 We have a good model with a very high adjusted R square of ninety point nine six two which says that the model accounts for roughly ninety six percent of the variation in sales, but recall that we also had one of our dummy variables for August was not significant. So what we probably should do is to rerun the model without August and where we're going to do that is just to delete the August column. Here I've switched over a copy of the data. I'm just going to select the August column and delete it and everything moves over. I'm going to go to data, data analysis, regression, bring up my box and I need to get my y-range, go over here and get my sales, those in and now I need to select my x variables which are all of the dummy variables in the period, the month. Get those all in, make sure they match. We have labels and I'm going to put the output and P1 and just click okay. Now I'm going to zoom over here. Okay, I've expanded the regression output and changed those scientific notations for the p-value. I'm going to move these into ordinary general format and you can see that we're still significant and all of the coefficients intercept down here are statistically significant. So that means the model might be better, but if we look up here on the R and R square and I'm going to reduce the decimals there a bit. We have an adjusted R square of 0.9616 and if I recall I'm going to click back here. The adjusted R square for the model with August was 0.9622, so that R is just a little bit better, R square is a little bit better than this R square. So we're better off to use this model to do our final forecast. We're going to forecast ahead for 12 months, so I'm going to select those three so I can drag my series down to October of 13. There's October of 13. In order to get the model we need to copy down our dummy variables as well. So we can go back up here and select these dummy variables from November to October. Copy those and then I'm going to paste those in so now we're set up and all we have to do is drag down the regression equation and we've got our values forecast ahead for 12 months using our best fit trend and seasonality regression equation. I added those data points down here to the bottom of our data for the scatter chart and you can see now here's our forecast ahead that has both the trend and the seasonality and that'll give us a pretty good, the forecast should have about 96% of the variation accounted for. So that's pretty doggone good.