Linear Regression Analysis

Abstract

Background: Regression analysis is an important statistical method for the analysis of medical data. It enables the identification and characterization of relationships among multiple factors. It also enables the identification of prognostically relevant risk factors and the calculation of risk scores for individual prognostication. Methods: This article is based on selected textbooks of statistics, a selective review of the literature, and our own experience. Results: After a brief introduction of the uni-and multivariable regression models, illustrative examples are given to explain what the important considerations are before a regression analysis is performed, and how the results should be interpreted. The reader should then be able to judge whether the method has been used correctly and interpret the results appropriately. Conclusion: The performance and interpretation of linear regression analysis are subject to a variety of pitfalls, which are discussed here in detail. The reader is made aware of common errors of interpretation through practical examples. Both the opportunities for applying linear regression analysis and its limitations are presented.

The Relationship between Behavioural Changes, Cognitive Symptoms, and Functional Disability in Primary Progressive Aphasia: A Longitudinal Study

“…Some limitations in this study include the relatively small sample size [44], which was related to the need to restrict to participants who had complete data at the follow-up visits. Still, this sample size is in line with other published longitudinal studies in PPA.…”

Patientenverfügung bei Patienten mit Nierenerkrankungen im Vergleich zu allgemein-internistischen Patienten

“…Zur statistischen Diskriminanzanalyse der Daten diente das logistische Regressionsmodell mit einer Backward-Selection. Das Effektmaß der logistischen Regression, die Odds Ratio mit dem Konfidenz-Intervall (95 %-KI) gibt an, wie stark der Einfluss einer der erklärenden Variablen auf die Zielvariable ist [3,29]. Schließt das Konfidenzintervall die Zahl 1,0 nicht mit ein, hat der erklärende Faktor einen signifikanten Einfluss auf die Zielvariable.…”

Methodological and statistical considerations in alcohol policy evaluation

“…While this involved extensive data collection (288 data points) to develop the policy scores, as we noted in our paper, 2 the small number of study areas under investigation reduced statistical power and, by default, restricted the choice of analytic method. The use of multiple regression analysis, as suggested by Duffy, 3 is therefore substantially underpowered 4 ( n  = 8 in the analysis presented by Duffy) and, in our paper, we cautioned against extrapolating the data for this reason. Future studies using the TEASE-16 to examine policy data from a larger number of countries will be better placed to conduct multiple regression analysis and account for the relationship between policy and consumption and alcohol-related harms.…”