within-subject blocking within a between-subject design

[ytgu]

Hello!

I wonder how I could counterbalance the order of "test_noise" and "test_accent" for a between-subject design? Subjects should be randomly assigned to Group 1, 2, or 3, but they will complete both noise test and accent test, and I wish to counterbalance the order of these 2 test blocks. Here's my code:


"block_sequence": [
                    "General_Instruction", "Subject_ID", "Sound_check",
                    {"Group1":["c_practice", "c_exposure_1", "c_exposure_2", "c_exposure_3", "c_test_noise", "c_test_accent", "c_baseline"],
                    "Group2":["n_practice", "n_exposure_1", "n_exposure_2", "n_exposure_3", "n_test_noise", "n_test_accent", "n_baseline"],
                    "Group3":["a_practice", "a_exposure_1", "a_exposure_2", "a_exposure_3", "a_test_noise", "a_test_accent", "a_baseline"]}
                    "Survey"] 


Thanks!

[Ting]

Hello! Perhaps one way you can do is to create subgroups of those 3 main groups. For example, you can have

Group1.1, Group1.2, Group1.3
Group2.1, Group2.2, Group2.3
Group3.1, Group3.2, Group3.3

(functionally they are 9 parallel groups, but semantically there are two levels)

Then, between the main groups, you can keep the C vs N vs A distinction. But within the groups (e.g., for Group1.1-Group1.3), you can create more blocks that semi-manually counterbalance the orders of those test blocks. For example, Groups 1.x will have the order A B C, Group 2.x B A C, and then Group 3.x C A B. That way you'll achieve between group counterbalance as well.

Let us know if this works!

[ytgu]

Hello! Perhaps one way you can do is to create subgroups of those 3 main groups. For example, you can have

Group1.1, Group1.2, Group1.3
Group2.1, Group2.2, Group2.3
Group3.1, Group3.2, Group3.3

(functionally they are 9 parallel groups, but semantically there are two levels)

Then, between the main groups, you can keep the C vs N vs A distinction. But within the groups (e.g., for Group1.1-Group1.3), you can create more blocks that semi-manually counterbalance the orders of those test blocks. For example, Groups 1.x will have the order A B C, Group 2.x B A C, and then Group 3.x C A B. That way you'll achieve between group counterbalance as well.

Let us know if this works!

Hi Ting!

Thank you for your suggestion! Yes I think this works well