Thank you so so much!! I´m from Brasil and this is my last year in college. I was with problems to understand the conjoint analisys to make my final work for college conclusion. Without this work, I could not reach the required score. After a lot searching and reading several books, your explanation was the onlyone that I understood perfectly. I´ll be forever grateful!
Coinjoint analysis is based on the ranking of alternatives with combined attributes (product features). It helps you to break down the combination of attributes into preferences for each individual attribute (part worth utilities), not the other way around.
Some good content here in this channel! I hope you will come back and post more, I'm sure you still have a lot to teach, specially with so much happening in the world today!
This has to be decided by a group of customers, they have to do the ranking first. With conjoint analysis you can then find out, which individual product attribute (in my example color, memory or delivery time) is the most important.
You code memory also 1,-1. It becomes more complex when you have more than 2 levels, because the number of possible combinations will increase rapidity. Then you could code -1, 0, 1 etc.
Oh, I'm a 2nd year graduate student learning data analytics. You've explained in 10 minutes, what took 5 hours of lecture. Your instruction was clear and concise. Thanks again.
Can you please explain out of 8 possible attributes combinations, which combination is most preferred so that we can use for our product(smart phone) building strategy?
Where does the rank come from? Customers select from pairs of products. How do you get the ranking from that? What are the odds of having a scenario where people prefer product A over B, B over C but C over A?
jpaokx Conjoint analysis is not about which product the customer chooses, but to find out which features (attributes) of the product impact customer's decision to buy, i.e. the part worth utilities. If you want to solve a problem, like which product to buy based on a list of criteria, you can use a decision making method as AHP (Analytic Hierarchy Process). See my other videos.
Thanks you so much. Can we decide which attributes combination is best out of 8 with the help of conjoint analysis?I understand ranking will give us the result. But its a stated result by customers. Can we have buyer best preferred combination without using ranking?
Habe im Moment keine konkreten Plaene, das Video zu uebersetzen, bedeutet ziemlichen Zeitaufwand. Ich nehme die Anregung aber gerne als Feedback auf. Falls noch mehr Viewer den gleichen Wunsch aeussern, koennte ich das ins Auge fassen.
hi! after I finished watching this video, I did some more research and found that in some sites the method is used by enumerating(1,2,3,..) the choices...trying to understand the method better, I swapped -1 with 2 and found that relative frequencies changed substantially!!!Can you explain this differences and elaborate\suggest a course of action?
@@GeeJamz Pathworth utilities (beta) are: -0.5 for color, -2 for memory and -1 for delivery. variation (x) is +/-1 +/-1 * -0.5 = 1 for color, +/-1 * -2 = 4 for memory and +/-1 * -1 = 2 for delivery. Total 1 + 4 + 2 = 7. Color: 1/7 = 14%, memory 4/7 = 57% and delivery 2/7 = 29%. Download my excel where you can see the detailed calculation: bpmsg.com/conjoint-analysis-example-in-excel/
hello, I am doing a research and the my scenario is I have, 6 attributes and each attributes with 5 levels ,so that is 15,625 combinations in full factorial. How do I go for solving it , and what should be my values for 5 levels (i.e -1,0,+1 ....etc )? And also full factorial makes it difficult for me to get ranks from respondants, so which conjoint method is best suited for my scenario or should I switch to AHP or LGPM ?
Maybe conjoint analysis is not the best method for your project. levels could be -2, -1, 0, 1, 2, but the problem seems to be, how to reduce the number of combinations.
Klaus Many thx I love your video on stated preference conjoint analysis! I ve just visited your page can't donate due to tim€ :) However if you or other team member visits #Cyprus Consider you have free accommodation in the region I promote Reg g
13 Jahre später und genau dieses Video habe ich gebraucht. Vielen Dank dafür!
Thank you so so much!! I´m from Brasil and this is my last year in college.
I was with problems to understand the conjoint analisys to make my final work for college conclusion. Without this work, I could not reach the required score.
After a lot searching and reading several books, your explanation was the onlyone that I understood perfectly. I´ll be forever grateful!
Ola, tudo bem?
Estou com dificuldades pra aplicar esse conceito em minha pesquisa, será que você poderia me dar uma força?
Ficaria muito grato
Coinjoint analysis is based on the ranking of alternatives with combined attributes (product features). It helps you to break down the combination of attributes into preferences for each individual attribute (part worth utilities), not the other way around.
I like the simplicity and clarity. The visuals are also very good. Thanks for uploading.
Some good content here in this channel! I hope you will come back and post more, I'm sure you still have a lot to teach, specially with so much happening in the world today!
As you have to rank the alternatives from 1 to 8, the constant is the average of {1,2,...8} = 4.5
Thanks a lot for your wonderful teaching!
Really the Great Explanation for Conjoint Analysis with the Practical Example
This has to be decided by a group of customers, they have to do the ranking first. With conjoint analysis you can then find out, which individual product attribute (in my example color, memory or delivery time) is the most important.
You code memory also 1,-1. It becomes more complex when you have more than 2 levels, because the number of possible combinations will increase rapidity. Then you could code -1, 0, 1 etc.
For 3 levels you could use -1, 0, +1 for example. But more than two level it becomes quite complex
Das ist sehr gut! Vielen Dank.
You are welcome! More info also on my blog bpmsg.com (slides and an example in Excel)
Thanks professor!
Info on conjoint analysis is really hard to find. Thanks for simplifying!
Great explanation!
Thank you so much for this video. You teach wonderfully.
Oh, I'm a 2nd year graduate student learning data analytics. You've explained in 10 minutes, what took 5 hours of lecture. Your instruction was clear and concise. Thanks again.
thanks very much brow...I understand know all of the presentation...
pure gold lecture
Can you please explain out of 8 possible attributes combinations, which combination is most preferred so that we can use for our product(smart phone) building strategy?
You are welcome. Thanks for the feedback. Feel free to visit my website bpmsg dot com.
Where does the rank come from? Customers select from pairs of products. How do you get the ranking from that? What are the odds of having a scenario where people prefer product A over B, B over C but C over A?
So, finally, which product should the customer choose? I see that there is a lot of explanation which didn't end up with the solution!
jpaokx Conjoint analysis is not about which product the customer chooses, but to find out which features (attributes) of the product impact customer's decision to buy, i.e. the part worth utilities. If you want to solve a problem, like which product to buy based on a list of criteria, you can use a decision making method as AHP (Analytic Hierarchy Process). See my other videos.
Vielen Dank, thank you very much! Very descriptive video
Thanks for explaining!
Excellent
Thnx professor
Really simple explanation, thx
Fantastic!
what if the attributes are 3*2*2? what will happen to the factorial design. Do I still need to use 1 and -1?
one more question, it should be described by cube or can be described by other shape? thank you very much
Thanks you so much. Can we decide which attributes combination is best out of 8 with the help of conjoint analysis?I understand ranking will give us the result. But its a stated result by customers. Can we have buyer best preferred combination without using ranking?
Habe im Moment keine konkreten Plaene, das Video zu uebersetzen, bedeutet ziemlichen Zeitaufwand. Ich nehme die Anregung aber gerne als Feedback auf. Falls noch mehr Viewer den gleichen Wunsch aeussern, koennte ich das ins Auge fassen.
Excellent
If the levels is more than 2 how do you assign the values for the matrix? In this case it was -1, +1 to distinguish between them, isn't it?
hi! after I finished watching this video, I did some more research and found that in some sites the method is used by enumerating(1,2,3,..) the choices...trying to understand the method better, I swapped -1 with 2 and found that relative frequencies changed substantially!!!Can you explain this differences and elaborate\suggest a course of action?
Y= 4,5 - 0,5 X color - 2 x memory - 1 x delivery...
can you explain to me how you get 4,5 please? thanks very much
Thank you very much for this. How did you calculate the relative weight please?
See at 6:30
@@BPMSG Thank you, sir. I didn't understand that part, pls can you explain further?
@@GeeJamz
Pathworth utilities (beta) are:
-0.5 for color,
-2 for memory and
-1 for delivery.
variation (x) is +/-1
+/-1 * -0.5 = 1 for color,
+/-1 * -2 = 4 for memory and
+/-1 * -1 = 2 for delivery.
Total 1 + 4 + 2 = 7.
Color: 1/7 = 14%, memory 4/7 = 57% and delivery 2/7 = 29%.
Download my excel where you can see the detailed calculation:
bpmsg.com/conjoint-analysis-example-in-excel/
@@BPMSG Thank you so much for this. I really appreciate it.
@@BPMSG How do you calculate the variation for each value? Your excel sheet seems to omit that calculation
hello, I am doing a research and the my scenario is I have, 6 attributes and each attributes with 5 levels ,so that is 15,625 combinations in full factorial. How do I go for solving it , and what should be my values for 5 levels (i.e -1,0,+1 ....etc )? And also full factorial makes it difficult for me to get ranks from respondants, so which conjoint method is best suited for my scenario or should I switch to AHP or LGPM ?
Maybe conjoint analysis is not the best method for your project. levels could be -2, -1, 0, 1, 2, but the problem seems to be, how to reduce the number of combinations.
Any suggestion of methods which can used in place of conjoint ? like LGPM or AHP
I am very familiar and prefer AHP. You can find information and a free SW and Excel template on my website. bpmsg dot com
Thanks a lot Professor .
Könnten Sie das Video freundlicherweise auch auf deutsch anbieten? Vielen Dank!
Sir why Y= 4.5 -( all...) how come u take 4.5?
It's the result of the linear regression (calculation)
Klaus
Many thx I love your video on stated preference conjoint analysis!
I ve just visited your page can't donate due to tim€ :)
However if you or other team member visits #Cyprus
Consider you have free accommodation in the region I promote
Reg
g
You use 32GB then the next is 16mb???? What ??
32 and 16 MB or GB - just an example.
@@BPMSG can I change the 16 into 32 so it's look the same 😄 thank brothe