Isingeniso Sokwabiwa Kwesampula
Emkhakheni wezibalo, ukusatshalaliswa kwamasampula kunendima ebalulekile, kusebenza njengesisekelo sezibalo zokucabanga kanye nokuhlaziywa kwedatha. Lesi sihloko sihlose ukucacisa umqondo wokusatshalaliswa kwamasampula, ukuchaza ukubaluleka kwawo, nokuhlola izici nezinhlobo zawo.
Kuyini Ukusatshalaliswa Kwesampula?
Ukuze siqale ukuhlola ukusatshalaliswa kwamasampula, kubalulekile ukuqonda amagama athile ayisisekelo kuqala.
Isibalo sabantu sibhekisela kulo lonke isethi yezinto noma abantu lapho idatha ingaqoqwa khona. Lapho kufundwa izici ezithile noma ukuziphatha kwabantu, ngokuvamile akunakwenzeka ukuhlola ilungu ngalinye. Lapha, isampula iba yinto ebaluleke kakhulu - iyingxenye encane yabantu emele kodwa ekwaziyo ukulawuleka ngobukhulu.
Ukusatshalaliswa kwesampula (noma ukusatshalaliswa kwesampula) ukusatshalaliswa kwamathuba kwesibalo esithile (njengesilinganiso noma ukuhlukahluka) ngokusekelwe kusampula engahleliwe. Ngamagama alula, yindlela isilinganiso sezibalo esivela kumasampula (isb., izindlela zesampula) esiziphatha ngayo uma uthatha amasampula aphindaphindwayo kubantu.
Ukubaluleka Kokusatshalaliswa Kwesampula
Izibalo Ezingafani
Ukusetshenziswa okuyinhloko kokusatshalaliswa kwamasampula kusezibalweni zokuqagela, lapho sinquma khona amapharamitha esibalo sabantu ngokusekelwe ezibalweni zamasampula. Isibonelo, ngokusatshalaliswa kwamasampula, singalinganisa izindlela zesibalo sabantu, ukuhlukahluka, kanye nezilinganiso, okusenza sikwazi ukwenza iziphetho ezinolwazi mayelana nesibalo sabantu bonke.
Ithiyori Yomkhawulo Ophakathi (i-CLT)
I-Central Limit Theorem ingenye yezimiso ezibaluleke kakhulu kwizibalo. Ithi uma usayizi wesampula mkhulu ngokwanele, ukusatshalaliswa kwesampula kwesilinganiso sesampula kuzosatshalaliswa cishe ngokujwayelekile, kungakhathaliseki ukusatshalaliswa kwabantu. Le theorem isekela izindlela eziningi zezibalo futhi iqinisekisa ukusetshenziswa kokusatshalaliswa okuvamile ekuhlolweni kwe-hypothesis kanye nokulinganisa isikhathi sokuzethemba.
Ukuhlolwa kwe-hypothesis
Ekuhlolweni kwe-hypothesis, ukusatshalaliswa kwesampula kuvumela abahlaziyi ukuthi banqume amathuba okubuka izibalo zesampula ngaphansi kwe-null hypothesis. Ngokuqhathanisa izibalo zesampula nokusatshalaliswa okulindelekile ngaphansi kwe-null hypothesis, singanquma ukuthi siyayenqaba i-null hypothesis noma cha.
Izikhathi Zokuzethemba
Ukusatshalaliswa kwamasampula kusiza ekwakheni izikhawu zokuzethemba, okunikeza ububanzi lapho ipharamitha yabantu ingase ilale khona. Lolu bubanzi, oluhambisana nezinga lokuzethemba (isb., 95%), lunikeza isilinganiso sokuqiniseka esingalinganiswa.
Izici Zokwabiwa Kwesampula
Ukuqonda izici ezibalulekile zokusatshalaliswa kwamasampula kubalulekile ekuchazeni nasekusebenziseni kahle.
Isilinganiso Sokusatshalaliswa Kwesampula
Isilinganiso sokusatshalaliswa kwesampula kwesilinganiso sesampula (esichazwa ngokuthi \( \mu_{\bar{x}} \)) silingana nesilinganiso sabantu ( \( \mu \) ). Ngokwezibalo, \( \mu_{\bar{x}} = \mu \). Lesi sici sibonisa ukuthi inani elilindelekile lesilinganiso sesampula lifana nesilinganiso sabantu.
Ukuguquguquka kanye nephutha elijwayelekile
Ukwehluka kokusatshalaliswa kwesampula kubanjwa yiphutha elijwayelekile (SE), elilinganisa ukusabalala kwezibalo zesampula ezizungeze ipharamitha yabantu. Ngokwesilinganiso sesampula, iphutha elijwayelekile libalwa kanje:
\[ SE_{\bar{x}} = \frac{\sigma}{\sqrt{n}} \]
lapho \( \sigma \) kuyi-population standard deviation, kanye \( n \) kuyi-sample size. Le fomula ikhombisa ukuthi amasampula amakhulu aholela ephutheni elincane elijwayelekile, ngaleyo ndlela kwandisa ukunemba kwesilinganiso sesampula njengesilinganiseli sesilinganiso sesibalo sabantu.
Isimo Sokusatshalaliswa
Ukuma kokusatshalaliswa kwamasampula kuncike kusayizi wesampula kanye nokusatshalaliswa kwabantu. Ngokusho kwe-Central Limit Theorem, osayizi abakhulu besampula bavame ukukhiqiza ukusatshalaliswa kwamasampula okucishe kube ukusatshalaliswa okuvamile, kungakhathaliseki ukuthi ukusatshalaliswa kwabantu kwasekuqaleni kwakunjani.
Izinhlobo Zokusatshalaliswa Kwesampula
Ukusatshalaliswa Kwesampula Kwesilinganiso
Ukusatshalaliswa kwesampula kwesilinganiso mhlawumbe ukusatshalaliswa kwesampula okuvame ukubonwa. Kumelela ukusatshalaliswa kwesampula okutholakala kuzo zonke izibonelo ezingaba khona zosayizi othile ezithathwe kubantu. Lokhu kusatshalaliswa kubalulekile ekulinganiseni izindlela zabantu kanye nokwakha izikhawu zokuzethemba.
Ukusatshalaliswa Kwesampula Kwesilinganiso
Uma bebhekene nedatha yezigaba, ososayensi bezibalo bavame ukusebenzisa isilinganiso sesampula. Ukusatshalaliswa kwesampula kwesilinganiso ukusatshalaliswa kwezilinganiso zesampula ezitholwe kumasampula ahlukene. Lokhu kusatshalaliswa kubalulekile ekulinganiseni izilinganiso zabantu kanye nokuhlolwa kwe-hypothesis yezilinganiso.
Ukusatshalaliswa kwe-T
Uma ukuphambuka okujwayelekile kwabantu (\( \sigma \)) kungaziwa futhi usayizi wesampula umncane, izazi zezibalo zisebenzisa ukusatshalaliswa kwe-t. Ukusatshalaliswa kwe-t kufana nokusatshalaliswa okuvamile kodwa kunemisila ejiyile, okubangela ukwehluka okwandisiwe ngenxa yosayizi omncane wesampula. Njengoba usayizi wesampula ukhula, ukusatshalaliswa kwe-t kuhlangana nokusatshalaliswa okuvamile.
Ukusatshalaliswa kwe-Chi-square
Ukusatshalaliswa kwe-chi-square kusetshenziswa ekuhlolweni kokuzimela kanye nokuhlolwa kokufaneleka. Kuyisisekelo futhi ekwakheni izikhawu zokuzethemba zokwehluka kanye nokuphambuka okujwayelekile.
Ukusatshalaliswa kwe-F
Ukusatshalaliswa kwe-F kusetshenziswa ekuqhathaniseni ukuhlukahluka nokuhlaziya ukuhlukahluka (i-ANOVA). Kususelwa esilinganisweni sokusatshalaliswa kwe-chi-square okubili futhi kusiza ekunqumeni ukuthi ukuhlukahluka kwabantu ababili kuhluke kakhulu yini.
Isibonelo: Ukuqonda Ukusatshalaliswa Kwesampula Ngesimo Sangempela
Ake sihlole isibonelo esisebenzayo ukuze siqinise ukuqonda kwethu. Ake sithi sifuna ukulinganisa ubude obujwayelekile bamadoda amadala edolobheni. Ukulinganisa umuntu ngamunye akunakwenzeka, ngakho-ke sikhetha isampula yamadoda ayi-100. Sibala isilinganiso sesampula (\( \bar{x} \)) sibe ngamasentimitha angu-68 kanye nokuphambuka okujwayelekile kwesampula kube ngamasentimitha angu-3.
Uma besingathatha amasampula amaningi kuleli nani labantu, izindlela zesampula bezizohluka kancane esimweni ngasinye, okudala ukusatshalaliswa kwezindlela zesampula. Ngokusho kwe-Central Limit Theorem, uma usayizi wethu wesampula mkhulu ngokwanele, lokhu kusatshalaliswa kuzoba cishe okuvamile.
Sisebenzisa idatha yesampula, singalinganisa isilinganiso sabantu bese sakha isikhawu sokuzethemba esingu-95%. Uma sicabanga ukuthi ukuphambuka kwesilinganiso sabantu akwaziwa, sisebenzisa ukusatshalaliswa kwe-t. Iphutha elijwayelekile yileli:
\[ SE_{\bar{x}} = \frac{s}{\sqrt{n}} = \frac{3}{\sqrt{100}} = 0.3 \]
Ngama-degree angu-99 enkululeko (n-1), inani elibalulekile elivela kuthebula lokusabalalisa i-t lezinga lokuzethemba elingu-95% cishe liyi-1.984. Ngakho-ke, i-margin of error (ME) ithi:
\[ ME = 1.984 \izikhathi 0.3 = 0.5952 \]
Ngenxa yalokho, isilinganiso sokuzethemba esingu-95% sabantu yilesi:
\[ (68 – 0.5952, 68 + 0.5952) = (67.4048, 68.5952) \]
Ngakho-ke, siqiniseka ngamaphesenti angama-95 ukuthi ubude obujwayelekile babo bonke abantu abadala edolobheni buphakathi kwamasentimitha angu-67.4048 namasentimitha angu-68.5952.
Isiphetho
Ukusatshalaliswa kwamasampula kwakha itshe eliyisisekelo lokucabanga kwezibalo, okusenza sikwazi ukwenza iziphetho ezinengqondo mayelana namapharamitha abantu ngokusekelwe kudatha yesampula. Ukuqonda izici zabo, izinhlobo, kanye nezinhlelo zokusebenza kubalulekile kunoma yimuphi umhlaziyi wedatha noma umcwaningi. Kusukela ekuhlolweni kwe-hypothesis kuya ekwakheni izikhawu zokuzethemba, ukusatshalaliswa kwamasampula kungamathuluzi abalulekile okuhumusha nokudweba ukuqonda okunengqondo kusuka kudatha.
