Table 3

Neural network prediction accuracy, neural network sensitivity analysis and PLS total effects on stickiness

Prediction accuracySensitivity analysis
Cross-validationRMSE trainingRMSE testingPresenceFlowUtilitarian valueHedonic valueSocial normsEngagement
10.1280.1330.2590.0610.1000.1260.1600.293
20.1300.1180.1040.1520.2560.0770.1020.310
30.1300.1150.1520.1490.1270.1240.1150.333
40.1310.1100.1180.1190.1600.1160.1410.346
50.1300.1250.1650.1150.1570.1230.1430.296
60.1280.1380.0750.1680.1590.1330.1130.353
70.1270.1490.0940.1330.2090.1660.1810.217
80.1300.1170.1080.0820.0550.2070.1950.353
90.1250.1660.0650.2240.0980.1830.1040.326
100.1310.1130.1590.1660.0700.1010.1510.353
Mean0.1290.128      
s.d0.0020.017      
Average importance 0.1300.1370.1390.1360.1400.318
Normalised importance (%) 40.85043.02843.76342.68044.170100.000
PLS analysis
Total effects0.2000.2400.3760.2810.3010.703
Normalised importance (%)28.45034.13953.48539.97242.817100.000

Source(s): Table by authors

or Create an Account

Close subscription notice
Close access options