Ancak \u00e7oklu do\u011frusal regresyon ger\u00e7ekle\u015ftirmeden \u00f6nce ilk olarak be\u015f varsay\u0131m\u0131n kar\u015f\u0131land\u0131\u011f\u0131ndan emin olmal\u0131y\u0131z:<\/span><\/p>\n 1. Do\u011frusal ili\u015fki:<\/strong> Her yorday\u0131c\u0131 de\u011fi\u015fken ile yan\u0131t de\u011fi\u015fkeni aras\u0131nda do\u011frusal bir ili\u015fki vard\u0131r.<\/span><\/p>\n 2. \u00c7oklu do\u011frusall\u0131\u011f\u0131n olmamas\u0131:<\/strong> yorday\u0131c\u0131 de\u011fi\u015fkenlerin hi\u00e7biri birbiriyle y\u00fcksek d\u00fczeyde korelasyona sahip de\u011fildir.<\/span><\/p>\n 3. Ba\u011f\u0131ms\u0131zl\u0131k:<\/strong> G\u00f6zlemler ba\u011f\u0131ms\u0131zd\u0131r.<\/span><\/p>\n 4. Homoskedastisite:<\/strong> art\u0131klar do\u011frusal modelin her noktas\u0131nda sabit bir varyansa sahiptir.<\/span><\/p>\n 5. \u00c7ok de\u011fi\u015fkenli normallik:<\/strong> Model art\u0131klar\u0131 normal da\u011f\u0131l\u0131ma sahiptir.<\/span><\/p>\n Bu varsay\u0131mlardan bir veya birka\u00e7\u0131 kar\u015f\u0131lanmazsa \u00e7oklu do\u011frusal regresyon sonu\u00e7lar\u0131 g\u00fcvenilir olmayabilir.<\/span><\/p>\n Bu yaz\u0131da her bir varsay\u0131m i\u00e7in bir a\u00e7\u0131klama, varsay\u0131m\u0131n kar\u015f\u0131lan\u0131p kar\u015f\u0131lanmad\u0131\u011f\u0131n\u0131n nas\u0131l belirlenece\u011fi ve varsay\u0131m kar\u015f\u0131lanmazsa ne yap\u0131laca\u011f\u0131na dair a\u00e7\u0131klamalar sunuyoruz.<\/span><\/p>\n \u00c7oklu do\u011frusal regresyon, her yorday\u0131c\u0131 de\u011fi\u015fken ile yan\u0131t de\u011fi\u015fkeni aras\u0131nda do\u011frusal bir ili\u015fki oldu\u011funu varsayar.<\/span><\/p>\n Bu varsay\u0131m\u0131n kar\u015f\u0131lan\u0131p kar\u015f\u0131lanmad\u0131\u011f\u0131n\u0131 belirlemenin en basit yolu, her \u00f6ng\u00f6r\u00fcc\u00fc de\u011fi\u015fkenin ve yan\u0131t de\u011fi\u015fkeninin bir da\u011f\u0131l\u0131m grafi\u011fini olu\u015fturmakt\u0131r.<\/span><\/p>\n Bu, iki de\u011fi\u015fken aras\u0131nda do\u011frusal bir ili\u015fki olup olmad\u0131\u011f\u0131n\u0131 g\u00f6rsel olarak g\u00f6rmenizi sa\u011flar.<\/span><\/p>\n Da\u011f\u0131l\u0131m grafi\u011findeki noktalar yakla\u015f\u0131k olarak d\u00fcz bir \u00e7apraz \u00e7izgi boyunca uzan\u0131yorsa de\u011fi\u015fkenler aras\u0131nda muhtemelen do\u011frusal bir ili\u015fki vard\u0131r.<\/span><\/p>\n \u00d6rne\u011fin, a\u015fa\u011f\u0131daki grafikteki noktalar\u0131n d\u00fcz bir \u00e7izgi \u00fczerinde d\u00fc\u015fmesi, bu \u00f6zel yorday\u0131c\u0131 de\u011fi\u015fken (x) ile yan\u0131t de\u011fi\u015fkeni (y) aras\u0131nda do\u011frusal bir ili\u015fki oldu\u011funu g\u00f6sterir:<\/span> <\/p>\n Bir veya daha fazla yorday\u0131c\u0131 de\u011fi\u015fken ile yan\u0131t de\u011fi\u015fkeni aras\u0131nda do\u011frusal bir ili\u015fki yoksa, birka\u00e7 se\u00e7ene\u011fimiz vard\u0131r:<\/span><\/p>\n 1.<\/strong> Tahmin de\u011fi\u015fkenine, \u00f6rne\u011fin log veya karek\u00f6k alarak do\u011frusal olmayan bir d\u00f6n\u00fc\u015f\u00fcm uygulay\u0131n. Bu \u00e7o\u011fu zaman ili\u015fkiyi daha do\u011frusal bir ili\u015fkiye d\u00f6n\u00fc\u015ft\u00fcrebilir.<\/span><\/p>\n 2.<\/strong> Modele ba\u015fka bir yorday\u0131c\u0131 de\u011fi\u015fken ekleyin. \u00d6rne\u011fin, x’e kar\u015f\u0131 y grafi\u011fi parabolik bir \u015fekle sahipse, modele ek bir tahmin de\u011fi\u015fkeni olarak X2’yi<\/sup> eklemek mant\u0131kl\u0131 olabilir.<\/span><\/p>\n 3.<\/strong> Tahmin de\u011fi\u015fkenini modelden \u00e7\u0131kar\u0131n. En u\u00e7 durumda, belirli bir yorday\u0131c\u0131 de\u011fi\u015fken ile yan\u0131t de\u011fi\u015fkeni aras\u0131nda do\u011frusal bir ili\u015fki yoksa, yorday\u0131c\u0131 de\u011fi\u015fkenin modele dahil edilmesi yararl\u0131 olmayabilir.<\/span><\/p>\n \u00c7oklu do\u011frusal regresyon, yorday\u0131c\u0131 de\u011fi\u015fkenlerden hi\u00e7birinin birbiriyle y\u00fcksek d\u00fczeyde korelasyona sahip olmad\u0131\u011f\u0131n\u0131 varsayar.<\/span><\/p>\n Bir veya daha fazla yorday\u0131c\u0131 de\u011fi\u015fken y\u00fcksek d\u00fczeyde korelasyona sahip oldu\u011funda, regresyon modeli \u00e7oklu do\u011frusall\u0131ktan<\/a> muzdarip olur ve bu da modelin katsay\u0131 tahminlerini g\u00fcvenilmez hale getirir.<\/span><\/p>\n Bu varsay\u0131m\u0131n kar\u015f\u0131lan\u0131p kar\u015f\u0131lanmad\u0131\u011f\u0131n\u0131 belirlemenin en basit yolu, her bir yorday\u0131c\u0131 de\u011fi\u015fken i\u00e7in VIF de\u011ferini hesaplamakt\u0131r.<\/span><\/p>\n VIF de\u011ferleri 1’den ba\u015flar ve \u00fcst s\u0131n\u0131r\u0131 yoktur. Genellikle 5*’in \u00fczerindeki VIF de\u011ferleri potansiyel \u00e7oklu do\u011frusall\u0131\u011f\u0131 g\u00f6sterir.<\/span><\/p>\n A\u015fa\u011f\u0131daki e\u011fitimler, \u00e7e\u015fitli istatistiksel yaz\u0131l\u0131mlarda VIF’nin nas\u0131l hesaplanaca\u011f\u0131n\u0131 g\u00f6sterir:<\/span><\/p>\n *Bazen ara\u015ft\u0131rmac\u0131lar, \u00e7al\u0131\u015fma alan\u0131na ba\u011fl\u0131 olarak bunun yerine 10 VIF de\u011ferini kullan\u0131rlar.<\/span><\/p>\n Bir veya daha fazla yorday\u0131c\u0131 de\u011fi\u015fkenin VIF de\u011feri 5’ten b\u00fcy\u00fckse, bu sorunu \u00e7\u00f6zmenin en kolay yolu, y\u00fcksek VIF de\u011ferlerine sahip yorday\u0131c\u0131 de\u011fi\u015fkeni\/de\u011fi\u015fkenleri kald\u0131rmakt\u0131r.<\/span><\/p>\n Alternatif olarak, her bir tahmin de\u011fi\u015fkenini modelde tutmak istiyorsan\u0131z, y\u00fcksek korelasyonlu tahmin de\u011fi\u015fkenlerini ele almak \u00fczere tasarlanm\u0131\u015f ridge regresyonu<\/a> , kement regresyonu<\/a> veya k\u0131smi en k\u00fc\u00e7\u00fck kareler regresyonu<\/a> gibi farkl\u0131 bir istatistiksel y\u00f6ntem kullanabilirsiniz.<\/span><\/p>\n \u00c7oklu do\u011frusal regresyon, veri setindeki her g\u00f6zlemin ba\u011f\u0131ms\u0131z oldu\u011funu varsayar.<\/span><\/p>\n Bu varsay\u0131m\u0131n kar\u015f\u0131lan\u0131p kar\u015f\u0131lanmad\u0131\u011f\u0131n\u0131 belirlemenin en basit yolu, art\u0131klar\u0131n (ve dolay\u0131s\u0131yla g\u00f6zlemlerin) otokorelasyon sergileyip sergilemedi\u011fini bize s\u00f6yleyen resmi bir istatistiksel test olan Durbin-Watson testini<\/a> ger\u00e7ekle\u015ftirmektir.<\/span><\/p>\n Bu varsay\u0131m\u0131n nas\u0131l ihlal edildi\u011fine ba\u011fl\u0131 olarak birka\u00e7 se\u00e7ene\u011finiz vard\u0131r:<\/span><\/p>\n \u00c7oklu do\u011frusal regresyon, art\u0131klar\u0131n do\u011frusal modeldeki her noktada sabit varyansa sahip oldu\u011funu varsayar. Durum b\u00f6yle olmad\u0131\u011f\u0131nda, art\u0131klar de\u011fi\u015fen varyanstan<\/a> muzdariptir.<\/span><\/p>\n Bir regresyon analizinde de\u011fi\u015fen varyans mevcut oldu\u011funda, regresyon modelinin sonu\u00e7lar\u0131 g\u00fcvenilmez hale gelir.<\/span><\/p>\n Spesifik olarak, heteroskedastisite, regresyon katsay\u0131s\u0131 tahminlerinin varyans\u0131n\u0131 artt\u0131r\u0131r, ancak regresyon modeli bunu hesaba katmaz. Bu, ger\u00e7ekte \u00f6yle olmad\u0131\u011f\u0131 halde, bir regresyon modelinin, modeldeki bir terimin istatistiksel olarak anlaml\u0131 oldu\u011funu iddia etme olas\u0131l\u0131\u011f\u0131n\u0131 \u00e7ok daha art\u0131r\u0131r.<\/span><\/p>\n Bu varsay\u0131m\u0131n kar\u015f\u0131lan\u0131p kar\u015f\u0131lanmad\u0131\u011f\u0131n\u0131 belirlemenin en kolay yolu, standartla\u015ft\u0131r\u0131lm\u0131\u015f art\u0131klar\u0131n tahmin edilen de\u011ferlere kar\u015f\u0131 bir grafi\u011fini olu\u015fturmakt\u0131r.<\/span><\/p>\n Bir veri k\u00fcmesine bir regresyon modeli yerle\u015ftirdikten sonra, yan\u0131t de\u011fi\u015fkeninin tahmin edilen de\u011ferlerini x ekseninde ve modelin standartla\u015ft\u0131r\u0131lm\u0131\u015f art\u0131klar\u0131n\u0131 x ekseninde g\u00f6r\u00fcnt\u00fcleyen bir da\u011f\u0131l\u0131m grafi\u011fi olu\u015fturabilirsiniz. y.<\/span><\/p>\n Da\u011f\u0131l\u0131m grafi\u011findeki noktalar bir trend sergiliyorsa, de\u011fi\u015fen varyans mevcut demektir.<\/span><\/p>\n A\u015fa\u011f\u0131daki grafik, heteroskedastisitenin sorun olmad\u0131\u011f\u0131 bir regresyon modeli \u00f6rne\u011fini g\u00f6stermektedir:<\/span> <\/p>\n Standartla\u015ft\u0131r\u0131lm\u0131\u015f art\u0131klar\u0131n net bir model olmadan s\u0131f\u0131r etraf\u0131nda da\u011f\u0131ld\u0131\u011f\u0131n\u0131 unutmay\u0131n.<\/span><\/p>\n A\u015fa\u011f\u0131daki grafik, de\u011fi\u015fen varyans\u0131n bir sorun oldu\u011fu<\/em> bir regresyon modeli \u00f6rne\u011fini g\u00f6stermektedir:<\/span> <\/p>\n Tahmin edilen de\u011ferler artt\u0131k\u00e7a standartla\u015ft\u0131r\u0131lm\u0131\u015f art\u0131klar\u0131n nas\u0131l daha fazla yay\u0131ld\u0131\u011f\u0131na dikkat edin. Bu \u201ckoni\u201d \u015fekli de\u011fi\u015fen varyans\u0131n klasik bir i\u015faretidir:<\/span> <\/p>\n Heteroskedasticity’yi d\u00fczeltmenin \u00fc\u00e7 yayg\u0131n yolu vard\u0131r:<\/span><\/p>\n 1. Yan\u0131t de\u011fi\u015fkenini d\u00f6n\u00fc\u015ft\u00fcr\u00fcn.<\/strong> Heteroskedasticity ile ba\u015fa \u00e7\u0131kman\u0131n en yayg\u0131n yolu, yan\u0131t de\u011fi\u015fkeninin t\u00fcm de\u011ferlerinin logunu, karek\u00f6k\u00fcn\u00fc veya k\u00fcp k\u00f6k\u00fcn\u00fc alarak yan\u0131t de\u011fi\u015fkenini d\u00f6n\u00fc\u015ft\u00fcrmektir. Bu genellikle heteroskedastisitenin ortadan kalkmas\u0131yla sonu\u00e7lan\u0131r.<\/span><\/p>\n 2. Yan\u0131t de\u011fi\u015fkenini yeniden tan\u0131mlay\u0131n.<\/strong> Yan\u0131t de\u011fi\u015fkenini yeniden tan\u0131mlaman\u0131n bir yolu, ham de\u011fer yerine bir oran<\/em> kullanmakt\u0131r. \u00d6rne\u011fin, bir \u015fehirdeki \u00e7i\u00e7ek\u00e7i say\u0131s\u0131n\u0131 tahmin etmek i\u00e7in n\u00fcfus b\u00fcy\u00fckl\u00fc\u011f\u00fcn\u00fc kullanmak yerine, ki\u015fi ba\u015f\u0131na d\u00fc\u015fen \u00e7i\u00e7ek\u00e7i say\u0131s\u0131n\u0131 tahmin etmek i\u00e7in n\u00fcfus b\u00fcy\u00fckl\u00fc\u011f\u00fcn\u00fc kullanabiliriz.<\/span><\/p>\n \u00c7o\u011fu durumda bu, \u00e7i\u00e7ek\u00e7ilerin say\u0131s\u0131ndan ziyade ki\u015fi ba\u015f\u0131na d\u00fc\u015fen \u00e7i\u00e7ek\u00e7ilerin say\u0131s\u0131n\u0131 \u00f6l\u00e7t\u00fc\u011f\u00fcm\u00fcz i\u00e7in daha b\u00fcy\u00fck pop\u00fclasyonlarda do\u011fal olarak olu\u015fan de\u011fi\u015fkenli\u011fi azalt\u0131r.<\/span><\/p>\n 3. A\u011f\u0131rl\u0131kl\u0131 regresyon kullan\u0131n.<\/strong> De\u011fi\u015fen varyansl\u0131l\u0131\u011f\u0131 d\u00fczeltmenin ba\u015fka bir yolu, her veri noktas\u0131na, uydurulmu\u015f de\u011ferinin varyans\u0131na dayal\u0131 olarak bir a\u011f\u0131rl\u0131k atayan a\u011f\u0131rl\u0131kl\u0131 regresyon kullanmakt\u0131r.<\/span><\/p>\n Temel olarak bu, daha y\u00fcksek varyansa sahip veri noktalar\u0131na d\u00fc\u015f\u00fck a\u011f\u0131rl\u0131k vererek bunlar\u0131n kalan karelerini azalt\u0131r. Uygun a\u011f\u0131rl\u0131klar kullan\u0131ld\u0131\u011f\u0131nda de\u011fi\u015fen varyans sorunu ortadan kald\u0131r\u0131labilir.<\/span><\/p>\n \u0130lgili<\/strong> : R’de A\u011f\u0131rl\u0131kl\u0131 Regresyon Nas\u0131l Ger\u00e7ekle\u015ftirilir<\/a><\/span><\/p>\n \u00c7oklu do\u011frusal regresyon, model art\u0131klar\u0131n\u0131n normal \u015fekilde da\u011f\u0131ld\u0131\u011f\u0131n\u0131 varsayar.<\/span><\/p>\n Bu varsay\u0131m\u0131n kar\u015f\u0131lan\u0131p kar\u015f\u0131lanmad\u0131\u011f\u0131n\u0131 kontrol etmenin iki yayg\u0131n yolu vard\u0131r:<\/span><\/p>\n 1. QQ grafiklerini<\/a> kullanarak<\/strong> hipotezi g\u00f6rsel olarak do\u011frulay\u0131n<\/span> .<\/p>\n Kantil-kantil grafi\u011finin k\u0131saltmas\u0131 olan QQ grafi\u011fi, bir modelin art\u0131klar\u0131n\u0131n normal da\u011f\u0131l\u0131ma uyup uymad\u0131\u011f\u0131n\u0131 belirlemek i\u00e7in kullanabilece\u011fimiz bir grafik t\u00fcr\u00fcd\u00fcr. Grafikteki noktalar kabaca d\u00fcz bir \u00e7apraz \u00e7izgi olu\u015fturuyorsa normallik varsay\u0131m\u0131 kar\u015f\u0131lanm\u0131\u015ft\u0131r.<\/span><\/p>\n A\u015fa\u011f\u0131daki QQ grafi\u011fi kabaca normal da\u011f\u0131l\u0131ma uyan bir kal\u0131nt\u0131 \u00f6rne\u011fini g\u00f6stermektedir:<\/span><\/p>\n Bununla birlikte, a\u015fa\u011f\u0131daki QQ grafi\u011fi, art\u0131klar\u0131n d\u00fcz bir \u00e7apraz \u00e7izgiden a\u00e7\u0131k\u00e7a sapt\u0131\u011f\u0131 ve normal da\u011f\u0131l\u0131ma uymad\u0131klar\u0131n\u0131 g\u00f6steren bir durumun \u00f6rne\u011fini g\u00f6stermektedir:<\/span><\/p>\n 2.<\/strong> Shapiro-Wilk, Kolmogorov-Smironov, Jarque-Barre veya D’Agostino-Pearson gibi resmi bir istatistiksel test kullanarak hipotezi do\u011frulay\u0131n.<\/span><\/p>\n Bu testlerin b\u00fcy\u00fck numune boyutlar\u0131na duyarl\u0131 oldu\u011funu unutmay\u0131n; yani, numune boyutunuz \u00e7ok b\u00fcy\u00fck oldu\u011funda genellikle art\u0131klar\u0131n normal olmad\u0131\u011f\u0131 sonucuna var\u0131rlar. Bu hipotezi do\u011frulamak i\u00e7in QQ grafi\u011fi gibi grafiksel y\u00f6ntemleri kullanman\u0131n genellikle daha kolay olmas\u0131n\u0131n nedeni budur.<\/span><\/p>\n Normallik varsay\u0131m\u0131 kar\u015f\u0131lanmazsa birka\u00e7 se\u00e7ene\u011finiz vard\u0131r:<\/span><\/p>\n 1.<\/strong> \u00d6ncelikle verilerde normallik varsay\u0131m\u0131n\u0131n ihlaline yol a\u00e7acak a\u015f\u0131r\u0131 u\u00e7 de\u011ferlerin mevcut olup olmad\u0131\u011f\u0131n\u0131 kontrol edin.<\/span><\/p>\n 2.<\/strong> Daha sonra yan\u0131t de\u011fi\u015fkenine, \u00f6rne\u011fin yan\u0131t de\u011fi\u015fkeninin t\u00fcm de\u011ferlerinin karek\u00f6k\u00fcn\u00fc, logunu veya k\u00fcp k\u00f6k\u00fcn\u00fc alarak do\u011frusal olmayan bir d\u00f6n\u00fc\u015f\u00fcm uygulayabilirsiniz. Bu genellikle model art\u0131klar\u0131n\u0131n daha normal bir da\u011f\u0131l\u0131m\u0131yla sonu\u00e7lan\u0131r.<\/span><\/p>\n A\u015fa\u011f\u0131daki e\u011fitimler \u00e7oklu do\u011frusal regresyon ve varsay\u0131mlar\u0131 hakk\u0131nda ek bilgi sa\u011flar:<\/span><\/p>\n \u00c7oklu Do\u011frusal Regresyona Giri\u015f<\/a> A\u015fa\u011f\u0131daki e\u011fitimlerde, farkl\u0131 istatistiksel yaz\u0131l\u0131mlar kullan\u0131larak \u00e7oklu do\u011frusal regresyonun nas\u0131l ger\u00e7ekle\u015ftirilece\u011fine ili\u015fkin ad\u0131m ad\u0131m \u00f6rnekler verilmektedir:<\/span><\/p>\n Excel’de \u00e7oklu do\u011frusal regresyon nas\u0131l ger\u00e7ekle\u015ftirilir<\/a> \u00c7oklu do\u011frusal regresyon, birden fazla \u00f6ng\u00f6r\u00fcc\u00fc de\u011fi\u015fken ile bir yan\u0131t de\u011fi\u015fkeni aras\u0131ndaki ili\u015fkiyi anlamak i\u00e7in kullanabilece\u011fimiz istatistiksel bir y\u00f6ntemdir. Ancak \u00e7oklu do\u011frusal regresyon ger\u00e7ekle\u015ftirmeden \u00f6nce ilk olarak be\u015f varsay\u0131m\u0131n kar\u015f\u0131land\u0131\u011f\u0131ndan emin olmal\u0131y\u0131z: 1. Do\u011frusal ili\u015fki: Her yorday\u0131c\u0131 de\u011fi\u015fken ile yan\u0131t de\u011fi\u015fkeni aras\u0131nda do\u011frusal bir ili\u015fki vard\u0131r. 2. \u00c7oklu do\u011frusall\u0131\u011f\u0131n olmamas\u0131: yorday\u0131c\u0131 de\u011fi\u015fkenlerin hi\u00e7biri birbiriyle y\u00fcksek […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[11],"tags":[],"class_list":["post-2436","post","type-post","status-publish","format-standard","hentry","category-rehber"],"yoast_head":"\n Hipotez 1: Do\u011frusal ili\u015fki<\/strong><\/span><\/h2>\n
Bu varsay\u0131m\u0131n kar\u015f\u0131lan\u0131p kar\u015f\u0131lanmad\u0131\u011f\u0131 nas\u0131l belirlenir<\/strong><\/h3>\n
![\"\"](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/hypotheseslinreg1.jpg\")
<\/p>\n Bu varsay\u0131ma uyulmazsa ne yap\u0131lmal\u0131?<\/strong><\/h3>\n
Hipotez 2: \u00e7oklu ba\u011flant\u0131 yok<\/strong><\/span><\/h2>\n
Bu varsay\u0131m\u0131n kar\u015f\u0131lan\u0131p kar\u015f\u0131lanmad\u0131\u011f\u0131 nas\u0131l belirlenir<\/strong><\/h3>\n
\n
Bu varsay\u0131ma uyulmazsa ne yap\u0131lmal\u0131?<\/strong><\/h3>\n
Hipotez 3: Ba\u011f\u0131ms\u0131zl\u0131k<\/strong><\/span><\/h2>\n
Bu varsay\u0131m\u0131n kar\u015f\u0131lan\u0131p kar\u015f\u0131lanmad\u0131\u011f\u0131 nas\u0131l belirlenir<\/strong><\/h3>\n
Bu varsay\u0131ma uyulmazsa ne yap\u0131lmal\u0131?<\/strong><\/h3>\n
\n
Hipotez 4: E\u015fvaranl\u0131k<\/strong><\/span><\/h2>\n
Bu varsay\u0131m\u0131n kar\u015f\u0131lan\u0131p kar\u015f\u0131lanmad\u0131\u011f\u0131 nas\u0131l belirlenir<\/strong><\/h3>\n
![\"\"](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/norme1.png\")
<\/p>\n![\"\"](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/norme2.png\")
<\/p>\n![\"\"](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/norme3.png\")
<\/p>\n Bu varsay\u0131ma uyulmazsa ne yap\u0131lmal\u0131?<\/strong><\/h3>\n
Varsay\u0131m 4: \u00c7ok de\u011fi\u015fkenli normallik<\/strong><\/span><\/h2>\n
Bu varsay\u0131m\u0131n kar\u015f\u0131lan\u0131p kar\u015f\u0131lanmad\u0131\u011f\u0131 nas\u0131l belirlenir<\/strong><\/h3>\n
Bu varsay\u0131ma uyulmazsa ne yap\u0131lmal\u0131?<\/strong><\/h3>\n
Ek kaynaklar<\/strong><\/span><\/h3>\n
Regresyon Analizinde De\u011fi\u015fken Varyans K\u0131lavuzu<\/a>
Regresyonda \u00c7oklu Ba\u011flant\u0131 ve VIF K\u0131lavuzu<\/a><\/p>\n
R’de \u00e7oklu do\u011frusal regresyon nas\u0131l ger\u00e7ekle\u015ftirilir<\/a>
SPSS’de \u00e7oklu do\u011frusal regresyon nas\u0131l ger\u00e7ekle\u015ftirilir<\/a>
Stata’da \u00e7oklu do\u011frusal regresyon nas\u0131l ger\u00e7ekle\u015ftirilir<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"