Speaker 1: Hello, welcome to my easy statistics. In this video, I'm going to discuss about conjoint analysis. Conjoint analysis is a survey based statistical technique used in marketing research that helps to determine how people value different attributes like features, functions, benefits that make up an individual product or service. For discussing this conjoint analysis, I'm taking a very simple example of cool drink. Here we have three attributes are there, flavor, quantity and price. In each attribute, I have taken three elements like a flavor, apple, orange and mango flavors we have taken and for quantity, there are three quantities, 250 ml, 500 ml, 1000 ml. Price less than 50 rupees and 51 to 75 rupees, above 75 rupees. Conjoint analysis will be conducted in two stages. In the first stage, we will create the combinations of these three attributes and in the second stage after the survey is conducted on these combinations, we will try to understand which is the most preferred combination and which is the least preferred combination so that the marketing people can introduce the most preferred combination in the market. So in the first stage of conjoint analysis, first we will create the combinations of the attributes. For creating the combinations, you must select data. In this data, select orthogonal design. In this, generate. Now this is the output screen. First we must create the three attributes. The first attribute is flavor, add it. Second attribute is quantity, add it. Third one is price. So first we have added three attributes. Now for each attribute, we need to add the levels. The flavor, we will add first level is apple, second one is orange, third attribute for this one is mango. Click continue and next for quantity, define value. First one is 250 ml, second level is 500 ml, third one is 1000 ml. Click continue and next price, define value. First one is less than 50 rupees, second one is between 51 to 75 rupees and third one is above 75 rupees. Click continue. Now we have three attributes with the levels. Next is we must create a data set for this combinations. We will call it as cool drink. Now after creating the data set, now we need to click on this options. Here comes the main thing. Minimum number of cases to generate. Here totally we will be getting 27 combinations because three attributes and each attribute is having three levels. So 3 into 3 into 3, totally 27 combinations we are going to get. But all combinations we may not use it because some combinations company is not interested and in some combinations customer may not prefer. So let us take 10 combinations we will see and in this minimum number of combinations that must be created that we call cases to generate is at 10. Number of holdout cases, holdout cases means we will create a combination and give the combination to the customer for giving preference but we don't use that combination in the analysis. So in this case I am not using any sort of holdouts. So I am expecting minimum number of cases to generate is 10. Click continue and click ok. So you can see here the plan is successfully generated with 16 cards. So let us see that. So this is the combinations which we got and let us see the data view. In the data view we will be seeing totally 16 cards. Cards means combinations and cards are given as a numbers they are given from 1 to 16. In this Cooldrink file system has created two variables automatically. The first one is card. Card is nothing but the combination number. You can see the first combination the system has created with mango flavor 250 ml price 51 rupees to 71 rupees. So in this way this card variable is automatically created which we have 16 combinations are there. The second variable which is automatically created is status. Here we have options like design and hold. Design means this combination will be used in the preference for the customers and this will be used in the analysis. If in case if I make this status as a hold it means combination 1 will be given to the respondent. But the preference will not be considered in the analysis means this combination will not be considered in the analysis. In this case I am not keeping any hold cases all I am taking is a design. Now you can see totally we have 16 combinations. In this combinations some combinations the company will not be interested. For example you can see apple which is given like you can see this one apple 1000 ml less than 50 rupees. But the company is not interested in giving this combination because it may not work for the company and customer may not prefer some combinations. For example you can see apple 250 ml above 75 rupees. So this combination customer may not prefer. So this also we are going to delete. In this way we must actually check all the combinations. Some combinations company may not be interested in some combinations customer may not prefer. Now we are interested only in taking 10 combinations we will even delete this all remaining combinations. Only 10 combinations I am taking and these 10 combinations must be given to the respondents and survey must be collected from them and these respondents will give their preference for each combination and after collecting the survey data that must be entered into SPSS. I have already created one file where we have 23 respondents are there customer ID you can see customer ID 1 2 3 till 23 customer respondents are there and see the first customer he has given preference from P1 to P10 and first preference is given to combination 8 second preference is given to combination 5 third preference to combination 9. In this way each customer has given their preferences from 1 to 10 and they are totally 23 respondents. Now we will be running this conjoint analysis on this file which I have named it as a preference. Now we will start the analysis conjoint analysis.