Skip to main content
. 2020 Jan 10;20:5. doi: 10.1186/s12874-019-0893-7

Table 3.

Validity analysis of the SAYCARE sedentary behavior questionnaire

Children (N=93) Q1 Accelerometry data Bland-Altman Analysis rho kd
Median (25th–75th percentile) Median (25th–75th percentile) Biasa 95%LOAb Trendc
 Week days (min/day) 66.1 (41.0 to 96.0) 522.4 (491.7 to 547.8) -459.5 ± 100.9 -661.4 to -257.6 - 0.57** - 0.10 0.14
 Weekend days (min/day) 200.0 (135.0 to 285.0) 546.6 (502.4 to 598.2) -332.6 ± 138.5 -609.6 to -55.6 0.20 0.40** 0.04
 Total days (min/day) 110.5 (74.0-154.3) 530.4 (497.2 to 560.2) -420.2 ± 100.3 -620.9 to -219.5 -0.45** 0.07 0.03
Adolescents (N=94) Q1 Accelerometry data Bland-Altman Analysis rho kd
Median (25th–75th percentile) Median (25th–75th percentile) Biasa 95%LOAb Trendc
 Week days (min/day) 102.0 (63.4 to 165.9) 570.3 (545.6 to 599.7) -449.3 ± 87.0 -623.3 to -275.3 - 0.16 -0.26 -0.11
 Weekend days (min/day) 180.0 (96.4 to 260.0) 599.2 (568.9 to 627.6) -399.7 ± 105.0 -609.6 to -189.7 0.66** 0.21 0.36**
 Total days (min/day) 131.1 (97.1 to 178.2) 582.2 (554.3 to 609.1) -435.1 ± 66.3 -566.6 to -302.6 0.08 0.06 0.05

Values are median (25th–75th percentile). Q1 Questionnaire first application, rho Spearman correlation coefficient

Moderate (or above) values of spearman correlation (rho ≥ 0.30) and kappa agreement (k ≥ 0.40) are in bold

aBias: average difference between methods (Q1 and accelerometer)

bLOA: Limits of agreement calculated using Bland-Altman test

cTrend, Pearson’s correlation coefficients between the absolute value of the difference versus the average of the two variables (Q1 vs accelerometer). Whether r>0 and p< 0.05, there is heteroscedasticity between the variables

d weighted (quadratic) Cohen’s kappa-coefficient for quartiles comparison

*p ≤.05

**p ≤ .01