Skip to content

Hypothesis Driven

Brian Pho edited this page Apr 28, 2021 · 7 revisions

Linear Regression

Here are the results for running a linear regression on the Yeo functional connectivity.

WISC_FSIQ
FPN: y = -5.46x + 102.78, r = -0.04, p = 0.3975
DMN: y = -15.09x + 104.13, r = -0.09, p = 0.0247
FPN+DMN: y = -13.05x + 102.95, r = -0.08, p = 0.0413
---
WISC_BD_Scaled
FPN: y = -1.93x + 10.43, r = -0.06, p = 0.1360
DMN: y = -3.78x + 10.60, r = -0.12, p = 0.0051
FPN+DMN: y = -4.37x + 10.46, r = -0.14, p = 0.0007
---
WISC_MR_Scaled
FPN: y = -0.01x + 10.02, r = -0.00, p = 0.9923
DMN: y = -2.37x + 10.49, r = -0.08, p = 0.0673
FPN+DMN: y = -2.03x + 10.31, r = -0.07, p = 0.1000
---
WISC_DS_Scaled
FPN: y = -0.31x + 9.98, r = -0.01, p = 0.7933
DMN: y = -2.35x + 10.35, r = -0.08, p = 0.0597
FPN+DMN: y = -1.63x + 10.11, r = -0.06, p = 0.1700
---
WISC_Coding_Scaled
FPN: y = 0.96x + 8.50, r = 0.03, p = 0.4279
DMN: y = 1.27x + 8.54, r = 0.04, p = 0.3155
FPN+DMN: y = 0.69x + 8.70, r = 0.02, p = 0.5651
---
WISC_Vocab_Scaled
FPN: y = -1.03x + 11.44, r = -0.03, p = 0.4318
DMN: y = -1.92x + 11.51, r = -0.06, p = 0.1584
FPN+DMN: y = -1.68x + 11.37, r = -0.05, p = 0.1960
---
WISC_FW_Scaled
FPN: y = -0.05x + 10.53, r = -0.00, p = 0.9654
DMN: y = -1.01x + 10.72, r = -0.03, p = 0.4131
FPN+DMN: y = -0.36x + 10.56, r = -0.01, p = 0.7627
---
WISC_VP_Scaled
FPN: y = -1.63x + 11.32, r = -0.05, p = 0.1913
DMN: y = -2.39x + 11.31, r = -0.08, p = 0.0652
FPN+DMN: y = -2.80x + 11.22, r = -0.09, p = 0.0231
---
WISC_PS_Scaled
FPN: y = 0.35x + 9.85, r = 0.01, p = 0.7782
DMN: y = -1.08x + 10.18, r = -0.03, p = 0.4039
FPN+DMN: y = -1.05x + 10.11, r = -0.04, p = 0.3940
---
WISC_SS_Scaled
FPN: y = 0.74x + 9.55, r = 0.03, p = 0.5253
DMN: y = 1.08x + 9.56, r = 0.04, p = 0.3735
FPN+DMN: y = 1.99x + 9.49, r = 0.07, p = 0.0832

And here are the significant correlations

WISC_FSIQ
DMN: y = -15.09x + 104.13, r = -0.09, p = 0.0247
FPN+DMN: y = -13.05x + 102.95, r = -0.08, p = 0.0413
---
WISC_BD_Scaled
DMN: y = -3.78x + 10.60, r = -0.12, p = 0.0051
FPN+DMN: y = -4.37x + 10.46, r = -0.14, p = 0.0007
---
WISC_VP_Scaled
FPN+DMN: y = -2.80x + 11.22, r = -0.09, p = 0.0231

Corrected using Bonferroni, where p = 0.05 / 10 = 0.005, leaves only one significant correlation: the one between BD_Scaled and FPN+DMN.

PLS

Here are the PLS results on 1000 iterations using 1 component.

WISC_FSIQ
Network: FPN
Train r^2: 0.0018283400914731493
Test r^2: -0.012116608253015063
---
Network: DMN
Train r^2: 0.009530242445772524
Test r^2: -0.002943105810554066
---
Network: FPN+DMN
Train r^2: 0.00845289598981106
Test r^2: -0.0035719535560264555
------------
WISC_BD_Scaled
Network: FPN
Train r^2: 0.0039090465478576215
Test r^2: -0.008552438738750312
---
Network: DMN
Train r^2: 0.014111109736510386
Test r^2: 0.0016658657080742699
---
Network: FPN+DMN
Train r^2: 0.021537190789571278
Test r^2: 0.009959707483367984
------------
WISC_MR_Scaled
Network: FPN
Train r^2: 0.0007358163380068505
Test r^2: -0.01242492047624511
---
Network: DMN
Train r^2: 0.00684541438783071
Test r^2: -0.00594487383277944
---
Network: FPN+DMN
Train r^2: 0.005790974179628912
Test r^2: -0.0071480596153403264
------------
WISC_DS_Scaled
Network: FPN
Train r^2: 0.0006613353045727055
Test r^2: -0.012048192555814945
---
Network: DMN
Train r^2: 0.007259321451551676
Test r^2: -0.005589042582520377
---
Network: FPN+DMN
Train r^2: 0.004092184550387628
Test r^2: -0.008316293500310911
------------
WISC_Coding_Scaled
Network: FPN
Train r^2: 0.0017045462670687345
Test r^2: -0.011351475906252453
---
Network: DMN
Train r^2: 0.0023772947354348773
Test r^2: -0.01032302500838807
---
Network: FPN+DMN
Train r^2: 0.0011658822552129977
Test r^2: -0.011459036028220945
------------
WISC_Vocab_Scaled
Network: FPN
Train r^2: 0.0018466476279057247
Test r^2: -0.011064137410537973
---
Network: DMN
Train r^2: 0.004430710647347619
Test r^2: -0.008936714519234442
---
Network: FPN+DMN
Train r^2: 0.0037824806228821998
Test r^2: -0.00840010528987866
------------
WISC_FW_Scaled
Network: FPN
Train r^2: 0.0007627161856192478
Test r^2: -0.012450146988884636
---
Network: DMN
Train r^2: 0.002151681321455725
Test r^2: -0.010662026576855112
---
Network: FPN+DMN
Train r^2: 0.0009515702293192052
Test r^2: -0.01139945174857129
------------
WISC_VP_Scaled
Network: FPN
Train r^2: 0.0034320722076159483
Test r^2: -0.010026685226548673
---
Network: DMN
Train r^2: 0.006833276464983547
Test r^2: -0.006223226518687049
---
Network: FPN+DMN
Train r^2: 0.010526502785213616
Test r^2: -0.002508416765924538
------------
WISC_PS_Scaled
Network: FPN
Train r^2: 0.0009837850767897127
Test r^2: -0.012339955331962003
---
Network: DMN
Train r^2: 0.0020690029416008983
Test r^2: -0.011676217774373313
---
Network: FPN+DMN
Train r^2: 0.0022194587118563535
Test r^2: -0.011698833607880654
------------
WISC_SS_Scaled
Network: FPN
Train r^2: 0.0009850363410559078
Test r^2: -0.011445889484246727
---
Network: DMN
Train r^2: 0.002041563722286701
Test r^2: -0.011294517511259698
---
Network: FPN+DMN
Train r^2: 0.005699493003347219
Test r^2: -0.006876009100127447