Speaker 1: Hello, everyone. Upon your request, I'm going to review logit regression analysis in marketing engineering. Here, the name is customer choice. Same thing, right? Customer choice is a logit regression model. And a logit regression model is a type of multiple regression model. But the difference of the unique point of logit regression is that it has dummy variable for its dependent variable. Here in the data table, we have customer choice, your choice. And the value is either 0 or 1. So it's a dummy variable. Nominal scale or categorical scale. So logit regression can deal with nominal scale or categorical scale, particularly with two values. So anyway, we can use this as our dependent variable using logit regression here. So 1 here, 1 means I will purchase the product. And 0 means I will not purchase the product. And the model is together with a group of independent variables. We have here 1, 2, 3, 4, 5 different independent variables that influence the dependent variable. So variables such as past purchases, expensive, convenient, service, large choice. So in our hypothesis, we can think that these can be influencing the dependent variable. For example, if you have purchased a product in the past, in the future, there is more chance for you to purchase the product. So it'll be 1 when you have certain number in the past purchases. Or when you have better service, more likely to purchase the product. So this will influence the choice, too. So we have our own theory. And the number of hypotheses we can create is 1, 2, 3, 4, 5, the number of independent variables. And we are ready with the data set here so we can run the analysis using marketing engineering. So go to Add-ins and Marketing Engineering. And Customer Choice, load it. And click Run Analysis. So I will not touch anything on this pop-up. Just use default option and click Next. All right, so I accidentally touched the cell, so I need to highlight the table manually, right? Sorry about that, but I can do that. OK, I'm going to highlight the whole table just like this. OK, so this whole table is being highlighted. And you can click OK. All right, so I have several different tables in the result, but let me click Segment here and look at the first table here. This is the major result from the analysis. So we have five different independent variables and their coefficients and their t-statistic. And we don't have p-value reported here, but that's OK because we have t-statistic, right? So to be significant of each hypothesis, t-value should be greater than 2. In the case of p-value, it should be less than 0.05. But t-value should be greater than 2 normally. We can find the relationship between t and p from online. There are a lot of different free tables out there, right? But anyway, look at the statistics and find value greater than 2. This one and this one, right? This one and this one, the two. The first two variables or first two hypotheses are significant. So we can make a conclusion regarding the first two variables. And we can say that the past purchases influence customer choice. And the price, particularly the price is expensive. They are less likely to buy the product, right? It's a negative relationship because we have negative value here in the t-statistics and coefficient, right? So there is negative relationship between the expensive price and customer choice, OK? So we can test each hypothesis using p-value or t-statistics. And to complete our regression equation, we can use coefficient, right? Coefficients. These are the unstandardized coefficient, b's. So we can input these numbers into the form, right? Form of regression equation. So y equals, for example, b0 plus b1 x1, the first variable, plus b2 x2, plus b3 x3, plus b4 x4, plus b5 x5, right? And then we can input b's, OK, from these numbers, OK? In the test, you may see a table more similar to the one in our online learning material, right? So maybe p-values will be reported there. And also with b's and betas, OK? But please don't use betas for your regression equation, right? We use b's, unstandardized coefficient, for our regression equation, right? So this was a simple review of loaded regression in marketing engineering, right? Good luck. Thank you very much.