma prom\u011bnn\u00fdmi (test nez\u00e1vislosti).<\/p>\n\n\n\n<p>V\u00fdzkumn\u00edci mohou testovat hypot\u00e9zy a ur\u010dovat z\u00e1konitosti pomoc\u00ed testu ch\u00ed-kvadr\u00e1t u kategori\u00e1ln\u00edch dat. Existuje n\u011bkolik d\u016fvod\u016f, pro\u010d je \u0161iroce pou\u017e\u00edv\u00e1n:<\/p>\n\n\n\n<ul>\n<li>Na rozd\u00edl od parametrick\u00fdch test\u016f nevy\u017eaduje p\u0159edpoklady o rozd\u011blen\u00ed dat.<\/li>\n\n\n\n<li>Lze ji pou\u017e\u00edvat v r\u016fzn\u00fdch oborech, tak\u017ee je univerz\u00e1ln\u00ed.<\/li>\n\n\n\n<li>Na z\u00e1klad\u011b zji\u0161t\u011bn\u00fdch vzorc\u016f pom\u00e1h\u00e1 p\u0159ij\u00edmat informovan\u00e1 rozhodnut\u00ed.<\/li>\n<\/ul>\n\n\n\n<h2>P\u0159edpoklady testu ch\u00ed-kvadr\u00e1t<\/h2>\n\n\n\n<p>Aby byla zaji\u0161t\u011bna platnost v\u00fdsledk\u016f ch\u00ed-kvadr\u00e1t testu, mus\u00ed b\u00fdt spln\u011bny ur\u010dit\u00e9 p\u0159edpoklady. Tyto p\u0159edpoklady pom\u00e1haj\u00ed zachovat p\u0159esnost a relevanci testu, zejm\u00e9na p\u0159i pr\u00e1ci s kategori\u00e1ln\u00edmi daty. Je t\u0159eba se zab\u00fdvat t\u0159emi kl\u00ed\u010dov\u00fdmi p\u0159edpoklady: n\u00e1hodn\u00fdm v\u00fdb\u011brem, kategorick\u00fdmi prom\u011bnn\u00fdmi a o\u010dek\u00e1van\u00fdmi po\u010dty \u010detnost\u00ed.<\/p>\n\n\n\n<p><strong>1. N\u00e1hodn\u00fd v\u00fdb\u011br vzork\u016f<\/strong><\/p>\n\n\n\n<p>Prvn\u00edm a nejz\u00e1kladn\u011bj\u0161\u00edm p\u0159edpokladem je, \u017ee data mus\u00ed b\u00fdt shrom\u00e1\u017ed\u011bna n\u00e1hodn\u00fdm v\u00fdb\u011brem. V\u00fdsledkem je, \u017ee vzorek zahrnuje ka\u017ed\u00e9ho jednotlivce nebo prvek rovnom\u011brn\u011b. N\u00e1hodn\u00fd vzorek minimalizuje zkreslen\u00ed, tak\u017ee v\u00fdsledky lze zobecnit na v\u011bt\u0161\u00ed populaci.<\/p>\n\n\n\n<p>Pokud vzorek nen\u00ed n\u00e1hodn\u00fd, mohou b\u00fdt v\u00fdsledky zkreslen\u00e9, co\u017e m\u016f\u017ee v\u00e9st k nespr\u00e1vn\u00fdm z\u00e1v\u011br\u016fm. V\u00fdsledky pr\u016fzkumu distribuovan\u00e9ho v\u00fdhradn\u011b ur\u010dit\u00e9 skupin\u011b v r\u00e1mci populace nemus\u00ed odr\u00e1\u017eet n\u00e1zory cel\u00e9 organizace, \u010d\u00edm\u017e je poru\u0161en p\u0159edpoklad n\u00e1hodn\u00e9ho v\u00fdb\u011bru.<\/p>\n\n\n\n<p><strong>2. Kategori\u00e1ln\u00ed prom\u011bnn\u00e9<\/strong><\/p>\n\n\n\n<p>\u00da\u010delem testu ch\u00ed-kvadr\u00e1t je analyzovat kategori\u00e1ln\u00ed prom\u011bnn\u00e9 - data, kter\u00e1 lze rozd\u011blit do r\u016fzn\u00fdch kategori\u00ed. Nem\u011bly by se vyskytovat \u017e\u00e1dn\u00e9 \u010d\u00edseln\u00e9 prom\u011bnn\u00e9 (a\u010dkoli pro pohodl\u00ed mohou b\u00fdt \u010d\u00edseln\u011b k\u00f3dov\u00e1ny) a m\u011bly by b\u00fdt seskupeny do jasn\u011b definovan\u00fdch skupin.<\/p>\n\n\n\n<p>Mezi p\u0159\u00edklady kategori\u00e1ln\u00edch prom\u011bnn\u00fdch pat\u0159\u00ed:<\/p>\n\n\n\n<ul>\n<li>Pohlav\u00ed (mu\u017e, \u017eena, nebin\u00e1rn\u00ed)<\/li>\n\n\n\n<li>Rodinn\u00fd stav (svobodn\u00fd, \u017eenat\u00fd, rozveden\u00fd)<\/li>\n\n\n\n<li>Barva o\u010d\u00ed (modr\u00e1, hn\u011bd\u00e1, zelen\u00e1)<\/li>\n<\/ul>\n\n\n\n<p>Ch\u00ed-kvadr\u00e1t test nelze pou\u017e\u00edt p\u0159\u00edmo pro spojit\u00e9 \u00fadaje, jako je v\u00fd\u0161ka nebo hmotnost, pokud nejsou p\u0159evedeny na kategorie. Aby m\u011bl ch\u00ed-kvadr\u00e1t test smysl, mus\u00ed b\u00fdt data kategori\u00e1ln\u00ed, nap\u0159\u00edklad \"mal\u00fd\", \"pr\u016fm\u011brn\u00fd\" nebo \"vysok\u00fd\".<\/p>\n\n\n\n<p><strong>3. O\u010dek\u00e1van\u00e1 \u010detnost<\/strong><\/p>\n\n\n\n<p>Dal\u0161\u00edm kritick\u00fdm p\u0159edpokladem ch\u00ed-kvadr\u00e1t testu je o\u010dek\u00e1van\u00e1 \u010detnost kategori\u00ed nebo pol\u00ed\u010dek v kontingen\u010dn\u00ed tabulce. Za p\u0159edpokladu, \u017ee plat\u00ed nulov\u00e1 hypot\u00e9za (tj. \u017ee prom\u011bnn\u00e9 spolu nesouvisej\u00ed), je o\u010dek\u00e1van\u00e1 \u010detnost teoretick\u00fdm po\u010dtem \u010detnost\u00ed, kter\u00e9 existuj\u00ed v ka\u017ed\u00e9 kategorii.&nbsp;<\/p>\n\n\n\n<p>Plat\u00ed pravidlo, \u017ee: O\u010dek\u00e1van\u00e1 \u010detnost pro ka\u017edou bu\u0148ku by m\u011bla b\u00fdt alespo\u0148 5. N\u00edzk\u00e1 o\u010dek\u00e1van\u00e1 \u010detnost m\u016f\u017ee v\u00e9st k nespolehliv\u00fdm v\u00fdsledk\u016fm, pokud je testovac\u00ed statistika zkreslen\u00e1. Fisher\u016fv exaktn\u00ed test by se m\u011bl zv\u00e1\u017eit, pokud o\u010dek\u00e1van\u00e9 \u010detnosti klesnou pod 5, zejm\u00e9na u mal\u00fdch velikost\u00ed vzorku.<\/p>\n\n\n\n<h2>Pr\u016fvodce krok za krokem k proveden\u00ed ch\u00ed-kvadr\u00e1t testu<\/h2>\n\n\n\n<ol>\n<li>Stanoven\u00ed hypot\u00e9z (nulov\u00e9 a alternativn\u00ed)<\/li>\n<\/ol>\n\n\n\n<ul>\n<li>Nulov\u00e1 hypot\u00e9za (H0): Mezi ob\u011bma porovn\u00e1van\u00fdmi v\u011bcmi neexistuje \u017e\u00e1dn\u00e1 souvislost. Ve\u0161ker\u00e9 rozd\u00edly, kter\u00e9 vid\u00edte, jsou pouze n\u00e1hodn\u00e9.<\/li>\n\n\n\n<li>Alternativn\u00ed hypot\u00e9za (H\u2081): To znamen\u00e1, \u017ee mezi ob\u011bma v\u011bcmi existuje skute\u010dn\u00e1 souvislost. Rozd\u00edly nejsou n\u00e1hodn\u00e9, ale smyslupln\u00e9.<\/li>\n<\/ul>\n\n\n\n<h3>2. Vytvo\u0159en\u00ed kontingen\u010dn\u00ed tabulky<\/h3>\n\n\n\n<p>Kontingen\u010dn\u00ed tabulky ukazuj\u00ed, jak \u010dasto se ur\u010dit\u00e9 v\u011bci vyskytuj\u00ed spole\u010dn\u011b. Tabulka nap\u0159\u00edklad ukazuje r\u016fzn\u00e9 skupiny (nap\u0159\u00edklad mu\u017ee a \u017eeny) a r\u016fzn\u00e9 mo\u017enosti (nap\u0159\u00edklad kter\u00fd v\u00fdrobek preferuj\u00ed). P\u0159i prohl\u00ed\u017een\u00ed tabulky zjist\u00edte, kolik lid\u00ed spad\u00e1 do jednotliv\u00fdch skupin a voleb.<\/p>\n\n\n\n<h3>3. V\u00fdpo\u010det o\u010dek\u00e1van\u00fdch \u010detnost\u00ed<\/h3>\n\n\n\n<p>Pokud by mezi porovn\u00e1van\u00fdmi v\u011bcmi neexistovala \u017e\u00e1dn\u00e1 skute\u010dn\u00e1 souvislost, o\u010dek\u00e1van\u00e9 \u010detnosti by byly takov\u00e9, jak\u00e9 byste o\u010dek\u00e1vali. K jejich v\u00fdpo\u010dtu lze pou\u017e\u00edt jednoduch\u00fd vzorec:<\/p>\n\n\n\n<p>O\u010dek\u00e1van\u00e1 \u010detnost = (celkov\u00fd po\u010det \u0159\u00e1dk\u016f \u00d7 celkov\u00fd po\u010det sloupc\u016f) \/ celkov\u00fd sou\u010det<\/p>\n\n\n\n<p>To v\u00e1m pouze \u0159\u00edk\u00e1, jak by \u010d\u00edsla m\u011bla vypadat, kdyby v\u0161e bylo n\u00e1hodn\u00e9.<\/p>\n\n\n\n<h3>4. V\u00fdpo\u010det ch\u00ed-kvadr\u00e1t statistiky<\/h3>\n\n\n\n<p>Ch\u00ed-kvadr\u00e1t test umo\u017e\u0148uje zm\u011b\u0159it, jak moc se pozorovan\u00e1 data odchyluj\u00ed od o\u010dek\u00e1van\u00fdch v\u00fdsledk\u016f, a pom\u00e1h\u00e1 ur\u010dit, zda existuj\u00ed vztahy. Vypad\u00e1 slo\u017eit\u011b, ale porovn\u00e1v\u00e1 skute\u010dn\u00e1 \u010d\u00edsla s o\u010dek\u00e1van\u00fdmi:<\/p>\n\n\n\n<p>\ud835\udf122=\u2211(Pozorovan\u00e9-O\u010dek\u00e1van\u00e9)2\/ O\u010dek\u00e1van\u00e9<\/p>\n\n\n\n<p>Tento postup provedete pro ka\u017ed\u00e9 pol\u00ed\u010dko v tabulce a pot\u00e9 je se\u010dtete a z\u00edsk\u00e1te jedno \u010d\u00edslo, co\u017e je va\u0161e statistika ch\u00ed-kvadr\u00e1t.<\/p>\n\n\n\n<h3>5. Ur\u010den\u00ed stup\u0148\u016f volnosti<\/h3>\n\n\n\n<p>K interpretaci v\u00fdsledk\u016f pot\u0159ebujete zn\u00e1t stupn\u011b volnosti. Na z\u00e1klad\u011b velikosti va\u0161\u00ed tabulky je vypo\u010dtete. Zde je vzorec:<\/p>\n\n\n\n<p>Stupn\u011b volnosti = ((po\u010det \u0159\u00e1dk\u016f -1)\u00d7(po\u010det sloupc\u016f-1))<\/p>\n\n\n\n<p>Je to jen m\u00f3dn\u00ed zp\u016fsob, jak zohlednit velikost dat.<\/p>\n\n\n\n<h3>6. Pou\u017eit\u00ed ch\u00ed-kvadr\u00e1t rozd\u011blen\u00ed k ur\u010den\u00ed p-hodnoty<\/h3>\n\n\n\n<p>Hodnotu p lze vypo\u010d\u00edtat pomoc\u00ed statistiky ch\u00ed-kvadr\u00e1t a stup\u0148\u016f volnosti. Kdy\u017e se pod\u00edv\u00e1te na p-hodnotu, m\u016f\u017eete ur\u010dit, zda byly pozorovan\u00e9 rozd\u00edly pravd\u011bpodobn\u011b zp\u016fsobeny n\u00e1hodou, nebo zda byly smyslupln\u00e9.<\/p>\n\n\n\n<p>Interpretace p-hodnoty:<\/p>\n\n\n\n<ul>\n<li>Obvykle mal\u00e1 p-hodnota znamen\u00e1, \u017ee zji\u0161t\u011bn\u00e9 rozd\u00edly nejsou n\u00e1hodn\u00e9, tak\u017ee nulovou hypot\u00e9zu zam\u00edtnete. M\u016f\u017eete vid\u011bt skute\u010dnou souvislost mezi t\u00edm, co studujete, a t\u00edm, co d\u011bl\u00e1te.<\/li>\n\n\n\n<li>Hodnota p v\u011bt\u0161\u00ed ne\u017e 0,05 znamen\u00e1, \u017ee rozd\u00edly jsou pravd\u011bpodobn\u011b n\u00e1hodn\u00e9, tak\u017ee byste m\u011bli ponechat nulovou hypot\u00e9zu. Neexistuje tedy mezi nimi \u017e\u00e1dn\u00e1 skute\u010dn\u00e1 souvislost.<\/li>\n<\/ul>\n\n\n\n<p>Pokud se dv\u011b v\u011bci stanou n\u00e1hodou nebo spolu souvisej\u00ed, m\u016f\u017eete pomoc\u00ed tohoto zjednodu\u0161en\u00e9ho postupu zjistit, zda spolu souvisej\u00ed!<\/p>\n\n\n\n<h2>Interpretace v\u00fdsledk\u016f testu ch\u00ed-kvadr\u00e1t<\/h2>\n\n\n\n<p>Ch\u00ed-kvadr\u00e1t statistika n\u00e1m \u0159\u00edk\u00e1, jak moc se skute\u010dn\u00e1 data (to, co jste pozorovali) li\u0161\u00ed od toho, co bychom o\u010dek\u00e1vali, kdyby mezi kategoriemi neexistoval \u017e\u00e1dn\u00fd vztah. V podstat\u011b m\u011b\u0159\u00ed, jak moc se na\u0161e pozorovan\u00e9 v\u00fdsledky li\u0161\u00ed od toho, co jsme p\u0159edpov\u00eddali na z\u00e1klad\u011b n\u00e1hody.<\/p>\n\n\n\n<ul>\n<li>Velk\u00e1 hodnota ch\u00ed-kvadr\u00e1tu: Rozd\u00edl mezi va\u0161\u00edm o\u010dek\u00e1v\u00e1n\u00edm a skute\u010dnost\u00ed je velk\u00fd. M\u016f\u017ee to znamenat, \u017ee se ve va\u0161ich datech d\u011bje n\u011bco zaj\u00edmav\u00e9ho.<\/li>\n\n\n\n<li>Mal\u00e1 hodnota ch\u00ed-kvadr\u00e1tu: To znamen\u00e1, \u017ee pozorovan\u00e1 data jsou velmi bl\u00edzk\u00e1 o\u010dek\u00e1van\u00e9 hodnot\u011b a nemus\u00ed se jednat o nic neobvykl\u00e9ho.<\/li>\n<\/ul>\n\n\n\n<p>To je sice pravda, ale samotn\u00e1 hodnota ch\u00ed-kvadr\u00e1tu v\u00e1m neposkytne v\u0161echny pot\u0159ebn\u00e9 informace. Pomoc\u00ed p-hodnoty m\u016f\u017eete zjistit, zda je rozd\u00edl v\u00fdznamn\u00fd, nebo zda se jedn\u00e1 pouze o n\u00e1hodu.<\/p>\n\n\n\n<h3>Co znamen\u00e1 p-hodnota<\/h3>\n\n\n\n<p>P-hodnoty v\u00e1m pomohou ur\u010dit, zda jsou rozd\u00edly mezi daty v\u00fdznamn\u00e9. Jin\u00fdmi slovy v\u00e1m \u0159ekne, jak\u00e1 je pravd\u011bpodobnost, \u017ee zji\u0161t\u011bn\u00e9 rozd\u00edly jsou v\u00fdsledkem n\u00e1hody.<\/p>\n\n\n\n<ul>\n<li>N\u00edzk\u00e1 p-hodnota (obvykle 0,05 nebo m\u00e9n\u011b): To znamen\u00e1, \u017ee rozd\u00edl pravd\u011bpodobn\u011b nen\u00ed zp\u016fsoben n\u00e1hodou. To znamen\u00e1, \u017ee pravd\u011bpodobn\u011b existuje skute\u010dn\u00fd rozd\u00edl a d\u011bje se n\u011bco zaj\u00edmav\u00e9ho. V d\u016fsledku toho byste zam\u00edtli domn\u011bnku, \u017ee \u017e\u00e1dn\u00fd vztah neexistuje (\"nulov\u00e1 hypot\u00e9za\").<\/li>\n<\/ul>\n\n\n\n<ul>\n<li>Vysok\u00e1 p-hodnota (vy\u0161\u0161\u00ed ne\u017e 0,05): To nazna\u010duje, \u017ee rozd\u00edl m\u016f\u017ee b\u00fdt snadno zp\u016fsoben n\u00e1hodou. V\u00fdsledkem je, \u017ee neexistuje \u017e\u00e1dn\u00fd siln\u00fd n\u00e1znak toho, \u017ee by se ve va\u0161ich datech vyskytovalo n\u011bco neobvykl\u00e9ho. Pokud mezi kategoriemi neexistuje \u017e\u00e1dn\u00fd vztah, nulovou hypot\u00e9zu nezam\u00edtnete.<\/li>\n<\/ul>\n\n\n\n<h3>Jak vyvodit z\u00e1v\u011bry<\/h3>\n\n\n\n<p>Jakmile z\u00edsk\u00e1te statistiku ch\u00ed-kvadr\u00e1t a p-hodnotu, m\u016f\u017eete vyvodit z\u00e1v\u011bry:<\/p>\n\n\n\n<p>Pod\u00edvejte se na p-hodnotu:<\/p>\n\n\n\n<ul>\n<li>Pokud je p-hodnota 0,05 nebo ni\u017e\u0161\u00ed, zam\u00edtnete domn\u011bnku, \u017ee mezi dv\u011bma kategoriemi neexistuje vztah. Pokud nap\u0159\u00edklad zkoum\u00e1te, zda pohlav\u00ed ovliv\u0148uje preference produktu, a p-hodnota je n\u00edzk\u00e1 (0,05 nebo ni\u017e\u0161\u00ed), m\u016f\u017eete \u0159\u00edci: \"Zd\u00e1 se, \u017ee pohlav\u00ed ovliv\u0148uje volbu lid\u00ed.\".<\/li>\n<\/ul>\n\n\n\n<ul>\n<li>Pokud je p-hodnota v\u011bt\u0161\u00ed ne\u017e 0,05, data nevykazuj\u00ed \u017e\u00e1dn\u00fd v\u00fdznamn\u00fd rozd\u00edl, tak\u017ee dojdete k z\u00e1v\u011bru, \u017ee kategorie spolu pravd\u011bpodobn\u011b nesouvis\u00ed. P\u0159i pou\u017eit\u00ed vysok\u00e9 p-hodnoty (v\u011bt\u0161\u00ed ne\u017e 0,05) m\u016f\u017eete \u0159\u00edci: \"Neexistuje \u017e\u00e1dn\u00fd siln\u00fd d\u016fkaz, \u017ee pohlav\u00ed ovliv\u0148uje preference produkt\u016f.<\/li>\n<\/ul>\n\n\n\n<h3>Nezapome\u0148te na v\u00fdznam v re\u00e1ln\u00e9m sv\u011bt\u011b<\/h3>\n\n\n\n<p>M\u011bli byste zv\u00e1\u017eit, zda m\u00e1 statisticky v\u00fdznamn\u00fd rozd\u00edl v\u00fdznam v re\u00e1ln\u00e9m \u017eivot\u011b, i kdy\u017e ukazuje statisticky v\u00fdznamn\u00fd rozd\u00edl. U velmi rozs\u00e1hl\u00e9ho souboru dat je mo\u017en\u00e9 pova\u017eovat za d\u016fle\u017eit\u00e9 i nepatrn\u00e9 rozd\u00edly, kter\u00e9 v\u0161ak v re\u00e1ln\u00e9m sv\u011bt\u011b nemus\u00ed m\u00edt v\u00fdznamn\u00fd dopad. M\u00edsto pouh\u00e9ho pohledu na \u010d\u00edsla v\u017edy zva\u017ete, co v\u00fdsledek znamen\u00e1 v praxi.<\/p>\n\n\n\n<p>Pomoc\u00ed statistiky ch\u00ed-kvadr\u00e1t v\u00e1m \u0159ekne, zda je rozd\u00edl mezi o\u010dek\u00e1van\u00fdm a z\u00edskan\u00fdm v\u00fdsledkem skute\u010dn\u00fd, nebo zda se jedn\u00e1 o n\u00e1hodu. Kdy\u017e data zkombinujete, m\u016f\u017eete zjistit, zda mezi nimi existuje smyslupln\u00fd vztah.<\/p>\n\n\n\n<h2>Vizualizace v\u00fdsledk\u016f ch\u00ed-kvadr\u00e1t testu pomoc\u00ed Mind the Graph<\/h2>\n\n\n\n<p>Test ch\u00ed-kvadr\u00e1t pom\u00e1h\u00e1 odhalit vzorce v datech, ale efektivn\u00ed prezentace t\u011bchto poznatk\u016f vy\u017eaduje poutav\u00e9 vizu\u00e1ln\u00ed zpracov\u00e1n\u00ed. <a href=\"https:\/\/mindthegraph.com\/science-figures\/?utm_source=blog&amp;utm_medium=cta-final&amp;utm_campaign=conversion\">Mind the Graph<\/a> poskytuje intuitivn\u00ed n\u00e1stroje pro vytv\u00e1\u0159en\u00ed \u00fa\u017easn\u00fdch vizualizac\u00ed v\u00fdsledk\u016f test\u016f ch\u00ed-kvadr\u00e1t, kter\u00e9 usnad\u0148uj\u00ed pochopen\u00ed slo\u017eit\u00fdch dat. A\u0165 u\u017e jde o akademick\u00e9 zpr\u00e1vy, prezentace nebo publikace, Mind the Graph v\u00e1m pom\u016f\u017ee srozumiteln\u011b a p\u016fsobiv\u011b p\u0159edat statistick\u00e9 poznatky. Prozkoumejte na\u0161i platformu je\u0161t\u011b dnes a prom\u011b\u0148te sv\u00e1 data v poutav\u00e9 vizu\u00e1ln\u00ed p\u0159\u00edb\u011bhy.<\/p>\n\n\n\n<figure class=\"wp-block-image size-full is-resized\"><img decoding=\"async\" loading=\"lazy\" src=\"https:\/\/mindthegraph.com\/blog\/wp-content\/uploads\/2023\/09\/mtg-80-plus-fields.gif\" alt=\"&quot;Animovan\u00fd GIF zobrazuj\u00edc\u00ed v\u00edce ne\u017e 80 v\u011bdeck\u00fdch obor\u016f dostupn\u00fdch na Mind the Graph, v\u010detn\u011b biologie, chemie, fyziky a medic\u00edny, co\u017e ilustruje v\u0161estrannost platformy pro v\u00fdzkumn\u00e9 pracovn\u00edky.&quot;\" class=\"wp-image-29586\" width=\"840\" height=\"555\"\/><figcaption class=\"wp-element-caption\">Animovan\u00fd GIF p\u0159edstavuj\u00edc\u00ed \u0161irokou \u0161k\u00e1lu v\u011bdeck\u00fdch obor\u016f, kter\u00e9 pokr\u00fdv\u00e1 <a href=\"https:\/\/mindthegraph.com\/science-figures\/?utm_source=blog&amp;utm_medium=cta-final&amp;utm_campaign=conversion\">Mind the Graph<\/a>.<\/figcaption><\/figure>\n\n\n\n<div class=\"is-content-justification-center is-layout-flex wp-container-1 wp-block-buttons\">\n<div class=\"wp-block-button\"><a class=\"wp-block-button__link has-background wp-element-button\" href=\"https:\/\/mindthegraph.com\/science-figures\/?utm_source=blog&amp;utm_medium=cta-final&amp;utm_campaign=conversion\" style=\"background-color:#7833ff\"><strong>Vytv\u00e1\u0159en\u00ed kr\u00e1sn\u00fdch graf\u016f pomoc\u00ed Mind the Graph<\/strong><\/a><\/div>\n<\/div>\n\n\n\n<p><\/p>","protected":false},"excerpt":{"rendered":"<p>Zjist\u011bte, jak pou\u017e\u00edvat ch\u00ed-kvadr\u00e1t test pro anal\u00fdzu kategori\u00e1ln\u00edch dat, testov\u00e1n\u00ed hypot\u00e9z a zkoum\u00e1n\u00ed vztah\u016f mezi prom\u011bnn\u00fdmi.<\/p>","protected":false},"author":27,"featured_media":55804,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[961,977],"tags":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v19.9 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Chi-square Test: Understanding and Applying This Statistical Tool - Mind the Graph Blog<\/title>\n<meta name=\"description\" content=\"Discover how to use the chi-square test for analyzing categorical data, testing hypotheses, and exploring relationships between variables.\" \/>\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\/chi-square-test\/\" \/>\n<meta property=\"og:locale\" content=\"cs_CZ\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Chi-square Test: Understanding and Applying This Statistical Tool - Mind the Graph Blog\" \/>\n<meta property=\"og:description\" content=\"Discover how to use the chi-square test for analyzing categorical data, testing hypotheses, and exploring relationships between variables.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/mindthegraph.com\/blog\/cs\/chi-square-test\/\" \/>\n<meta property=\"og:site_name\" content=\"Mind the Graph Blog\" \/>\n<meta property=\"article:published_time\" content=\"2024-12-12T12:00:00+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2024-12-09T17:05:01+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/mindthegraph.com\/blog\/wp-content\/uploads\/2024\/12\/chi-square_test.png\" \/>\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\/png\" \/>\n<meta name=\"author\" content=\"Aayushi Zaveri\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Aayushi Zaveri\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"10 minutes\" \/>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Chi-square Test: Understanding and Applying This Statistical Tool - Mind the Graph Blog","description":"Discover how to use the chi-square test for analyzing categorical data, testing hypotheses, and exploring relationships between variables.","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\/chi-square-test\/","og_locale":"cs_CZ","og_type":"article","og_title":"Chi-square Test: Understanding and Applying This Statistical Tool - Mind the Graph Blog","og_description":"Discover how to use the chi-square test for analyzing categorical data, testing hypotheses, and exploring relationships between variables.","og_url":"https:\/\/mindthegraph.com\/blog\/cs\/chi-square-test\/","og_site_name":"Mind the Graph Blog","article_published_time":"2024-12-12T12:00:00+00:00","article_modified_time":"2024-12-09T17:05:01+00:00","og_image":[{"width":1123,"height":612,"url":"https:\/\/mindthegraph.com\/blog\/wp-content\/uploads\/2024\/12\/chi-square_test.png","type":"image\/png"}],"author":"Aayushi Zaveri","twitter_card":"summary_large_image","twitter_misc":{"Written by":"Aayushi Zaveri","Est. reading time":"10 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/mindthegraph.com\/blog\/chi-square-test\/","url":"https:\/\/mindthegraph.com\/blog\/chi-square-test\/","name":"Chi-square Test: Understanding and Applying This Statistical Tool - Mind the Graph Blog","isPartOf":{"@id":"https:\/\/mindthegraph.com\/blog\/#website"},"datePublished":"2024-12-12T12:00:00+00:00","dateModified":"2024-12-09T17:05:01+00:00","author":{"@id":"https:\/\/mindthegraph.com\/blog\/#\/schema\/person\/bbd0a706ff2842e8aff298830658ddbd"},"description":"Discover how to use the chi-square test for analyzing categorical data, testing hypotheses, and exploring relationships between variables.","breadcrumb":{"@id":"https:\/\/mindthegraph.com\/blog\/chi-square-test\/#breadcrumb"},"inLanguage":"cs-CZ","potentialAction":[{"@type":"ReadAction","target":["https:\/\/mindthegraph.com\/blog\/chi-square-test\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/mindthegraph.com\/blog\/chi-square-test\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/mindthegraph.com\/blog\/"},{"@type":"ListItem","position":2,"name":"Chi-square Test: Understanding and Applying This Statistical Tool"}]},{"@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\/bbd0a706ff2842e8aff298830658ddbd","name":"Aayushi Zaveri","image":{"@type":"ImageObject","inLanguage":"cs-CZ","@id":"https:\/\/mindthegraph.com\/blog\/#\/schema\/person\/image\/","url":"https:\/\/secure.gravatar.com\/avatar\/b2a0d532f6fba932612a6cef1fc289c0?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/b2a0d532f6fba932612a6cef1fc289c0?s=96&d=mm&r=g","caption":"Aayushi Zaveri"},"description":"Aayushi Zaveri majored in biotechnology engineering. She is currently pursuing a master's degree in Bioentrepreneurship from Karolinska Institute. She is interested in health and diseases, global health, socioeconomic development, and women's health. As a science enthusiast, she is keen in learning more about the scientific world and wants to play a part in making a difference.","sameAs":["http:\/\/linkedin.com\/in\/aayushizaveri"],"url":"https:\/\/mindthegraph.com\/blog\/cs\/author\/aayuyshi\/"}]}},"_links":{"self":[{"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/posts\/55803"}],"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\/27"}],"replies":[{"embeddable":true,"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/comments?post=55803"}],"version-history":[{"count":1,"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/posts\/55803\/revisions"}],"predecessor-version":[{"id":55805,"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/posts\/55803\/revisions\/55805"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/media\/55804"}],"wp:attachment":[{"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/media?parent=55803"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/categories?post=55803"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/mindthegraph.com\/blog\/cs\/wp-json\/wp\/v2\/tags?post=55803"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}