Before continuing, it will be better to explain here these basic statistical terms because I will decide whether my model is sufficient or not by looking at these figures. What is the Industry Email List p-value? P-value or probability value shows statistical significance. Say you have an assumption that your brand's average keyword CTR is 70% or higher and its p-value is 0.02. This means that there is a 2% chance that you will get CTRs from your branded keywords below 70%. Is this Industry Email List statistically significant? 0.05 is usually used for the upper limit (95% confidence level),
So if you have a p-value less than 0.05, yes! It is significant. The smaller the p-value, the better your results! Now let's look at the summary table. My 4 variables have p-values showing their relationships whether significant or Industry Email List not significant with the annual amount Industry Email List spent . As you can see, time spent on the website is statistically insignificant because its p-value is 0.180. It would therefore be better to abandon it. What is R Squared and Adjusted R Squared? R squared is a simple yet powerful metric that shows the variance explained by the model.
It counts all the variables you defined in X and gives a percentage explanation. It's something like your model abilities. Adjusted R-squared Industry Email List is also similar to R-squared, but it only counts statistically significant variables. That's why it's best to constantly watch the adjusted R-squared. In my model, 98.4% of the variance can be explained, which is really high. What is Coef? These are coefficients of the variables that give us the model equation. So it's over? Nope! I have a Time on Website variable in my model, which is statistically insignificant. Now Industry Email List I'm going to create another model and remove the Time on Website variable :