Nheyo dzekugovera Samples
Pendauluan
Kugovera sampuli ipfungwa huru muhuwandu hwevanhu inotarisa pahunhu hwekugovera sampuli dzakatorwa kubva kuvanhu. Nheyo yekugovera sampuli inokosha mukufungidzira kwenhamba nekuti inotibvumira kufungidzira nekufanotaura parameters yevanhu zvichibva padata remuenzaniso.
Munyika chaiyo, kuunganidza data kubva kuvanhu vose kazhinji hazvigoneki kana kutoti hazvigoneki. Nokudaro, vaongorori vanotora sampuli kubva kuvanhu vakawanda uye vanoshandisa misimboti yekugovera sampuli kuti vawane mhedziso dzakakodzera nezvevanhu.
Chinyorwa chino chichakurukura misimboti yekugoverwa kwemasampuli, pamwe chete nedzimwe pfungwa huru dzine chekuita nekugoverwa kwemasampuli, dzakadai sekugoverwa kwemasampuli kweavhareji, dzidziso yepakati yemuganhu, uye kugoverwa kwemasampuli kwehuwandu.
Nheyo dzekutanga dzekugovera mienzaniso
Huwandu hwevanhu vs. Muenzaniso
Vanhu vose kana zvinhu zviri munyaya dzekutsvaga kana dzekuongorora nhamba. Kusiyana neizvi, muenzaniso iboka revanhu vanosarudzwa kuti vaongorore uye vaongorore. Nzira iyi inoshandiswa nekuti kuyera kana kuona vanhu vose kwakaoma kana kuti hazvigoneki.
Maparamita neManhamba
Paramendi inhamba inotsanangura hunhu hwehuwandu hwevanhu, senge avhareji, variance, kana proportion. Kune rumwe rutivi, nhamba inhamba inotorwa kubva mumuenzaniso uye inoshandiswa kufungidzira huwandu hwevanhu. Semuenzaniso, kana tichida kuziva avhareji yehurefu hwevanhu, tinogona kutora muenzaniso kubva kuvanhu, kuverenga avhareji yehurefu hwemuenzaniso (statistic), uye kushandisa izvi kufungidzira huwandu hwevanhu (parameter).
Kugoverwa kwemuenzaniso
Kugoverwa kwemasampuli kunoreva kugoverwa kwehuwandu hwehuwandu hwemuenzaniso. Ngatitii takatora sampuli dzinoverengeka kubva kuvanhu vakafanana toverenga avhareji yemuenzaniso kune imwe neimwe, kugoverwa kwemasampuli aya ndiko kugoverwa kwemasampuli kweavhareji.
Kugoverwa kwemasampuli kunopa ruzivo rwekuti nhamba dzemuenzaniso dzinoita sei kana pakaitwa ongororo dzakasiyana. Izvi zvakakosha kuti tinzwisise kusiyana kuripo munhamba dzemuenzaniso uye kuti tiite fungidziro dzakarurama dzehuwandu hwevanhu.
Dzidziso yeMuganho wePakati (Dzidziso yeMuganho wePakati)
Imwe yepfungwa dzakakosha dzine chekuita nekugoverwa kwesampuli iCentral Limit Theorem (CLT). Iyi dzidziso inotaura kuti, zvisinei nechimiro chekugoverwa kwevanhu, huwandu hwesampuli huchasvika padanho renguva dzose (kugoverwa kweGaussian) kana saizi yemuenzaniso yakakura zvakakwana, kazhinji n ≥ 30.
Kunzwisisa Theorem yeCentral Limit
Zviri pachena, Central Limit Theorem inoti kana tikatora sampuro yakakura zvakakwana kubva kuvanhu vane mean µ uye variance σ², saka sampling distribution yemuenzaniso iyoyo ichafungidzira normal distribution ine mean µ uye standard error (SE) ye σ/√n, apo n iri sample size.
Zvinorehwa neCentral Limit Theorem
CLT ine zvazvinoreva zvakakosha pakufungidzira kwezviverengero nekuti inotibvumira kushandisa mitemo yekugoverwa kwakajairika pakufungidzira nekuyedza fungidziro, kunyangwe kana data rekutanga risingawanzo kugoverwa. Izvi zvine simba guru mukuita kwezviverengero zvezuva nezuva nekuti zvinoita kuti matekiniki mazhinji ezviverengero akavakirwa patsika dzemazuva ese ave akajairika mukushandiswa kwawo.
Kugovera Mienzaniso yeAvhareji
Imwe yenzira huru dzekushandisa Central Limit Theorem ndeyekunzwisisa kugoverwa kweavhareji yesampuli. Patinotora sampuli isina kurongeka kubva kuvanhu toverenga avhareji yemuenzaniso, tinoda kuziva kuti avhareji yemuenzaniso uyu inosiyana sei kubva pamuenzaniso mumwe kuenda kune mumwe.
Vari pakati nepakati uye Kusiyana
Pamasaizi makuru emuenzaniso, kugoverwa kwesampuru kweavhareji kuchasvika pakugoverwa kwakajairika kune avhareji yakaenzana neavhareji yevanhu (μ) uye musiyano mudiki we σ²/n, apo σ iri kutsauka kwehuwandu hwevanhu uye n iri saizi yesampuru.
Standard Error
Chikanganiso chakajairwa (SE) ichiyero chekutsauka kwekugoverwa kwesampuro kubva paavhareji. Chinopa chiyero chekuti avhareji yesampuro inotarisirwa kutsauka kusvika papi kubva paavhareji yevanhu. SE inoverengerwa se σ/√n, zvichiratidza kuti kuwedzera saizi yesampuro kuchaderedza SE uye kuita kuti avhareji yevanhu ive yakarurama.
Kugoverwa kweZvikamu zveMienzaniso
Kugoverwa kwesampuli kwechikamu kwakafanana nekugoverwa kwesampuli kweavhareji, asi tinotarisa pachikamu kwete paavhareji. Semuenzaniso, ngatitii tinoda kufungidzira chikamu chevanhu vane hunhu hwakasiyana, senge chikamu chevanhu vanoputa muhuwandu.
Avhareji uye Kusiyana kweZviyero
Kana p iri chikamu chehuwandu hwevanhu chine hunhu hwakati, saka kugoverwa kwesampuli kwechikamu p (p-hat) kuchafungidzira kugoverwa kwakajairika neavhareji p uye variance (pq/n), apo q = 1 - p na n ndiyo saizi yemuenzaniso.
Chikanganiso Chakajairika cheChikamu
Chikanganiso chakajairwa chechikamu chinoverengerwa se √[p(1-p)/n]. Izvi zvinopa chiyero chekuti chiyero chemuenzaniso (p-hat) chiri kure zvakadii kubva pachiyero chechokwadi chehuwandu hwevanhu (p).
Mhedziso
Nheyo dzekugovera sampuli ndidzo hwaro hwezvinhu zvakawanda zve inferential statistics. Kunzwisisa pfungwa idzi kunobvumira vaongorori kuita fungidziro dzakakodzera uye kuita bvunzo dzefungidziro zvichibva pamienzaniso mishoma. NeCentral Limit Theorem, tinogona kushandisa nheyo dzekugovera kwakajairika mumamiriro akasiyana-siyana uye kuita fungidziro dzakarurama kunyangwe data rekutanga risingawanzo kugoverwa.
Nekuongorora kugoverwa kwemuenzaniso weavhareji nechiyero, tinogona kuwana kunzwisisa kwakadzama kwekusiyana kwenhamba yemuenzaniso uye kuita fungidziro dziri nani nezvehuwandu hwevanhu. Nheyo idzi, kunyange dzichiita sedzisina kujeka, dzine mashandisirwo akawanda anoshanda munzvimbo dzakasiyana-siyana dzekutsvagisa, kubva kusayenzi yemagariro evanhu kusvika kusayenzi yezvisikwa nebhizinesi. Chinangwa chikuru ndechekuita sarudzo dziri nani zvichibva padata riripo, kunyangwe kana data iroro riri chikamu chidiki chechokwadi chikuru.
