Effect Of Different Normalization Techniques On The Results Of Criteria Weightıng Methods: An Application Based On The CRITIC Method

Saadettin Aydın

Makale detayları

Effect Of Different Normalization Techniques On The Results Of Criteria Weightıng Methods: An Application Based On The CRITIC Method

Yazarlar:

Saadettin Aydın

Yayıncı:

Van Yuzuncu Yıl University

Ana Başlıklar:

Sosyolojide Niceliksel Yöntemler

Dergi:

Yüzüncü Yıl Üniversitesi Sosyal Bilimler Enstitüsü Dergisi Issue 67, pp.121-130

Belge Tipi:

Makale

Yayın Tarihi:

3/2025

Belge Dili:

Türkçe

Document ID (Org No.):

2001342611

pISSN:

1302-6879

Anahtar kelimeler:

Özet:

 Multi-Criteria Decision Making MCDM methods enable the selection of the most appropriate option among multiple alternatives by considering various criteria. In these methods, the first step involves the weighting of criteria to determine their relative importance. In practice, the presence of different measurement units among criteria complicates direct comparison and necessitates their transformation into a dimensionless form. Accordingly, normalization techniques play a crucial role in this process. However, there is no consensus in the literature regarding the most suitable normalization technique for a given criteria weighting method. This study aims to comprehensively analyze the impact of six widely used normalization techniques on the results of criteria weighting methods using the CRITIC method. The conformity of the CRITIC method results, obtained using six different normalization techniques, to a normal distribution was examined using the Shapiro-Wilk test. Pearson correlation analysis was then applied to the weight scores that followed a normal distribution to assess the relationships between them. According to the research findings, the Z-Score Standardization Method was identified as the best alternative to replace the Linear Constant Normalization technique, which is originally used in the CRITIC method. Additionally, a strong and significant correlation was found between the weight scores obtained from the Linear Ratio Normalization, Linear Normalization by Sum, and Vector Normalization techniques, indicating that these three techniques could be used interchangeably as alternatives. On the other hand, the Logarithmic Normalization technique exhibited the poorest performance, leading to the conclusion that it is not suitable for use within the CRITIC method.

 Multi-Criteria Decision Making MCDM methods enable the selection of the most appropriate option among multiple alternatives by considering various criteria. In these methods, the first step involves the weighting of criteria to determine their relative importance. In practice, the presence of different measurement units among criteria complicates direct comparison and necessitates their transformation into a dimensionless form. Accordingly, normalization techniques play a crucial role in this process. However, there is no consensus in the literature regarding the most suitable normalization technique for a given criteria weighting method. This study aims to comprehensively analyze the impact of six widely used normalization techniques on the results of criteria weighting methods using the CRITIC method. The conformity of the CRITIC method results, obtained using six different normalization techniques, to a normal distribution was examined using the Shapiro-Wilk test. Pearson correlation analysis was then applied to the weight scores that followed a normal distribution to assess the relationships between them. According to the research findings, the Z-Score Standardization Method was identified as the best alternative to replace the Linear Constant Normalization technique, which is originally used in the CRITIC method. Additionally, a strong and significant correlation was found between the weight scores obtained from the Linear Ratio Normalization, Linear Normalization by Sum, and Vector Normalization techniques, indicating that these three techniques could be used interchangeably as alternatives. On the other hand, the Logarithmic Normalization technique exhibited the poorest performance, leading to the conclusion that it is not suitable for use within the CRITIC method.

 Multi-Criteria Decision Making MCDM methods enable the selection of the most appropriate option among multiple alternatives by considering various criteria. In these methods, the first step involves the weighting of criteria to determine their relative importance. In practice, the presence of different measurement units among criteria complicates direct comparison and necessitates their transformation into a dimensionless form. Accordingly, normalization techniques play a crucial role in this process. However, there is no consensus in the literature regarding the most suitable normalization technique for a given criteria weighting method. This study aims to comprehensively analyze the impact of six widely used normalization techniques on the results of criteria weighting methods using the CRITIC method. The conformity of the CRITIC method results, obtained using six different normalization techniques, to a normal distribution was examined using the Shapiro-Wilk test. Pearson correlation analysis was then applied to the weight scores that followed a normal distribution to assess the relationships between them. According to the research findings, the Z-Score Standardization Method was identified as the best alternative to replace the Linear Constant Normalization technique, which is originally used in the CRITIC method. Additionally, a strong and significant correlation was found between the weight scores obtained from the Linear Ratio Normalization, Linear Normalization by Sum, and Vector Normalization techniques, indicating that these three techniques could be used interchangeably as alternatives. On the other hand, the Logarithmic Normalization technique exhibited the poorest performance, leading to the conclusion that it is not suitable for use within the CRITIC method.

 Çok Kriterli Karar Verme ÇKKV yöntemleri, birden fazla kriter göz önünde bulundurularak birçok alternatif arasından en uygun seçimin yapılmasını sağlar. Bu yöntemlerde ilk olarak kriterlerin önem derecesini ortaya koymak için kriter ağırlıklandırma işlemi yapılmaktadır. Uygulamada kriterlerin farklı ölçü birimlerine sahip olması, bunların doğrudan karşılaştırılmasını güçleştirmekte ve boyutsuz hale getirilmesini zorunlu kılmaktadır. Bu doğrultuda, normalizasyon teknikleri kritik bir rol oynamaktadır. Ancak, belirli bir kriter ağırlıklandırma yöntemi için hangi normalizasyon tekniğinin en uygun olduğu konusunda literatürde ortak bir görüş bulunmamaktadır. Bu çalışmada, literatürde yaygın olarak kullanılan 6 farklı normalizasyon tekniğinin kriter ağırlıklandırma yöntemlerinin sonuçları üzerindeki etkisi, CRITIC yöntemi kullanılarak kapsamlı bir şekilde analiz edilmesi amaçlanmaktadır. 6 farklı normalizasyon tekniği kullanılarak elde edilen CRITIC yöntemi sonuçlarının normal dağılıma uygunluğu Shapiro Wilk testi kullanılarak incelenmiş, normal dağılım gösterdiği belirlenen ağırlık skorlarına Pearson korelasyon testi uygulanarak bu ağırlık skorları arasındaki ilişki incelenmiştir. Araştırma bulgularına göre, CRITIC yönteminin orijinal halinde kullanılan teknik olan Doğrusal Sabit Normalizasyonu yerine kullanılabilecek en iyi alternatif tekniğin z Skor Standartlaştırma Yöntemi olduğu belirlenmiştir. Ayrıca Doğrusal Orantı Normalizasyonu, Toplama Göre Doğrusal Normalizasyon ve Vektör Normalizasyonu tekniklerinin ağırlık skorları arasında yüksek düzeyde anlamlı korelasyon ilişkileri bulunduğu tespit edilmiş, bu üç normalizasyon tekniğinin birbirine alternatif olarak kullanılabileceği görülmüştür. Logaritmik normalizasyon tekniğinin en düşük performansı sergilediği ve CRITIC yöntemi kapsamında kullanılmasının uygun olmadığı sonucuna ulaşılmıştır.

Makale detayları

Effect Of Different Normalization Techniques On The Results Of Criteria Weightıng Methods: An Application Based On The CRITIC Method

Yazarlar:

Saadettin Aydın

Yayıncı:

Van Yuzuncu Yıl University

Ana Başlıklar:

Sosyolojide Niceliksel Yöntemler

Dergi:

Yüzüncü Yıl Üniversitesi Sosyal Bilimler Enstitüsü Dergisi Issue 67, pp.121-130

Belge Tipi:

Makale

Yayın Tarihi:

3/2025

Belge Dili:

Türkçe

Document ID (Org No.):

2001342611

pISSN:

1302-6879

Normalization

Criterion Weighting

CRITIC.: Normalizasyon

Kriter Ağırlıklandırma

CRITIC

Sayı

Makale detayları

Effect Of Different Normalization Techniques On The Results Of Criteria Weightıng Methods: An Application Based On The CRITIC Method

Yazarlar: Saadettin Aydın

Yayıncı: Van Yuzuncu Yıl University

Ana Başlıklar: Sosyolojide Niceliksel Yöntemler

Dergi: Yüzüncü Yıl Üniversitesi Sosyal Bilimler Enstitüsü Dergisi Issue 67, pp.121-130

Belge Tipi: Makale

Yayın Tarihi: 3/2025

Belge Dili: Türkçe

Document ID (Org No.): 2001342611

pISSN: 1302-6879

Normalization

Criterion Weighting

CRITIC.: Normalizasyon

Kriter Ağırlıklandırma

CRITIC

Sayı

Yazarlar:

Saadettin Aydın

Yayıncı:

Van Yuzuncu Yıl University

Ana Başlıklar:

Sosyolojide Niceliksel Yöntemler

Dergi:

Yüzüncü Yıl Üniversitesi Sosyal Bilimler Enstitüsü Dergisi Issue 67, pp.121-130

Belge Tipi:

Makale

Yayın Tarihi:

3/2025

Belge Dili:

Türkçe

Document ID (Org No.):

2001342611

pISSN:

1302-6879