L9(3,4) table desinged by SPSS is different from the table in google image that google image contains 4 colum (1 blank colum), and the ordinary of variables are different. what does the differences of variants order could influence?
@@jensk.perret6794 the second table in this page. there are 4 columns in the table and ordinary of the variant levels are different. and i tried to creat 4 variant 3 level table by SPSS, and there are 13 cards totally. a L9 (3,4) table like the image that i refferenced can do 3 and 4 variant analysis. which table is more reliable? www.google.com/search?rlz=1C2EJFA_zh-CNUS760US760&biw=1279&bih=650&tbm=isch&sa=1&ei=i6PbW6_zJ-THjwSww6q4Cg&q=orthogonal+table&oq=orthogonal+table&gs_l=img.3..35i39k1j0i19k1j0i8i30i19k1l3j0i5i30i19k1.42400.43413.0.44015.18.7.0.0.0.0.127.693.4j3.7.0....0...1c.1.64.img..16.1.126....0.aeHD-fOgsBw#imgrc=BXXVCSe-rQZbxM:
@@chideleyve533 If you generate an orthogonal design you get the minimal number of cards that are needed to generate a stable output. To increase the stability of the output you can increase the number of cards by using additional hold-out cases. Thus, if you tell SPSS to include a certain number of hold-out cases you will get a larger table.
If I understand your question correctly you are looking for a way to analyze the data you collected for an orthogonal design you created. In this case you can look up the video on how to run a conjoint analysis in SPSS: ua-cam.com/video/LZ1aMpidn0Y/v-deo.html If this is not what you are looking for, then please rephrase your question.
Hai Sir, how can i make all the total 9 card into several choice set, for example; that 9 cards has grouped into 3 choice set and each choice set have 3 cards, how can i do it with using SPSS. I am already generate 16 cards using Orthogonal Design SPSS.. but it is to many to be rank by respondents in a one row. (from the most preferred until the less preferred) I am already setting the minimum card number which are 15 and the holdout number are 5, then SPSS generate 16 Design with 5 Holdout. so, it is to many to be rank in a one row by respondent. can it be divide/ group into several group? and what is the right method to be use? I am very appreciate if you can answer my question. thanks Sir.
Dear Ms. Mizuki, in the way I understand your question I would either abstain from using ranking and have the interviewee evaluate the different designs on a given scale and then in the context of the conjoint analysis. (You see this in my video on using the conjoint analysis.) you can select SCORE as the way to interpret the input data. Alternatively you can use pairwise comparisons between the 16 cards. However, here SPSS will not be a big help in deciding which cards to compare and how to extract a comprehensive rating from the pariwise comparisons. Here I would advise to either read up on the corresponding literature in this regard or switch to a more suitable program. I think some time ago I saw some R code online that would do this decently enough.
klasse video 👏 wirklich super erklärt
This really help me. Thank you very much
This was helpful. Thankyou
L9(3,4) table desinged by SPSS is different from the table in google image that google image contains 4 colum (1 blank colum), and the ordinary of variables are different. what does the differences of variants order could influence?
If you specify which Google image you refer to I might help answering your question.
@@jensk.perret6794 the second table in this page. there are 4 columns in the table and ordinary of the variant levels are different. and i tried to creat 4 variant 3 level table by SPSS, and there are 13 cards totally. a L9 (3,4) table like the image that i refferenced can do 3 and 4 variant analysis. which table is more reliable?
www.google.com/search?rlz=1C2EJFA_zh-CNUS760US760&biw=1279&bih=650&tbm=isch&sa=1&ei=i6PbW6_zJ-THjwSww6q4Cg&q=orthogonal+table&oq=orthogonal+table&gs_l=img.3..35i39k1j0i19k1j0i8i30i19k1l3j0i5i30i19k1.42400.43413.0.44015.18.7.0.0.0.0.127.693.4j3.7.0....0...1c.1.64.img..16.1.126....0.aeHD-fOgsBw#imgrc=BXXVCSe-rQZbxM:
@@chideleyve533 If you generate an orthogonal design you get the minimal number of cards that are needed to generate a stable output. To increase the stability of the output you can increase the number of cards by using additional hold-out cases. Thus, if you tell SPSS to include a certain number of hold-out cases you will get a larger table.
Hi sir, sometime I want to remove or add additional value and label, and rerun again. Also factor
Thanks a lot.
Hi, how can I analyze the orthogonal array for ANOVA analysis or regression using spss?
If I understand your question correctly you are looking for a way to analyze the data you collected for an orthogonal design you created. In this case you can look up the video on how to run a conjoint analysis in SPSS: ua-cam.com/video/LZ1aMpidn0Y/v-deo.html
If this is not what you are looking for, then please rephrase your question.
Hai Sir, how can i make all the total 9 card into several choice set, for example; that 9 cards has grouped into 3 choice set and each choice set have 3 cards, how can i do it with using SPSS. I am already generate 16 cards using Orthogonal Design SPSS.. but it is to many to be rank by respondents in a one row. (from the most preferred until the less preferred)
I am already setting the minimum card number which are 15 and the holdout number are 5, then SPSS generate 16 Design with 5 Holdout. so, it is to many to be rank in a one row by respondent. can it be divide/ group into several group?
and what is the right method to be use?
I am very appreciate if you can answer my question.
thanks Sir.
Dear Ms. Mizuki,
in the way I understand your question I would either abstain from using ranking and have the interviewee evaluate the different designs on a given scale and then in the context of the conjoint analysis. (You see this in my video on using the conjoint analysis.) you can select SCORE as the way to interpret the input data.
Alternatively you can use pairwise comparisons between the 16 cards. However, here SPSS will not be a big help in deciding which cards to compare and how to extract a comprehensive rating from the pariwise comparisons. Here I would advise to either read up on the corresponding literature in this regard or switch to a more suitable program. I think some time ago I saw some R code online that would do this decently enough.