Now we are back to the analysis. So we are back to analysis. And if you look at your level 2 model here, okay? You'll have that your intercept can have a overall mean plus a random error, but your slope the, beta1 does not have a random slope activated yet. So you needed to double click on this mu here, mu1 to activate the random slope. So you'll see if you double click it, now your mu1 is black, it's not grey anymore. We are ready to run our model. Just click on the run the model, and we'll see what will happen. Hm, HLM is not very much friendly, so you will get this screen that's black with a bunch of things and then probably you'll get an information about convergence. So you needed to click on y on your keyboard and then the program will keep it running. Well, once you get that what will happen is you'll not see the black screen anymore. How do we get the output? You are needed to go to File and all the way down to View Output. If you click on that then you'll have the output file. Yeah, I know it takes a few steps to get to the output, but that's how HLM was developed. So the first thing that you see is this information about your sample. So 600 observations at level-1, 200 observations at level-2. My outcome variable is PERFORMAnce. And if we scroll down your output file a little bit, you'll see the equations. The level-1 equation, the level-2 equation, and a mixed model equation. It's important to notice that the level-2 models, we have one for the intercept and one for the slope. This gamma 00 is the overall average for the intercept. If there is variance across teams, across groups, in this case across departments, that's represented by mu. For the beta 1, this gamma 10 is the mean of the slopes. If the slopes vary across department that will be represented by your mu1, okay? So now that we are here, we are looking at your variables. We are looking at your mixed model equation. And, let's take a look at the results, the coefficients. If we were to go all the way down to the final estimation of fixed facts, with robust standard errors. And, I want you to take a look at the coefficient for the intercept, and here the coefficient for the intercept is 0.17, okay, with a standard error of 0.34. If we move to the p-value, we have that we find that they intercept, is not statistically different from 0, because p is not the less than 0.05. And if we look at the slope beta1, the coefficient is 0.44, the standard error 0.04. And if we look at the p-value, the p-value is the last at 0.05. So we do have a significant slope for job satisfaction, the effect of job satisfaction on performance. And now we needed to see if this coefficient, the intercept, and the slope vary across departments, or between departments. And we look at this part of the output here. We notice that for the intercept, our mu0, the standard deviation there is 1.54. Well, if we look at the p-value, we find that the standard deviation is definitely different from 0. So what that means is that, well, although our intercept, the overall intercept is not different from 0, statistically different from 0, we do have enough variance between departments. Or a variance between the departments that is making, this is meaningful information. All the departments, they don't have the same starting point, the same intercept. But if we look at the job satisfaction, the slope here, the standard deviation is very small, it's 0.05. And if we look at the p-value, we see that p-value is not less than 0.05. So what we find here is that well the slopes across departments are pretty much the same. Although we do have a significant slope, the slope is different from 0 across all the departments, they tended to be pretty much the same. There is not much variance in the slope. So what we can conclude here is that, well, the intercept is not different from 0 for most, for all, when we take all the departments together, the intercept is pretty much close to 0. But there is variance across departments, with the slope, we find that the slope is significant. So the slope is different from 0, that slope's not flat. But we also find that that slope does not change based on the department. So the slope is pretty much the same across all the departments. Well, in this session, we covered, we ran a HLM model. We ran a multilevel analysis model and we showed that we can have models in which the slope vary across departments or across groups. We can have slopes that are statistically not different from 0 but we could have variance on how the slopes happen across departments. We have differences in slope and we have differences in intercepts. We also talked a little bit about centering, which is an extremely important topic for HLM and multi-level analysis. And we mentioned that we can have grand mean centering or group mean centering. I hope that this session was enough to keep you interested in multi-level analysis. This is a complex analysis that we can have mediation models, interaction models, moderation models, conditional indirect effect models. This is the simplest type of analysis that we can have at multi-level, in which we look at the direct effect.