Version

Weighted

Visual Summary

Dist+Subcort Excl

Sample Table
Sub03 Individual R2 Contributions with Distance+Subcortical Exclusion
R2_ed R2_pl R2_co Total
0.0582069 0.0125362 0.0256489 0.0963921
0.0603327 0.0323289 0.0164754 0.1091371
0.0012437 0.0044102 0.0009150 0.0065689
0.0521764 0.0118130 0.0699642 0.1339536
0.0282109 0.0222681 0.0164370 0.0669161
0.0076379 0.0307104 0.0336740 0.0720223
0.0109981 0.0035723 0.0026478 0.0172182
0.0648037 0.0051982 0.0021384 0.0721403
0.0148981 0.0165036 0.0618014 0.0932031
0.0229694 0.0308875 0.0199420 0.0737989
Plot

Dist+Subcort Excl+2SDceil

Sample Table
Sub03 Individual R2 Contributions with Distance+Subcortical Exclusion+2SD ceil
R2_ed R2_pl_2SD R2_co_2SD Total
0.0583187 0.0072004 0.0125237 0.0780428
0.0532253 0.0282768 0.0137331 0.0952351
0.0009224 0.0024866 0.0092722 0.0126813
0.0521638 0.0073939 0.0532127 0.1127705
0.0287627 0.0211401 0.0125461 0.0624489
0.0095261 0.0355451 0.0460438 0.0911150
0.0101260 0.0032751 0.0017405 0.0151415
0.0619713 0.0058854 0.0127016 0.0805583
0.0130481 0.0169751 0.0572742 0.0872973
0.0227211 0.0349849 0.0210238 0.0787297
Plot

Binarized

Individual Plots

Specific Analyses:PL-Based

0.05 is the mean R2 value across various thresholds

Functional Associations with Path Length Comparing Low and High R2 Category

0.01

0.02

0.04

0.05

0.06

0.08
## Error in t.test.default(x = c(0.350026007710802, 0.231482008727155, -0.217682024847822, : not enough 'y' observations
## Error in t.test.default(x = c(0.0272469407805695, 0.0240571261242433, : not enough 'y' observations

## Error in stats::oneway.test(func ~ pl, data = df, var.equal = FALSE): not enough observations
## Error in stats::oneway.test(func ~ pl, data = df, var.equal = FALSE): not enough observations
0.10
## Error in t.test.default(x = c(0.0272469407805695, 0.0240571261242433, : not enough 'y' observations

## Error in stats::oneway.test(func ~ pl, data = df, var.equal = FALSE): not enough observations
0.15
## Error in t.test.default(x = numeric(0), y = numeric(0), p.adjust.method = "holm", : not enough 'x' observations
## Error in t.test.default(x = numeric(0), y = numeric(0), p.adjust.method = "holm", : not enough 'x' observations

## Error in `contrasts<-`(`*tmp*`, value = contr.funs[1 + isOF[nn]]): contrasts can be applied only to factors with 2 or more levels
## Error in `contrasts<-`(`*tmp*`, value = contr.funs[1 + isOF[nn]]): contrasts can be applied only to factors with 2 or more levels
0.20

0.25

0.30

Are the two MLR models (PL_cont, PL_cat) different?

AIC analysis
  • In PL_cat model, PL==1 coded as 0, all other PLs coded as 1
  • I calculated AIC values for each model and then subtracted AIC(model with PL continuous) from AIC(model with PL categorical). If model with PL categorical is a better fit, then AIC_PLcat - AIC_Plcont should be negative (suggesting that AIC_PLcat has a smaller AIC value). Values hover around 0, with slight “leaning” toward the left (negative) for more conservative thresholds.

T-test (R2Total @ 0.01 threshold)

  • The R2 contributions across all ROIs differ when considering the total R2 for an MLR model with PL_continuous versus PL_categorical. This difference was assessed by a paired, two-sided t-test.

  • t = -5.7132, df = 463, p-value = 1.986e-08

  • 95 percent confidence interval: -0.005609400 -0.002738195

  • mean of the differences: -0.004173797 (R2Total_PLcategorical - R2Total_PLcontinuous)

SUMMARY OF ANALYSES

Analytic Procedure Map