11.10
Multiple regression is a statistical tool used to analyze the relationship between more than two variables.
Multiple regression can be modeled into a simple equation that estimates the linear relationship between the response or dependent variable, with more than one predictor or independent variables.
For example, the water consumption of athletes is positively correlated with both the temperature and the total amount of time practiced.
Here, the temperature and the total amount of time practiced are the predictor variables that can be independently set. The water consumption is the response variable as it depends on the other two variables.
Since the manual calculation of the multiple regression equation is generally complex, software is used to solve it.
The multiple coefficient of determination is calculated to measure how well the equation fits the data set. It means that the changes in temperature and the total amount of time practiced can explain 97% of the variation in water consumption.
However, as more variables are used, R2 generally increases.
In such cases, the adjusted coefficient of determination is calculated, which accounts for the sample size and the number of predictor variables.
Bij meervoudige regressie wordt een lineair verband geanalyseerd tussen één respons- of afhankelijke variabele en twee of meer onafhankelijke variabelen. Deze methode kent diverse praktische toepassingen.
Landbouwkundigen kunnen meervoudige regressie toepassen om de gewasopbrengst te voorspellen op basis van meerdere factoren, zoals de beschikbaarheid van water, bemesting, bodemeigenschappen, enzovoort. In dit geval is de gewasopbrengst de afhankelijke variabele, aangezien deze wordt beïnvloed door de onafhankelijke variabelen. De analyse vereist het opstellen van een spreidingsdiagram, gevolgd door een meervoudige lineaire regressievergelijking om de meervoudige determinatiecoëfficiënt, R^2, te berekenen. Stel dat de waarde van R^2 96% bedraagt; dan kan men concluderen dat de verschillende combinaties van water en bemesting 96% van de variatie in de gewasopbrengst verklaren.
De waarde van R^2 neemt echter toe naarmate het aantal onafhankelijke variabelen toeneemt. Daarom wordt tijdens de analyse een aangepaste determinatiecoëfficiënt gebruikt, die zowel rekening houdt met de steekproefomvang als met het aantal onafhankelijke variabelen.
Multiple regression is a statistical tool used to analyze the relationship between more than two variables.
Multiple regression can be modeled into a simple equation that estimates the linear relationship between the response or dependent variable, with more than one predictor or independent variables.
For example, the water consumption of athletes is positively correlated with both the temperature and the total amount of time practiced.
Here, the temperature and the total amount of time practiced are the predictor variables that can be independently set. The water consumption is the response variable as it depends on the other two variables.
Since the manual calculation of the multiple regression equation is generally complex, software is used to solve it.
The multiple coefficient of determination is calculated to measure how well the equation fits the data set. It means that the changes in temperature and the total amount of time practiced can explain 97% of the variation in water consumption.
However, as more variables are used, R2 generally increases.
In such cases, the adjusted coefficient of determination is calculated, which accounts for the sample size and the number of predictor variables.
From Chapter 11:
Now Playing
Correlation and Regression
3.6K Views
Correlation and Regression
12.7K Views
Correlation and Regression
8.0K Views
Correlation and Regression
6.7K Views
Correlation and Regression
7.7K Views
Correlation and Regression
5.4K Views
Correlation and Regression
7.4K Views
Correlation and Regression
4.9K Views
Correlation and Regression
6.5K Views
Correlation and Regression
2.6K Views