Grup Karşılaştırmalarında Alternatif Bir Yaklaşım Olarak ANOM Testi

Abstract

Bu çalışmanın amacı tüm grupların aynı anda eşit olmasına ve alternatif hipotez olarak en az birinin farklı olmasına ilişkin hipotez testinde ANOVA veya Kruskal Wallis yöntemine alternatif bir yaklaşım olarak Standart ANOM veya Parametrik Olmayan ANOM testinin kullanılabilir olduğunu göstermektir. Veriler YÖK resmi istatistik sayfasından alınmış olup, 6 fakülte arasında öğrenci, öğretim üyesi sayısı ve öğretim üyesi başına düşen öğrenci sayısı açısından fark olup olmadığı hipotezleri R yazılımı ”ANOM” paketi kullanılarak test edilmiştir. Verilerin normal dağılıma uygun olmamasından dolayı, Parametrik Olmayan ANOM testi kullanılmıştır. Parametrik Olmayan ANOM’a göre Eğitim Bilimleri Fakültesi (p=0,023) ve Fen Edebiyat Fakültesindeki (p=0,033) öğrenci sayısı genel ortalamaya göre yüksek; Tıp Fakültesi ve Sağlık Bilimleri Fakültesi (p<0,001) ise genel ortalamaya göre düşük bulunmuştur. Tıp Fakültesi (p<0,001) ve Fen Edebiyat Fakültesindeki (p=0,004) öğretim üyesi sayısı genel ortalamaya göre yüksek; İktisadi-İdari Bilimler Fakültesi (p<0,001) ve Sağlık Bilimleri Fakültesi (p<0,001) ise genel ortalamaya göre düşük bulunmuştur. Eğitim Bilimleri Fakültesi (p=0,001) ve İktisadi-İdari Bilimler Fakültesindeki (p<0,001) öğretim üyesi başına düşen öğrenci sayısı genel ortalamaya göre yüksek; Tıp Fakültesi (p<0,001) ise genel ortalamaya göre düşük bulunmuştur. Parametrik Olmayan ANOM sonuçları ile Kruskal Wallis sonuçları birbirinden farklı bulunmuştur.

Keywords

• ANOM
• Ortalamaların analizi
• Yükseköğrenim

ANOM Test as an Alternative Approach to Group Comparisons

Abstract

The aim of this study is ANOM or Non Parametric ANOM test can be used approach as an alternative to ANOVA or Kruskal Wallis method in that all groups are equal and at least one alternative hypothesis is different. Data taken from the official statistics page of YÖK and 6 faculties were selected and hypothesis that there is difference in terms of number of students, number of academician and number of student for each academician between the faculties is tested using R software "ANOM" package. Because Data is non normal, Non Parametric ANOM test is used. According to Non Parametric ANOM, number of students in Faculty of Education (p=0.023) and Faculty of Arts and Science (p=0.033) were higher than sample mean; Faculty of Medicine and Faculty of Healthy Science (p<0.001) were lower than sample mean. Number of academician in Faculty of Medicine (p<0.001) and Faculty of Arts and Science (p=0.004) were higher than sample mean; Faculty of Economics and Administrative Science (p<0.001) and Faculty of Healthy Science (p<0.001) were lower than sample mean. Number of student for each academician in Faculty of Education (p=0.001) and Faculty of Economics and Administrative Science (p<0.001) were higher than sample mean; Faculty of Healthy Science (p<0.001) was lower than sample mean. Kruskal Wallis and Non Parametric ANOM results are different, which was found.

Keywords

• ANOM
• Analysis of means
• University education

References

1. [1] Tabachnick, B.G. and Fidell, L.S. (2013) Using Multivariate Statistics. 6th ed. Pearson Education, Boston.
2. [2] Mendeş, M. (2015) Uygulamalı Bilimler İçin İstatistik ve Araştırma Yöntemleri. 2nd ed. Kriter Yayın Evi, İstanbul.
3. [3] Wludyka, P.S., Nelson, P.R., and Silva, P.R. (2001) Power Curves for The Analysis of Means for Variances. Journal of Quality Technology. 33 (1), 60–65.
4. [4] Halperin, M., Greenhouse, S.W., Cornfield, J., and Zalokar, J. (1955) Tables of Percentage Points for the Studentized Maximum Absolute Deviate in Normal Samples. Journal of American Statistical Ssociation. 50 185–195.
5. [5] Ott, E.R. (1967) Analysis of Means - a Graphical Procedure. Journal of Quality Technology. 24 101–109.
6. [6] Schilling, E.G. (1973) Systematic Approach To the Analysis of Means - 1. Analysis of Treatment Effect. Journal of Quality Technology. 5 (3), 93–108.
7. [7] Wludyka, P.S. (1999) Non-Parametric Analysis of Means Type Tests for Homogeneity of Variances. Journal of Applied Statistics. 26 (2), 243–256.
8. [8] Sheesley, J.H. (1980) Comparison of K Samples Involving Variables or Attribute Data Using The Analysis of Means. Journal of Quality Techology. 12 (1), 47–52.

9. [9] Sheesley, J.H. (1981) Factors for Analysis of Means When The Standard Deviation is Estimated with The Range. Journal of Quality Technology. 13 (3), 184–185.
10. [10] Ohta, H. (1981) A Procedure for Pooling Data by The Analysis of Means. Journal of Quality Technology. 13 (2), 115–119.
11. [11] Wludyka, P.S. and Nelson, P.R. (1997) Analysis-of-Means-Type Test for Variances from Normal Populations. Technometrics 39. 274–285.
12. [12] Wludyka, P.S. and Nelson, P.R. (1997) Analysis of Means Type Tests for Variances Using Jackknifing and Subsampling. American Journal of Mathematical and Management Sciences. 17 (1–2), 31–60.
13. [13] West, R. (2009) The multiple facets of cigarette addiction and what they mean for encouraging and helping smokers to stop. COPD: Journal of Chronic Obstructive Pulmonary Disease. 6 (4), 277–283.
14. [14] Nelson, P.R. (1985) Power curves for the analysis of means. Technometrics. 27 (1), 65–73.
15. [15] Nelson, P.R. (1993) Additional uses for the analysis of means and extended tables of critical values. Technometrics. 35 (1), 61–71.
16. [16] Nelson, P.R. (1983) A Comparison of Sample Sizes for The Analysis of Means and The Analysis of Variance. Journal of Quality Technology. 15 (1), 33–39.
17. [17] Ryan, T.P. (1978) Modern Experimental Design. John Wiley & Sons Inc, New Jersey.
18. [18] Mendeş, M. and Yiğit, S. (2013) Comparison of ANOVA-F and ANOM Tests with Regard to Type 1 Error Rate and Test Power. Journal of Statistical Computation and Simulation. 8 (11), 2093–2104.
19. [19] Gülsüm, Ü.K. (2010) Bölünmüş Parseller Deney Tasarımı ve Bir Uygulama, Gazi Üniversitesi, Yüksek Lisans Tezi, Ankara, 2010.
20. [20] Hasgül, Ö. (2011) Ürün ve süreçlerin geliştirilmesinde deney tasarımı: gıda sektöründe bir uygulama. Yönetim ve Ekonomi Araştırmaları Dergisi. 42–67.
21. [21] Dumlupınar, E., Parlar, A., and Üçkardeş, F. (2015) ANOM Testinin Tıp Alanında MİNİTAB Programıyla Kullanımı, 9. Uluslararası İstatistik Kongresi Bildiriler Kitabı. 267-268.
22. [22] Pallmann, P. and Hothorn, L.A. (2016) Analysis of means: a generalized approach using R. Journal of Applied Statistics. 43 (8), 1541–1560.
23. [23] Pohlert, T. (2014) The Pairwise Multiple Comparison of Mean Ranks Package (PMCMR). R Package.
24. [24] Yüksek Öğrenim Kurumu (n.d.) https://istatistik.yok.gov.tr. Yüksek Öğrenim Kurumu.
25. [25] Gamgam, H. and Altunkaynak, B. (2008) Parametrik Olmayan Yöntemler SPSS Uygulamalı. .
26. [26] Bakir, S.T. (1989) Analysis of means using ranks. Communications in Statistics - Simulation and Computation. 18 (2), 757–776. [27] Bakir, S.T. (1994) Analysis of means using ranks for the randomized complete block design. Communications in Statistics - Simulation and Computation. 23 (2), 547–568.
27. [28] Konietschke, F., Hothorn, L.A., and Brunner, E. (2012) Rank-based multiple test procedures and simultaneous confidence intervals. Electronic Journal of Statistics. 6 (January 2014), 738–759.
28. [29] Konietschke, F., Placzeck, M., Schaarschmidt, F., and Hothorn, L.A. (2015) nparcomp:An R SoftwarevPackage for Nonparametric Multiple Comparisons and Simultaneous Confidence Intervals. Journal of Statistical Software. 64 (9), 1–17.
29. [30] Gao, X. and Alvo, M. (2008) Nonparametric multiple comparison procedures for unbalanced two-way layouts. Journal of Statistical Planning and Inference. 138 (12), 3674–3686.
30. [31] Ruymgaart, F.H. (1980) A unified approach to the asymptotic distribution theory of certain midrank statistics. Springer, Berlin.