Hi Jens, your explanation of SPSS really helped me and since I have a big SPSS deadline due in about two weeks (and this may be a bit of a strange question), it would be absolutely amazing if you would find the time to help me out a bit, if you want! I am really struggling and because my lecturers can’t really provide me with good feedback for my exam of course, I cannot find anyone who thoroughly understands SPSS and can help me out! Please, let me know! I would be so incredibly thankful. Kindest regards, Donna from The Netherlands!
To assess the significance difference between socio-demographic factors and consumer preferences, how would you utilize the conjoint analysis results in ANOVA testing?
In your case, as I understand it, I would propose to not conduct the conjoint analysis as in this tutorial but rather slightly transform your data and use it in regression estimation. This way you could introduce additional factors into the regression and thus conjoint analysis and use a stepwise approach to see in how far they impact the results, i.e., wtp or other measures.
Hi Jens, I had a problem with output of my SPSS. I Have 6 factors and 2 levels each in conjoint analysis using SPSS it got 8 cards and it runs well, when i have 7 factors and 2 levels each (12 cards) the standard error was empty or blank and we got warning *There are not enough degrees of freedom to compute the standard errors. No reversals occurred*. and i tried to change 6 factors with 2 levels and 1 one factor with 3 level it runs well. Could you help me to solve this Best Regards Sukma
The simple answer would be that by using the standard command as presented in this video it is not possible. However, there are some workarounds: The simplest solution would be to run a (multinomial) logit regression, controlling for potential problems that might arise. Alternatively, you can first translate the choices into a sequence of preferences and then use the datatype SEQUENCE. Here you would however have to refer to external approaches to transform your dataset. Aside from these two approaches there are some additional alternatives which however cannot be easily realized in SPSS.
@@jensk.perret6794 Great video, however I am still wondering how to do this for a choice-based conjoint analysis. Especially how to input your survey data is not clear. Could you please show us how to do that in another video? Thanks again!
@@dionvanrooij5301 As mentioned SPSS only offers three possibilities to read the data. SEQUENCE requires an ordering for all data points. RANK requires a ranking of all data points and SCORE requires an evaluation of the data points. If you can order your answers because they are all transitive then you can use SEQUENCE. If there are however logical inconsistencies then you need to apply a different approach first to generate an ordering or a ranking. In this regard SPSS cannot help you.
so it is okay if the number not in sequence? if I have 8 cards but I've taken the data preferences of the variables from the likert scale so there will be some cards have the same number.
Hi Jens, your explanation of SPSS really helped me and since I have a big SPSS deadline due in about two weeks (and this may be a bit of a strange question), it would be absolutely amazing if you would find the time to help me out a bit, if you want! I am really struggling and because my lecturers can’t really provide me with good feedback for my exam of course, I cannot find anyone who thoroughly understands SPSS and can help me out! Please, let me know! I would be so incredibly thankful.
Kindest regards, Donna from The Netherlands!
Are you studying at RUG haha
can I ask what the SPSS file with the model in it looks like and how was this compiled?
To assess the significance difference between socio-demographic factors and consumer preferences, how would you utilize the conjoint analysis results in ANOVA testing?
In your case, as I understand it, I would propose to not conduct the conjoint analysis as in this tutorial but rather slightly transform your data and use it in regression estimation. This way you could introduce additional factors into the regression and thus conjoint analysis and use a stepwise approach to see in how far they impact the results, i.e., wtp or other measures.
Hi Jens, the utility are negative for all levels of atributes. Are these results ok?
In this case all categories are perceived to be worse than the base category which might actually be the case.
Hi Jens, I had a problem with output of my SPSS. I Have 6 factors and 2 levels each in conjoint analysis using SPSS it got 8 cards and it runs well, when i have 7 factors and 2 levels each (12 cards) the standard error was empty or blank and we got warning *There are not enough degrees of freedom to compute the standard errors. No reversals occurred*. and i tried to change 6 factors with 2 levels and 1 one factor with 3 level it runs well. Could you help me to solve this
Best Regards
Sukma
Same problem!! :(
Hi Jens, could you also explain how this would be done for a choice-based conjoint analysis?
The simple answer would be that by using the standard command as presented in this video it is not possible.
However, there are some workarounds:
The simplest solution would be to run a (multinomial) logit regression, controlling for potential problems that might arise.
Alternatively, you can first translate the choices into a sequence of preferences and then use the datatype SEQUENCE. Here you would however have to refer to external approaches to transform your dataset.
Aside from these two approaches there are some additional alternatives which however cannot be easily realized in SPSS.
@@jensk.perret6794 Great video, however I am still wondering how to do this for a choice-based conjoint analysis. Especially how to input your survey data is not clear. Could you please show us how to do that in another video? Thanks again!
@@dionvanrooij5301 As mentioned SPSS only offers three possibilities to read the data. SEQUENCE requires an ordering for all data points. RANK requires a ranking of all data points and SCORE requires an evaluation of the data points. If you can order your answers because they are all transitive then you can use SEQUENCE. If there are however logical inconsistencies then you need to apply a different approach first to generate an ordering or a ranking. In this regard SPSS cannot help you.
From where doyou get that code
If work through the SPSS documentation from IBM you can piece it together.
so it is okay if the number not in sequence? if I have 8 cards but I've taken the data preferences of the variables from the likert scale so there will be some cards have the same number.
As I understand your question you can simply use the datatype SCORE and it should work for your data.
@@jensk.perret6794 Waa thank you for the response. I really appreciate it.
Thanks. Please mind the correct pronunciation of "variables" ;-)