I feel like we discussed the importance of selecting criteria in the Experiences page, but I can't recall if that was for recommendations or something else.
I think option A is the way to go. By changing the Control traffic allocation to 50%, we can test both Criteria side-by-side within a single Recommendations activity.
I feel pretty confident about this one. I just need to make sure I understand the difference between the data subject rights mentioned and apply them to the specific scenario. I think I can nail this question.
Okay, I've got this. Fractional factorial designs sacrifice some information by confounding main effects with interactions. That's the key disadvantage compared to a full factorial design. I'll select that option.
B is the answer, no doubt. An A/B test is the way to go. I wonder if the product team will also be testing the product recommendations against a 'random recommendations' control group. That would be the real test!
Gearldine
5 months agoMy
5 months agoMarlon
5 months agoAltha
5 months agoGabriele
5 months agoElina
6 months agoVicki
6 months agoMila
6 months agoLevi
6 months agoCarmelina
6 months agoTy
6 months agoAileen
6 months agoHermila
6 months agoElise
6 months agoNobuko
6 months agoCarmelina
6 months agoPaola
11 months agoDominga
10 months agoAlesia
10 months agoJerlene
10 months agoTwana
11 months agoWynell
11 months agoBrendan
10 months agoTimothy
10 months agoChauncey
11 months agoFranchesca
12 months agoMarisha
10 months agoMaurine
10 months agoTony
11 months agoGlen
11 months agoVirgie
1 year agoPrecious
10 months agoQuentin
10 months agoWava
11 months agoYuriko
12 months agoJamal
1 year agoTenesha
1 year agoCiara
1 year ago