Рет қаралды 10,220
Simple linear regression , statistical method, relationship, two variables, outcome, dependant variable, predictor, explanatory variable, fit best regression line into the data, generate regression equation, predict outcome from predictor, normality assumptions , Q-Q plot of residuals , Cook’s distance, regression model, adjusted R square, r , ANOVA test, standardised estimate and confidence interval for estimate, coefficient of correlation , strong positive correlation, hypothesis testing, slope, P-value, significant, intercept, equality of variance, Regression equation, Y=b0+ b1*x.
Simple linear regression is a statistical method used to determine if there is a relationship between two variables one of the variables is an outcome or a dependant variable and the other valuable is a predictor or an explanatory variable.
The objective of the simple linear regression analysis is to fit the best regression line into the data and then generate regression equation and from the regression equation it will be possible to predict the outcome from a given predictor.
The test will determine whether a predictor can predict the outcome. I have two variables the first one is the age which is a predictor or an explanatory or an independent variable and an outcome which is systolic blood pressure which is a dependant variable and order response variable both valuables has been checked for normality and the assumptions for performing simple linear regressions will be not discussed or explained in this video.
To perform simple linear regression analysis and to find out whether the age can predict the systolic blood pressure.
In the analyse section click on the regression and then select linear regression and in the new window appear move systolic blood pressure into the dependent variable and the age into the covariates spontaneously.
2 tables are generated at for linear regression.
Scroll down to model builder and click on
For assumptions check tick on the normality test and check on the Q-Q plot of residual and check on residual plot and also Cook’s distance for the model and tick on the adjusted R square
Tick on their ANOVA test, standardised estimate and confidence interval for estimate
looking at the table generated from the outcome of linear regression analysis for the two variables the outcome which is systolic blood pressure and the age which is a predictor, Model 1 shows that a value of 0.7 and this represents the coefficient of correlation between the two variables and the value is high which indicates that there is a strong positive correlation between age and between the systolic blood pressure an increase, in age will lead to increase in systolic blood pressure are square.
Square of R means that 53% off the variation in the systolic part pressure which is the outcome can P-value is calculated and is less than 0.05
Second table is the ANOVA table shows that there is a P-value less than 0.05. The table test the hypothesis that the slope is not equal to zero and it shows that the P-value is significant so there is a relationship between the systolic blood pressure and the age.
Third table shows the model for coefficient and the first value which is the intercept is equal to 86.2 and the slope is equal to 1.1. This value means that with every increase in one year of age there is an increase of 1.1 in the systolic blood pressure. The P value here is significant less than 0.05 .
Q-Q plot shows that Y axis is standardised residual and on the X axis is theoretical quantiles residual. Second plot shows that there is deviation from the equality of variance
Regression equation can be generated as the intercept is given and the slope is given
Y=86.2 + 1.1*X
This regression equation can be used to predict the outcome based on a given value for the predictor or the age.