Your data analysis solution
Name is required.
Email address is required.
Invalid email address
Answer is required.
Exceeding max length of 5KB

Incomplete block design

By Robin Alikhani | Jun 11, 2017 03:09PM CEST

Dear Support
According to XLSTAT tutorial of choice based conjoint with hierarchical baye, there are two steps in defining profiles and choice sets.
"The statistical process is separated into 2 steps:

- Fractional factorial designs or D-optimal designs are used to generate the profiles.

- Once the profiles have been generated they are allocated in the comparison groups using incomplete block designs"
in the mentioned tutorial (https://help.xlstat.com/customer/en/portal/articles/2062399-choice-based-conjoint-analysis-with-hierarchical-bayes-cbc-hb-?b_id=9283) number of profiles generated by optimize approach is 12. Therefore, total number of choice set with 3 profile in each is 220. However, XLSTAT only provides one block of questions (20 comparisons) to be sent to all respondent. Is this correct? Can incomplete block design have only one block? Would you please provide a reference in literature. what is the statistical consequence of this design? Please explain in detail

Thanks

Best Answer
By Sébastien | Jun 12, 2017 10:45AM CEST | XLSTAT Agent

Dear Robin,

Thank you for your message.

In CBC, each question is a block, and each profile within the block is a treatment. In complete block designs, all profiles appear in each and every block. Whereas in incomplete block designs, you have a reduced number of profiles within each block. If the design tends to be balanced (which XLSTAT aims at achieving while respecting your constraints), then the design is consistent.

Things are well explained here: http://www.maths.qmul.ac.uk/~rab/DOEbook/doeweb11.pdf

Jean Paul


Up

0

Down

By Sébastien | Jun 12, 2017 10:45AM CEST | XLSTAT Agent

Dear Robin,

Thank you for your message.

In CBC, each question is a block, and each profile within the block is a treatment. In complete block designs, all profiles appear in each and every block. Whereas in incomplete block designs, you have a reduced number of profiles within each block. If the design tends to be balanced (which XLSTAT aims at achieving while respecting your constraints), then the design is consistent.

Things are well explained here: http://www.maths.qmul.ac.uk/~rab/DOEbook/doeweb11.pdf

Jean Paul

Post Your Public Answer

Contact Us

1c26995d494fb3061dd0ae8571ffc0a4@xlstat.desk-mail.com
https://cdn.desk.com/
false
desk
Loading
seconds ago
a minute ago
minutes ago
an hour ago
hours ago
a day ago
days ago
about
false
Invalid characters found
/customer/portal/articles/autocomplete
9283