Hi all,
I'm new in this forum and new in the study of the statistic too.
I have a very basic question about the Time Series analisys: what is the correct way to use the linear regression to understand the trend of the serie?
For example, I have two series:
In the first case, I have positive slope and the regression line is all in the positive part, so I can deduce that the trend is ok.
In the second case I have positive slope too, but the regression line is in the negative part.
So my confusion is this, the positivity of the trend doesn't tell me the "positivity" of the trend. In every of the two cases I have positive trend, but If I calculate the future value of the series (forecasting) I have two different trend.
Which is the correct mode to analize the trend of the series using the slope of the RL? Maybe there is something that I haven't understood!
Thank you to all.
Igor.
I'm new in this forum and new in the study of the statistic too.
I have a very basic question about the Time Series analisys: what is the correct way to use the linear regression to understand the trend of the serie?
For example, I have two series:
In the first case, I have positive slope and the regression line is all in the positive part, so I can deduce that the trend is ok.
In the second case I have positive slope too, but the regression line is in the negative part.
So my confusion is this, the positivity of the trend doesn't tell me the "positivity" of the trend. In every of the two cases I have positive trend, but If I calculate the future value of the series (forecasting) I have two different trend.
Which is the correct mode to analize the trend of the series using the slope of the RL? Maybe there is something that I haven't understood!
Thank you to all.
Igor.