You recall our first model, first initial judgment which says the sales is determined by two factors, promotion and advertising? But let us say our question is whether we can improve that, you know, understanding the relationship among those factors and values. So, somehow I think that in addition to those two variables, in other words, promotion and advertising, probably the overall economic condition might have some impact on the size of the company. So, once I saw my first model, I, you know, sort of improved my understanding and I become more confident about my model, but using the knowledge, somehow I have to see whether I can further improve my understanding about this cause and effect relationship. So let's say I want to expand my thermometer. Now I want to include another variable, another factor which is index, [INAUDIBLE] index. And so now I think that my [INAUDIBLE] is assessed is. Not just this promotion. And advertising but also I think that there is impact of economic index. So, that's my second model. In other words, improved model. I somehow improved. Of course, you know, I don't know yet whether that's improvement or not. And I do the same thing and I get this digression model, digression analysis, and then eventually, I have these values. As I already talked about. If I answer the question about how confident am I. I got to look at these p values. And it's very interesting to note that all of these p values are very, very small. That means that these three variables indeed. Have a very strong cause and effect relationship with the sales, so it's pretty good. One interesting point is that as in our first model, commotion has a positive impact on sales. Advertising has also a positive impact on sales, but economic index has negative impact on sales. Meaning that if the economy becomes better the demand of that product is going down. Or if the economy becomes bad the sells of the product increase. So we see that the product is a sort of inferior product. In other words when people have more money, they don't usually consume that, they probably want to consume much more expensive item than that. You can think about what kind of a product might have that kind of, you know, relationship but anyway for the purpose of our discussion. It's enough to see that all this valuables are statistical strong. And also I wanted to get these [UNKNOWN] scale. Then it is larger than other the scale we had previously. And you probably recall that the scale we had initially was 0.5237, okay? So this is more, you know, much more improved, right? Because initially we had 0.5 to 0.37. Okay? So I would just say that I already mentioned that. Let's assume that our scale somehow measures our learning capability. Of course, I'm not saying that that's the only measure, I just want to explain the concept in simple terms. And we happen to have our scale, is pretty good. So if we use the regression analysis, our scale is somehow, you know, proxy [UNKNOWN] our understanding. Or the actors of the whole relationship. Okay? And now we want to also ask a question. Can I improve my model further? You have got to think about the ways, right? You have to use your management judgment further. What kind of additional factors, what kind of additional variables should be there in order to enhance our level of learning? So I think that probably based on your own experience and based on some brainstorming with your team members or, you know, your colleague, managers, and so on and so forth, you probably can think of a better model. But I would just say that I believe that there might be some relationship in between, previous period of the promotion and current period of the sales, in other words if I use T as Index, let's say, says T is the sales figure. In peer to the T and I think that this one might be affected by promotion in the previous period T-1. And likewise I think that it might be possible that the previous period of the advertising affects the current period of the sales. So, I, you know, further define my mental model. I further define my mental model and now the model says. Okay. So we can here you go. Our model say that the sales is determined by now five variables, beta one, Pt. Okay, now we need the T and beta two, 80 plus and beta three index T. And now beta four promotion T -1, beta five advertising T -1. As I, what a dimension T-1 implies, the previous period, right? If this is [INAUDIBLE] five [INAUDIBLE] this says in corner five is determined to buy quarter commotion in quarter five and also we believe that commotion in the previous period, quarter four. Advertising quote of four. Okay? [SOUND] And now, usually, as usual, we have to ask a question, how confident am I? I'm very much confident because all this [INAUDIBLE]. But, one thing very interesting. These size are the same as before. There is possible relationship between promotion and sales. Positive relationship between advertising and sales, and negative relationship between economic push back and the sales. What is interesting is, this one. Okay? This one is very interesting. I mean that this one is negative which implies that the previous years, previous periods of the promotion actually has negative impact on the current period assessed. On the other hand, the sales has still positive. In other words, the previous advertising, I'm sorry, advertising has positive impact on the sales. The previous period advertisings tier has positive impact on the sales, the current period sales. Can you explain why? Why? I think we, we enriched our understanding about this relationship greatly. To a great extent, right? Promotion is fundamentally different from advertising. It's easy to see, right? Promotion is something like that actually, you know, buy things. You are supposed to buy things next period, but you buy now because there is some promotion going on. So promotion is not actually generating users. It actually shifts the sales point, right? Says, the future sales occur now. It's not generating new sales. Well as advertising as sort of a long term impact. Another interesting point is that if you look at this magnitude, here, the current period advertising's 1.9 whereas the previous advertising has larger coefficient. That means that the previous period advertising has more impact, the larger impact on the current period, sale that the current period advertising. So as you can see, our understanding becomes richer and we know more about these complex cause and effect relationships. And moreover our [INAUDIBLE] improves. For that. That means that we are explaining more and more variances. More and more variations in the complex cause and effect relationship. Meaning that we know more about distance. So we become more confident. In our managerial list you are making regarding these values. So that's how we improve our learning capability, our cause and. Our understanding about the cause and effect relationship because clearer, better, more accurate. And that's what I mean by our learning capability becomes stronger. More and more noises are eliminated in doing so. Right? As we go through this thing. More and more noises. More and more wrong information, wrong data, wrong impression. Those things are eliminated and we see the true cause and effect relationship more clearly. And that's what I mean by this learning process. As we go through the process. Our learning, capability becomes stronger and stronger.