Whether I should use mixed model for my data?

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

Dear all,
I am writing to get your advice on my thesis data analysis.

I am analyzing what factors affect how much funding crowdfunding projects can attract. In my data, I have 125 projects and 10 pairs of these projects are created by the same creator (in total, 115 unique creators). So 10 creators in my data created 2 projects.
All my IVs are at project level though, i.e., duration of the campaign, # of funders, etc., even the tenure of creators is at project level since they will be different for every project. DV is at project level as well, i.e., how much money each project can attract.
Do you see a reason I should control creator effect? I ran a null in Spss with mixed model. I put creatorID in "subjects", and DV at the DV. Is this the right way to run null model in SPSS? If yes, the ICC calculated from covariance estimate is 75%, which I have a hard time to comprehend. 92% of creators have 1 projects, and 8% creators (10) have 10 projects, it is hard to believe that grouping has such a big effect.

Also if I need to run a mixed model to control for creator effect, do I just include creatorID as "subject" or "random effects" in SPSS mixed model?
Thanks for any tip.

Best,

Li

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：mixed model Whether Should Mixed model