Skip to main content
. 2021 Jul 13;23(7):e26063. doi: 10.2196/26063

Table 3.

Discriminant validity of the measurement model: Fornell-Larcker criterion and heterotrait-monotrait ratio.

Variables BIa PEb EEc SId FCe HMf PVg HAh MOi EDj GAk PAl Age GENm EXPn
BI .856 o .879 .646 .414 .623 .604 .473 .795 .423 .218 .241 N/Ap N/A N/A N/A
PE .875 .835 .651 .464 .594 .694 .405 .747 .635 .368 .378 N/A N/A N/A N/A
EE .637 .648 .823 .181 .785 .435 .614 .341 .321 .147 .179 N/A N/A N/A N/A
SI .407 .455 .168 .911 .135 .536 .057 .616 .366 .366 .375 N/A N/A N/A N/A
FC .584 .561 .871 .090 .733 .394 .678 .281 .278 .146 .178 N/A N/A N/A N/A
HM .607 .693 .446 .517 .363 .881 .254 .650 .515 .458 .571 N/A N/A N/A N/A
PV .467 .412 .619 .046 .645 .266 .873 .181 .199 .077 .097 N/A N/A N/A N/A
HA .592 .569 .180 .590 .091 .536 .027 .811 .470 .316 .366 N/A N/A N/A N/A
MO .423 .630 .319 .356 .253 .519 .203 .404 .806 .683 .712 N/A N/A N/A N/A
ED .222 .365 .148 .364 .125 .451 .078 .303 .680 .861 .632 N/A N/A N/A N/A
GA .243 .366 .188 .346 .156 .549 .107 .339 .706 .637 .794 N/A N/A N/A N/A
PA .133 .130 .073 .032 .060 .176 .067 .079 .046 .104 .036 N/A N/A N/A N/A
Age .038 .026 .003 −.036 .053 −.034 .084 .001 .055 −.033 −.011 −.035 N/A N/A N/A
GEN .019 .064 .118 −.092 .058 −.038 .041 −.016 .157 .063 .096 −.057 .061 N/A N/A
EXP .095 .043 .159 −.140 .196 .009 .179 −.099 −.040 −.068 −.061 .084 .051 −.011 N/A

aBI: behavioral intentions to use the fitness app.

bPE: performance expectancy.

cEE: effort expectancy.

dSI: social influence.

eFC: facilitating conditions.

fHM: hedonic motivation.

gPV: price value.

hHA: habit.

iMO: motivation-related app features.

jED: education-related app features.

kGA: gamification-related app features.

lPA: intentions of being physically active.

mGEN: gender.

nEXP: user experience with fitness apps.

oTerms in italics along the diagonal are square roots of average variance extracted. Below the diagonal, the lower left metrics test the discriminant validity according to the Fornell-Larcker criterion. Discriminant validity is fulfilled if the square roots of the average variance extracted are larger than the relevant interconstruct correlations. Furthermore, above the diagonal, the upper right metrics refer to the heterotrait-monotrait ratio, where <0.85 or <0.90 indicates good discriminant validity.

pN/A: not applicable.