Application of the log-linear analysis to choose determinants of disability among the elderly residents of south-eastern Poland
DOI:
https://doi.org/10.15584/medrev.2016.4.5Keywords:
disability, elderly, log-linear model, PolandAbstract
Introduction: Many variables in research in the area of medical and health sciences are qualitative in nature. A common statistical tool used to analyze them is the χ2 test. However, it does not allow us to assess the relationship between a number of variables and distinguish the factors determining the investigated phenomenon. A more accurate tool is the log-linear analysis, which enable the researcher to evaluate the dependences and interactions between the studied variables.
Purpose: Description of the use of log-linear analysis on the example of the cross-sectional study on disability of the elderly.
Material and methods: The assessment of disability and the choice of the factors that determine it was carried out on the results of a survey of 800 randomly selected people aged 71-80 years from the area of south-eastern Poland. The research tool was a WHODAS 2.0 questionnaire and a respondent’s particulars. The log-linear model was used for the analysis. In order to evaluate the fitting of the model, the Pearson’s χ2 and the χ2 maximum likelihood statistics, R2 and A coefficients were used.
Results: Education, adjustments of a house / flat, physical exercises have a significant impact on the prevalence of disability in the study group.
Conclusions: The log-linear analysis allows us to determine the effect not only of individual variables on the formation of an independent variable, but also their interactions and the determination the odds occurrence of a dependent variable according to different qualitative categories of dependent variables. The information obtained in this way is a valuable clue to take practical action to decrease or increase the severity of the studied phenomenon.
Downloads
References
Kwasiborski PJ, Sobol M. Test niezależności chi-kwadrat i jego zastosowanie w interpretacji wyników badań klinicznych. Kardiochir Torakochi 2011;4:550–554.
Stanisz A. Przystępny kurs statystyki z zastosowaniem Statistica PL na przykładach z medycyny, tom 1, Statystyki podstawowe. Kraków: StatSoft; 2006.
Stanisz A. Przystępny kurs statystyki z zastosowaniem Statistica PL na przykładach z medycyny, tom 3, Analizy wielowymiarowe. Kraków: StatSoft; 2007.
Officer A, Groce NE. Key concepts in disability. Lancet 2009;374(9704):1795–1796.
World Health Organization. World Report on Disability. Geneva, Switzerland: World Health Organization; 2011.
Salamaga M. Wykorzystanie analizy log-liniowej do wyboru czynników opisujących sytuację ekonomiczną gospodarstw domowych. Przegląd Statystyczny 2008;55(4):40-51.
Bąk I. Wybór czynników determinujących wyjazdy turystyczne rencistów. W: Borkowski B, redaktor. Metody ilościowe w badaniach ekonomicznych. Warszawa: Szkoła Główna Gospodarstwa Wiejskiego Wydział Zastosowań Informatyki i Matematyki Katedra Ekonometrii i Statystyki; 2011:1–10.
Brzezińska J. Wykorzystywanie modeli logarytmiczno-liniowych do analizy bezrobocia w Polsce w latach 2004-2012. Pr Nauk Uniw Ekon Wroc 2014;15(327):95-103.
Üstün TB, Kostanjsek N, Chatterji S, Rehm J. Measuring Health and Disability Manual for WHO Disability Assessment Schedule WHODAS 2.0. Geneva, Switzerland: World Health Organization; 2010.
World Health Organization. International Classification of Functioning, Disability and Health. Geneva, Switzerland: World Health Organization; 2011.
Dobosz M. Wspomagana komputerowo statystyczna analiza wyników badań. Warszawa: Akademicka Oficyna Wydawnicza ELIT; 2004.
Bąk I. Wykorzystanie analizy log-liniowej do wyboru czynników opisujących aktywność turystyczną domowych gospodarstw emerytów w Polsce. Folia Pomer. Univ. Technol. Stetin. Oeconomica 2012;297(68):5–16.
Goodmann LA. Analyzing Qualitative/Categorical Data. In: Magidson J, edotors. Log−Linear Models and Latent−Structure Analysis. Cambridge; 1978.
Akaike H. Information theory and an extension of the maximum likelihood principle. In: Petrov BN, Czaki F, edotors. Proceedings of the 2and International Symposium on Information. Budapest; 1973.
Kaczmarek E, Małkiewicz E. Przydatność modelowania log-liniowego do wieloczynnikowej analizy epidemiologicznej – przykład badania populacyjnego występowania rozszczepów podniebienia pierwotnego i/lub wtórnego w latach 1998–1999 w Polsce. Dent Med Probl 2005;42(3):419-424.
Peng W, Zhang H, Li Y, Li C, Yan W. Log-linear model used in the hybrid design of case-parents triad/control-mother dyad. Zhonghua Liu Xing Bing Xue Za Zhi 2016;37(6):886-890.
Zhu M, Nie S, Yan W. Application of log-linear model in the case-parent triad study. Zhonghua Liu Xing Bing Xue Za Zhi 2012;33(4):435-438.
Helmy A, Timofeev I, Palmer C, Gore A, Menon D, Hutchinson P. Hierarchical log linear analysis of admission blood parameters and clinical outcome following traumatic brain injury. Acta Neurochirurgica 2010;152(6):953-957.
Nyman H, Pensar J, Koski T, Corander J. Context-specific independence in graphical log-linear models. Computational Statistics 2016;31(4):1493-1512.
Półtorak M. Modele log-liniowe i ich zastosowania w psychologii. Przegląd Psychologiczny 2007;50(1):25-44.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2016 European Journal of Clinical and Experimental Medicine
This work is licensed under a Creative Commons Attribution 4.0 International License.
Our open access policy is in accordance with the Budapest Open Access Initiative (BOAI) definition: this means that articles have free availability on the public Internet, permitting any users to read, download, copy, distribute, print, search, or link to the full texts of these articles, crawl them for indexing, pass them as data to software, or use them for any other lawful purpose, without financial, legal, or technical barriers other than those inseparable from having access to the Internet itself.
All articles are published with free open access under the CC-BY Creative Commons attribution license (the current version is CC-BY, version 4.0). If you submit your paper for publication by the Eur J Clin Exp Med, you agree to have the CC-BY license applied to your work. Under this Open Access license, you, as the author, agree that anyone may download and read the paper for free. In addition, the article may be reused and quoted provided that the original published version is cited. This facilitates freedom in re-use and also ensures that Eur J Clin Exp Med content can be mined without barriers for the research needs.
