Now that we have seen the mathematical method, we are going to see how to carry out the regression with Síagro with the following data:
To study the regression between the Average Daily Gain (ADG), the dependent variable, and the Average Daily Consumption (ADC), the independent variable, we only have to access to Prediction Models / Linear from the control panel and select our variables. It’s that simple, and the application will return the following output:
|(Intercept)|| -23,0443|| 97.7|| -0.236|| 0.819|
| CMD|| 0.4986|| 0.091|| 5.48||0.000588|| ***|
| 0.79|| 0.763|| 25.9|| 30||0.000588|| 1|| -45.6|| 97.2|| 98.1||5.36e+03|| 8|| 10|
We are only going to explain some of the data that appears in this output, leaving a broad explanation for the next article (but do not hesitate to contact our team if you have further doubts). What would be the equation of the regression line? Just recall that in our case the variable y is y = ADG and the variable x is x = ADC.
ADG = -23.0443 + 0.4986 ADC
That is, the estimated coefficients of a and b are what Síagro (and R) calls Estimate. From the above equation we can say that for each unit of ADC, the ADG increases 0.4986.
But … can we really trust that relation? Or in another way: is the coefficient b different from 0?
This question involves going from description to inference, as we said at the beginning of the article. This is answered with a t-test with null hypothesis b = 0 and alternative b ≠ 0. In this case, the p-value is less than 0, 05 (p-value: 0.000588) and therefore we reject the null hypothesis, that is, b is different from 0.
We can also say that the CMD explains 78.96% of the GMD.
We will see in the next article further explanations of variability. We recommend you to be subscribed to our newsletter if you do not want to miss it!!