Table 3 Mean and standard deviation daily percent overlap of non-wear in wellness study according to different methods of detection – (Day level) (n = 31,175).

From: A standardized workflow for long-term longitudinal actigraphy data processing using one year of continuous actigraphy from the CAN-BIND Wellness Monitoring Study

Missing

5,536

   

% Agreement Between All Methods of Detection (excluding Troiano)

82.35 ± 28.19

   

% Agreement Between All Methods of Detection (Including Troiano)

79.32 ± 27.71

   

% Agreement Between All Algorithms (Troiano, Choi, van Hees)

91.55 ± 14.96

   
 

Wear sensor

Choi algorithm

Troiano algorithm

Van Hees algorithm

Wear Sensor

1

   

Choi Algorithm

85.25 ± 25.61

1

  

Troiano Algorithm

82.94 ± 24.90

95.82 ± 5.67

1

 

Van Hees Algorithm

83.96 ± 27.80

95.12 ± 14.30

92.10 ± 14.68

1

Median daily percent overlap of non-wear in wellness study according to different methods of detection

% Agreement Between All Methods of Detection (excluding Troiano)

94.72

   

% Agreement Between All Methods of Detection (Including Troiano)

90.42

   

% Agreement Between All Algorithms (Troiano, Choi, van Hees)

95.42

   
 

Wear sensor

Choi algorithm

Troiano algorithm

Van Hees algorithm

Wear Sensor

1

   

Choi Algorithm

96.88

1

  

Troiano Algorithm

93.04

98.61

1

 

Van Hees Algorithm

96.53

100

95.63

1

Back to article page