{"id":21111,"date":"2020-01-13T16:40:18","date_gmt":"2020-01-13T13:40:18","guid":{"rendered":"https:\/\/www.enago.com\/academy\/tr\/?p=21111"},"modified":"2020-10-29T09:51:50","modified_gmt":"2020-10-29T06:51:50","slug":"should-researchers-include-a-confidence-index-of-their-results","status":"publish","type":"post","link":"https:\/\/www.enago.com\/academy\/tr\/should-researchers-include-a-confidence-index-of-their-results\/","title":{"rendered":"Ara\u015ft\u0131rmac\u0131lar Bulgular\u0131n\u0131n G\u00fcven Endeksini Dahil Etmeli mi?"},"content":{"rendered":"<p>S\u00fcregelen yeniden \u00fcretilebilirlik problemi, deneylerin raporlanma y\u00f6ntemine dikkat \u00e7ekmektedir. Endi\u015fe uyand\u0131r\u0131c\u0131 bir \u015fekilde, bir Nitelik ara\u015ft\u0131rmas\u0131 , bilim adamlar\u0131n\u0131n %70&#8217;inin ba\u015fka bir bilim adam\u0131n\u0131n deneylerini yeniden \u00fcretemedi\u011fini ve bilim adamlar\u0131n\u0131 %50&#8217;sinin de kendi deneylerini yeniden \u00fcretemedi\u011fini ortaya koydu. Sonu\u00e7lar\u0131n\u0131z\u0131 istatistiksel olarak p-de\u011feri ile desteklemek yayg\u0131n bir bilimsel uygulamad\u0131r. Yine de yeniden \u00fcretilebilirlik krizi bu uygulaman\u0131n yeterli olmad\u0131\u011f\u0131n\u0131 g\u00f6stermektedir. Verilerin yorumlanmas\u0131na a\u00e7\u0131kl\u0131k getirmenin bir yolu, sonu\u00e7lar\u0131n\u0131z\u0131n do\u011fruluk ihtimalini g\u00f6steren bir g\u00fcven endeksi kullanmak olabilir .<\/p>\n<div id=\"om-ceko1gku0mkfe69nxkbs-holder\"><\/div>\n<p><script type=\"text\/javascript\" src=\"https:\/\/a.omappapi.com\/app\/js\/api.min.js\" data-campaign=\"ceko1gku0mkfe69nxkbs\" data-user=\"41855\" async><\/script><\/p>\n<h2>G\u00fcven Endeksi nedir?<\/h2>\n<p>Bilimsel deneylerde, bir tedaviyi bir ara\u015ft\u0131rma evreni \u00f6rne\u011fi \u00fczerinde test etmektesiniz. Ara\u015ft\u0131rma evreninden rastgele bir temsilci \u00f6rnek (\u00f6rne\u011fin klinik deney kat\u0131l\u0131mc\u0131lar\u0131) se\u00e7ti\u011finizden emin olursunuz. G\u00fcven endeksi, tedavinizin t\u00fcm ara\u015ft\u0131rma evreni \u00fczerinde (deneylerinizde oldu\u011fu gibi) ayn\u0131 etkiye sahip olaca\u011f\u0131ndan ne kadar emin olabilece\u011finizi size s\u00f6yleyecektir.<br \/>\nVerileri bir p-de\u011feri ile raporlamak bilim adamlar\u0131 aras\u0131nda yayg\u0131n bir uygulamad\u0131r ve g\u00fcven endeksine benzese de ayn\u0131s\u0131 de\u011fildir. Aralar\u0131ndaki fark; p-de\u011feri bulgular\u0131n\u0131z\u0131n %95 oran\u0131nda bir do\u011frulu\u011fu oldu\u011funu ileri s\u00fcrerken, g\u00fcven endeksinin bulgular\u0131n\u0131z\u0131n do\u011frulu\u011fu konusundaki ihtimalini g\u00f6stermesidir.<br \/>\nP-de\u011feri hesaplamalar\u0131 \u015funlara ba\u011fl\u0131d\u0131r :<br \/>\n\u2022 \u00d6rnek b\u00fcy\u00fckl\u00fc\u011f\u00fc daha b\u00fcy\u00fck \u00f6rnek boyutlar\u0131 daha do\u011fru sonu\u00e7lar verecektir.<br \/>\n\u2022 Cevap s\u0131kl\u0131\u011f\u0131- belirli bir yan\u0131t ne kadar \u00e7ok al\u0131n\u0131rsa, sonu\u00e7lar o kadar do\u011fru olur.<br \/>\n\u2022 Ara\u015ft\u0131rma evreni b\u00fcy\u00fckl\u00fc\u011f\u00fc- bu durum sadece ara\u015ft\u0131rma evreniniz k\u00fc\u00e7\u00fck oldu\u011funda \u00f6nemlidir.<br \/>\nBayes olas\u0131l\u0131\u011f\u0131n\u0131 temel alabilecek bir g\u00fcven endeksi, a\u015fa\u011f\u0131daki etkenleri i\u00e7erir:<br \/>\n\u2022 Rastgele de\u011fi\u015fkenler- bilgi eksikli\u011finden kaynaklanan bilinmeyen fakt\u00f6rler.<br \/>\n\u2022 \u00d6nsel olas\u0131l\u0131k- mevcut bilginin dikkate al\u0131nmas\u0131.<br \/>\n\u2022 Hipotez do\u011frulu\u011fu- hipotezinizin do\u011fru veya yanl\u0131\u015f olmas\u0131 olas\u0131l\u0131\u011f\u0131.<br \/>\nKlinik deneyler benzeri ileri deneyler hakk\u0131nda karar verirken g\u00fcven endeksi \u00f6nem arz etmektedir. G\u00fcven endeksi ayr\u0131ca \u00e7al\u0131\u015fmalar\u0131n k\u0131s\u0131tl\u0131l\u0131klar\u0131 hakk\u0131nda fikir vermelidir.<\/p>\n<h2>P-De\u011feri \u0130le Bulgular\u0131n Do\u011frulanmas\u0131<\/h2>\n<p>Yayg\u0131n olarak, bilim adamlar\u0131 bulgular\u0131n\u0131 yay\u0131nlanabilir olarak de\u011ferlendirmek i\u00e7in p&lt;0,05 de\u011ferine itimat etmektedir. Di\u011fer ara\u015ft\u0131rmac\u0131lara bulgular\u0131n\u0131z\u0131n ge\u00e7erli oldu\u011funu g\u00f6sterir. P-de\u011feriniz 0,05&#8217;ten b\u00fcy\u00fckse, verileri b\u00fcy\u00fck ihtimalle yay\u0131nlamazs\u0131n\u0131z. Baz\u0131lar\u0131 bu durumun literat\u00fcrde bulunmayan de\u011ferli ara\u015ft\u0131rmalara sebep oldu\u011funu s\u00f6ylemektedir.<\/p>\n<p>Veriler genel olarak a\u015fa\u011f\u0131daki \u015fekilde raporlan\u0131r. \u201cModerat\u00f6r analizleri, uyku vaktinde y\u00fcksek s\u0131cakl\u0131\u011f\u0131n d\u00fc\u015f\u00fck uyku verimlili\u011fi (UV) ile ili\u015fkili oldu\u011funu g\u00f6sterdi \u201c(b = &#8211; 11,6 pp; p = 0,020).\u201d<br \/>\nP de\u011ferinin 0,05&#8217;ten az oldu\u011funu g\u00f6rebiliriz, ancak bu durum, bu ifadenin do\u011fru oldu\u011fu anlam\u0131na m\u0131 geliyor?<br \/>\nAmerikan \u0130statistik Kurumu (ASA) 2016 y\u0131l\u0131nda, p-de\u011feri kullanma k\u0131lavuzlar\u0131 yay\u0131nlad\u0131 ve p de\u011ferinin tek ba\u015f\u0131na \u201cbir model veya hipotezle ilgili iyi bir kan\u0131t sa\u011flamad\u0131\u011f\u0131n\u0131\u201d belirtti.<br \/>\nBunun nedeni, p-de\u011feri size:<br \/>\n1. \u00d6rne\u011finizin ara\u015ft\u0131rma evreninizin bir temsilcisi olup olmad\u0131\u011f\u0131n\u0131<br \/>\n2. \u00c7al\u0131\u015f\u0131lan hipotezin do\u011fru veya yanl\u0131\u015f olup olmad\u0131\u011f\u0131n\u0131<br \/>\n3. Verilerin tesad\u00fcfen \u00fcretilip \u00fcretilmedi\u011fini<br \/>\ng\u00f6stermez.<\/p>\n<h2>G\u00fcven Endeksi&#8217;nin Dahil Edilmesi<\/h2>\n<p>Yukar\u0131dakiler g\u00f6z \u00f6n\u00fcne al\u0131nd\u0131\u011f\u0131nda, ASA, p de\u011ferinin \u201cbilimsel mant\u0131\u011f\u0131n yerine ge\u00e7memesi\u201d gerekti\u011fi ve \u201ciyi \u00e7al\u0131\u015fma tasar\u0131m\u0131 ve davran\u0131\u015f\u0131\u201d gibi di\u011fer bir\u00e7ok fakt\u00f6r\u00fcn \u00f6nemli oldu\u011fu konusunda uyarmaktad\u0131r.<br \/>\nEvet, p-de\u011feri ile ilgili herhangi bir problem, deneysel y\u00f6ntemlerin \u015feffafl\u0131\u011f\u0131 ve verilerin \u00f6nceki literat\u00fcr ba\u011flam\u0131nda analiz edilmesiyle ele al\u0131nmal\u0131d\u0131r. Ne de olsa, ara\u015ft\u0131rmac\u0131lar, bir ara\u015ft\u0131rman\u0131n sonu\u00e7lar\u0131n\u0131n ge\u00e7erlili\u011fini, ilgili t\u00fcm fakt\u00f6rleri dikkate alarak de\u011ferlendirmek \u00fczere e\u011fitilir.<br \/>\nYine de Steven Goodman, ara\u015ft\u0131rman\u0131n bulgular\u0131n bir g\u00fcven endeksi ile \u00f6l\u00e7\u00fclmesi durumunda ara\u015ft\u0131rman\u0131n netle\u015ftirilebilece\u011fini d\u00fc\u015f\u00fcnmektedir. Bu, s\u00f6zde p-hacking (p&lt;0,05 de\u011ferinin elde edilmesi i\u00e7in verilerle oynanmas\u0131) riskini azaltabilir ve bunun yerine ger\u00e7ek olas\u0131l\u0131k de\u011ferlerini verebilir. Bu netli\u011fin yeniden \u00fcretilebilirlik sorununun \u00e7\u00f6z\u00fclmesinde yard\u0131mc\u0131 olabilece\u011fini d\u00fc\u015f\u00fcn\u00fcyor musunuz? A\u015fa\u011f\u0131daki yorumlar b\u00f6l\u00fcm\u00fcnde d\u00fc\u015f\u00fcncelerinizi bize bildirin.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>S\u00fcregelen yeniden \u00fcretilebilirlik problemi, deneylerin raporlanma y\u00f6ntemine dikkat \u00e7ekmektedir. Endi\u015fe uyand\u0131r\u0131c\u0131 bir \u015fekilde, bir Nitelik [&hellip;]<\/p>\n","protected":false},"author":8267,"featured_media":21115,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[1504,1503],"tags":[1583],"ppma_author":[1656],"class_list":["post-21111","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-phds-postdocs","category-career-corner","tag-scientific-transparency"],"acf":[],"single_webinar_page_date":null,"single_webinar_page_time":null,"session_agenda":null,"who_should_attend_this_session":null,"about_the_speaker_field":null,"co-webinar-sec":null,"co_webinar_sec_one":null,"speaker-name":null,"webinar-date":null,"webinar-time":null,"webinar-s-image":null,"custum_webinar_category":null,"authors":[{"term_id":1656,"user_id":8267,"is_guest":0,"slug":"developersid","display_name":"developersid","avatar_url":"https:\/\/secure.gravatar.com\/avatar\/ac1364def40b04953d83a9610d34a15281886a26dfb5260bf0ec6b4f98db3d40?s=96&d=mm&r=g","author_category":"1","first_name":"Nilesh","last_name":"Waghela","user_url":"","job_title":"","description":""}],"_links":{"self":[{"href":"https:\/\/www.enago.com\/academy\/tr\/wp-json\/wp\/v2\/posts\/21111","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.enago.com\/academy\/tr\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.enago.com\/academy\/tr\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.enago.com\/academy\/tr\/wp-json\/wp\/v2\/users\/8267"}],"replies":[{"embeddable":true,"href":"https:\/\/www.enago.com\/academy\/tr\/wp-json\/wp\/v2\/comments?post=21111"}],"version-history":[{"count":5,"href":"https:\/\/www.enago.com\/academy\/tr\/wp-json\/wp\/v2\/posts\/21111\/revisions"}],"predecessor-version":[{"id":22282,"href":"https:\/\/www.enago.com\/academy\/tr\/wp-json\/wp\/v2\/posts\/21111\/revisions\/22282"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.enago.com\/academy\/tr\/wp-json\/wp\/v2\/media\/21115"}],"wp:attachment":[{"href":"https:\/\/www.enago.com\/academy\/tr\/wp-json\/wp\/v2\/media?parent=21111"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.enago.com\/academy\/tr\/wp-json\/wp\/v2\/categories?post=21111"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.enago.com\/academy\/tr\/wp-json\/wp\/v2\/tags?post=21111"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/www.enago.com\/academy\/tr\/wp-json\/wp\/v2\/ppma_author?post=21111"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}