Participants

SIC01

Summary Stats

With PL in model

Summary Statistics
Variable N Mean Std. Dev. Min Pctl. 25 Pctl. 75 Max
% variance explained by Euclidean Distance toward FC 423 0.03579 0.03967 0.00012 0.00642 0.05124 0.25727
% variance explained by Path Length toward FC 423 0.01741 0.02181 9e-05 0.00421 0.0208 0.17647
% variance explained by Communicability toward FC 423 0.02165 0.03046 1e-05 0.00256 0.02634 0.20276
Total % variance explained toward FC 423 0.07486 0.06785 0.00086 0.02579 0.10288 0.42991

Without PL in model

Summary Statistics
Variable N Mean Std. Dev. Min Pctl. 25 Pctl. 75 Max
% variance explained by Euclidean Distance toward FC 423 0.03897 0.0453 0 0.00567 0.05409 0.25695
% variance explained by Communicability toward FC 423 0.025 0.03635 2e-05 0.00232 0.03149 0.25656
Total % variance explained toward FC 423 0.06397 0.06541 2e-05 0.01547 0.09321 0.36261

Paired t-tests

co and ed (pl+)

T-test results: co and ed (pl+)  

  • t = 5.7828, df = 422, p-value = 1.431e-08
  • alternative hypothesis: true difference in means is not equal to 0
  • 95 percent confidence interval:0.009220713 0.018716694
  • mean of the differences = 0.0139687

 

Cohens-D results: co and ed (pl+)  

  • ed vs. co: ES=0.28, magnitude small-to-moderate
  • That is, the number of standard deviations between the means of R[^2] values for ED and CO is small-to-moderate
  • Further, the effect (mean of R[^2] values) is bigger for ED than for CO.

co and pl (pl+)

T-test results: co and pl (pl+)  

  • t = -3.0488, df = 422, p-value = 0.002442
  • alternative hypothesis: true difference in means is not equal to 0
  • 95 percent confidence interval:-0.006979699 -0.001507693
  • mean of the differences = -0.004243696

 

Cohens-D results: co and pl (pl+)  

  • pl vs. co: ES=-0.15, magnitude negligible
  • That is, the number of standard deviations between the means of R[^2] values for PL and CO is negligible.
  • Further, the effect (mean of R[^2] values) is bigger for CO than for PL.

co and ed (pl-)

T-test results: co and ed (pl-)  

  • t = 5.7828, df = 422, p-value = 1.431e-08
  • alternative hypothesis: true difference in means is not equal to 0
  • 95 percent confidence interval:0.009220713 0.018716694
  • mean of the differences = 0.0139687

 

Cohens-D results: co and ed (pl-)  

  • ed vs. co: ES=0.28, magnitude small-to-moderate
  • That is, the number of standard deviations between the means of R[^2] values for ED and CO is small-to-moderate.
  • Further, the effect (mean of R[^2] values) is bigger for ED than for CO.

SIC02

Summary Stats

With PL in model

Summary Statistics
Variable N Mean Std. Dev. Min Pctl. 25 Pctl. 75 Max
% variance explained by Euclidean Distance toward FC 420 0.02032 0.02103 5e-05 0.00473 0.0287 0.11
% variance explained by Path Length toward FC 420 0.01714 0.02349 0.00018 0.00454 0.01958 0.17171
% variance explained by Communicability toward FC 420 0.02465 0.02722 1e-04 0.00552 0.03478 0.14573
Total % variance explained toward FC 420 0.06211 0.04894 0.0022 0.02656 0.08447 0.28768

Without PL in model

Summary Statistics
Variable N Mean Std. Dev. Min Pctl. 25 Pctl. 75 Max
% variance explained by Euclidean Distance toward FC 420 0.02066 0.02162 3e-05 0.00453 0.02991 0.11275
% variance explained by Communicability toward FC 420 0.02891 0.03433 0 0.00554 0.03799 0.18339
Total % variance explained toward FC 420 0.04957 0.04345 0.00021 0.01837 0.06848 0.2325

Paired t-tests

co and ed (pl+)

T-test results: co and ed (pl+)  

  • t = -2.7547, df = 419, p-value = 0.006131
  • alternative hypothesis: true difference in means is not equal to 0
  • 95 percent confidence interval: -0.007419167 -0.001240153
  • mean of the differences = -0.00432966

 

Cohens-D results: co and ed (pl+)  

  • ed vs. co: ES=-0.13, magnitude negligible
  • That is, the number of standard deviations between the means of R[^2] values for ED and CO is small.
  • Further, the effect (mean of R[^2] values) is bigger for CO than for ED.

co and pl (pl+)

T-test results: co and pl (pl+)  

  • t = -5.2953, df = 419, p-value = 1.923e-07
  • alternative hypothesis: true difference in means is not equal to 0
  • 95 percent confidence interval:-0.01029289 -0.00472000
  • mean of the differences = -0.007506445

 

Cohens-D results: co and pl (pl+)  

  • pl vs. co: ES=-0.26, magnitude small
  • That is, the number of standard deviations between the means of R[^2] values for PL and CO is small.
  • Further, the effect (mean of R[^2] values) is bigger for CO than for PL.

co and ed (pl-)

T-test results: co and ed (pl-)  

  • t = -4.5146, df = 419, p-value = 8.254e-06
  • alternative hypothesis: true difference in means is not equal to 0
  • 95 percent confidence interval:-0.011850560 -0.004661373
  • mean of the differences = -0.008255966

 

Cohens-D results: co and ed (pl-)  

  • ed vs. co: ES=-0.22, magnitude small
  • That is, the number of standard deviations between the means of R[^2] values for ED and CO is small.
  • Further, the effect (mean of R[^2] values) is bigger for CO than for ED.

SIC03

Summary Stats

With PL in model

Summary Statistics
Variable N Mean Std. Dev. Min Pctl. 25 Pctl. 75 Max
% variance explained by Euclidean Distance toward FC 464 0.02177 0.02372 3e-05 0.00441 0.02921 0.13427
% variance explained by Path Length toward FC 464 0.02399 0.02703 0.00011 0.00536 0.03188 0.15033
% variance explained by Communicability toward FC 464 0.01829 0.02403 1e-04 0.0037 0.02203 0.16486
Total % variance explained toward FC 464 0.06405 0.04796 0.00085 0.02844 0.08702 0.24639

Without PL in model

Summary Statistics
Variable N Mean Std. Dev. Min Pctl. 25 Pctl. 75 Max
% variance explained by Euclidean Distance toward FC 464 0.02301 0.0261 1e-05 0.00347 0.03429 0.17895
% variance explained by Communicability toward FC 464 0.02221 0.0283 1e-05 0.00359 0.03003 0.20344
Total % variance explained toward FC 464 0.04521 0.04104 0.00011 0.01646 0.06039 0.23286

Paired t-tests

co and ed (pl+)

T-test results: co and ed (pl+)  

  • t = 2.4712, df = 463, p-value = 0.01382
  • alternative hypothesis: true difference in means is not equal to 0
  • 95 percent confidence interval: 0.0007129128 0.0062489153
  • mean of the differences = 0.003480914

 

Cohens-D results: co and ed (pl+)  

  • ed vs. co: ES=0.12, magnitude negligible
  • That is, the number of standard deviations between the means of R[^2] values for ED and CO is small.
  • Further, the effect (mean of R[^2] values) is bigger for ED than fo C0.

co and pl (pl+)

T-test results: co and pl (pl+)  

  • t = 3.7859, df = 463, p-value = 0.0001733
  • alternative hypothesis: true difference in means is not equal to 0
  • 95 percent confidence interval:0.002744671 0.008669086
  • mean of the differences = 0.005706879

 

Cohens-D results: co and pl (pl+)  

  • pl vs. co: ES=0.18, magnitude small
  • That is, the number of standard deviations between the means of R[^2] values for PL and CO is small.
  • Further, the effect (mean of R[^2] values) is bigger for PL than for CO.

co and ed (pl-)

T-test results: co and ed (pl-)  

  • t = 0.4805, df = 463, p-value = 0.6311
  • alternative hypothesis: true difference in means is not equal to 0
  • 95 percent confidence interval:-0.002465744 0.004061851
  • mean of the differences = 0.0007980538

 

Cohens-D results: co and ed (pl-)  

  • ed vs. co: ES=-0.0.2, magnitude negligible
  • That is, the number of standard deviations between the means of R[^2] values for ED and CO is small.
  • Further, the effect (mean of R[^2] values) is slightly bigger for CO than for ED.