{"id":4537,"date":"2023-07-10T07:36:27","date_gmt":"2023-07-10T07:36:27","guid":{"rendered":"https:\/\/statorials.org\/tr\/anova-sonuclarini-sas-cinsinden-yorumlayin\/"},"modified":"2023-07-10T07:36:27","modified_gmt":"2023-07-10T07:36:27","slug":"anova-sonuclarini-sas-cinsinden-yorumlayin","status":"publish","type":"post","link":"https:\/\/statorials.org\/tr\/anova-sonuclarini-sas-cinsinden-yorumlayin\/","title":{"rendered":"Tam k\u0131lavuz: sas'ta anova sonu\u00e7lar\u0131 nas\u0131l yorumlan\u0131r?"},"content":{"rendered":"

<\/p>\n

\n

\u00dc\u00e7 veya daha fazla ba\u011f\u0131ms\u0131z grubun ortalamalar\u0131 aras\u0131nda istatistiksel olarak anlaml\u0131 bir fark olup olmad\u0131\u011f\u0131n\u0131 belirlemek i\u00e7in tek y\u00f6nl\u00fc ANOVA<\/a> kullan\u0131l\u0131r.<\/span><\/p>\n

A\u015fa\u011f\u0131daki \u00f6rnek, SAS’ta tek y\u00f6nl\u00fc ANOVA sonu\u00e7lar\u0131n\u0131n nas\u0131l yorumlanaca\u011f\u0131n\u0131 g\u00f6sterir.<\/span><\/p>\n

\u00d6rnek: ANOVA sonu\u00e7lar\u0131n\u0131 SAS’ta yorumlama<\/strong><\/span><\/h2>\n

Bir ara\u015ft\u0131rmac\u0131n\u0131n bir \u00e7al\u0131\u015fmaya kat\u0131lmak \u00fczere 30 \u00f6\u011frenciyi i\u015fe ald\u0131\u011f\u0131n\u0131 varsayal\u0131m. \u00d6\u011frenciler bir s\u0131nava haz\u0131rlanmak i\u00e7in \u00fc\u00e7 \u00e7al\u0131\u015fma y\u00f6nteminden birini kullanmak \u00fczere rastgele atan\u0131r<\/a> .<\/span><\/p>\n

Her \u00f6\u011frencinin s\u0131nav sonu\u00e7lar\u0131 a\u015fa\u011f\u0131da g\u00f6sterilmektedir:<\/span> <\/p>\n

\"\"<\/p>\n

Bu veri k\u00fcmesini SAS’ta olu\u015fturmak i\u00e7in a\u015fa\u011f\u0131daki kodu kullanabiliriz:<\/span><\/p>\n

 \/*create dataset*\/\n<\/span>data<\/span> my_data;\n    input<\/span> Method $Score;\n    datalines<\/span> ;\nAt 78\nAt 81\nAt 82\nAt 82\nAt 85\nAt 88\nAt 88\nAt 90\nB 81\nB 83\nB 83\nB85\nB 86\nB 88\nB90\nB91\nC 84\nC 88\nC 88\nC 89\nC 90\nC 93\nC 95\nC 98\n;\nrun<\/span> ;\n<\/strong><\/span><\/pre>\n
 Daha sonra tek y\u00f6nl\u00fc ANOVA’y\u0131 ger\u00e7ekle\u015ftirmek i\u00e7in proc ANOVA’y\u0131<\/strong> kullanaca\u011f\u0131z:<\/span><\/p>\n
 \/*perform one-way ANOVA*\/\n<\/span>proc ANOVA<\/span> data<\/span> =my_data;\n    classMethod<\/span> ;\n    modelScore<\/span> = Method;\n    means<\/span> Method \/ tukey cldiff<\/span> ;\nrun<\/span> ;<\/span><\/span><\/strong><\/span><\/pre>\n
 Not<\/strong> : Tek y\u00f6nl\u00fc ANOVA’dan elde edilen genel p de\u011feri istatistiksel olarak anlaml\u0131ysa, Tukey post-hoc testinin (g\u00fcven aral\u0131klar\u0131yla) ger\u00e7ekle\u015ftirilmesi gerekti\u011fini belirtmek i\u00e7in, tukey<\/strong> ve cldiff<\/strong> se\u00e7enekleriyle birlikte ortalama<\/strong> ifadesini kulland\u0131k.<\/span><\/p>\n
 \u00d6ncelikle sonu\u00e7taki ANOVA tablosuna bakaca\u011f\u0131z:<\/span> <\/p>\n
\"SAS'ta<\/p>\n
 \u00c7\u0131kt\u0131daki her de\u011ferin nas\u0131l yorumlanaca\u011f\u0131 a\u015fa\u011f\u0131da a\u00e7\u0131klanm\u0131\u015ft\u0131r:<\/span><\/p>\n
 DF modeli:<\/strong> De\u011fi\u015fken y\u00f6ntemin<\/em> serbestlik derecesi. Bu #groups -1 olarak hesaplan\u0131r. Bu durumda 3 farkl\u0131 \u00e7al\u0131\u015fma y\u00f6ntemi vard\u0131, dolay\u0131s\u0131yla bu de\u011fer: 3-1 = 2<\/strong> .<\/span><\/p>\n
 DF hatas\u0131:<\/strong> art\u0131klar\u0131n serbestlik derecesi. Bu, #toplam g\u00f6zlemler \u2013 #gruplar olarak hesaplan\u0131r. Bu durumda 24 g\u00f6zlem ve 3 grup oldu\u011fundan bu de\u011fer: 24-3 = 21<\/strong> olur.<\/span><\/p>\n
 D\u00fczeltilmi\u015f toplam<\/strong> : DF modelinin ve DF hatas\u0131n\u0131n toplam\u0131. Bu de\u011fer 2 + 21 = 23’t\u00fcr<\/strong> .<\/span><\/p>\n
 Kareler Toplam\u0131 Modeli:<\/strong> De\u011fi\u015fkene ili\u015fkin karelerin toplam\u0131 y\u00f6ntemi<\/em> . Bu de\u011fer 175.583’t\u00fcr<\/strong> .<\/span><\/p>\n
 Kareler Toplam\u0131 Hatas\u0131:<\/strong> Art\u0131klar veya \u201chatalar\u201d ile ili\u015fkili karelerin toplam\u0131. Bu de\u011fer 350,25’tir<\/strong> .<\/span><\/p>\n
 D\u00fczeltilmi\u015f Kareler Toplam\u0131<\/strong> : SS modeli ve SS hatas\u0131n\u0131n toplam\u0131. Bu de\u011fer 525.833’t\u00fcr<\/strong> .<\/span><\/p>\n
 Ortalama kareler modeli:<\/strong> y\u00f6ntemle<\/em> ili\u015fkili karelerin ortalama toplam\u0131. Bu, SS modeli \/ DF modeli veya 175,583 \/ 2 = 87,79<\/strong> olarak hesaplan\u0131r.<\/span><\/p>\n
 Ortalama kare hatas\u0131:<\/strong> art\u0131klarla ili\u015fkili karelerin ortalama toplam\u0131. Bu, SS Hatas\u0131 \/ DF Hatas\u0131 olarak hesaplan\u0131r ve bu da 350,25 \/ 21 = 16,68<\/strong> olur.<\/span><\/p>\n
 F de\u011feri:<\/strong> ANOVA modelinin genel F istatisti\u011fi. Bu, model ortalama karesi\/ortalama kare hatas\u0131 veya 87,79\/16,68 = 5,26<\/strong> olarak hesaplan\u0131r.<\/span><\/p>\n
 Pr >F:<\/strong> Pay df = 2 ve payda df = 21 olan F istatisti\u011fiyle ili\u015fkili p de\u011feri. Bu durumda p de\u011feri 0,0140<\/strong> olur.<\/span><\/p>\n
 Sonu\u00e7 k\u00fcmesindeki en \u00f6nemli de\u011fer p de\u011feridir \u00e7\u00fcnk\u00fc bize \u00fc\u00e7 grup aras\u0131nda ortalama de\u011ferler aras\u0131nda anlaml\u0131 bir fark olup olmad\u0131\u011f\u0131n\u0131 s\u00f6yler.<\/span><\/p>\n
 Tek y\u00f6nl\u00fc bir ANOVA’n\u0131n a\u015fa\u011f\u0131daki bo\u015f ve alternatif hipotezleri kulland\u0131\u011f\u0131n\u0131 hat\u0131rlay\u0131n:<\/span><\/p>\n