Skip to main content

Table 3 Reporting probabilities and underreporting

From: How should social mixing be measured: comparing web-based survey and sensor-based methods

Network

Quantity

< 5 min

5 - 15 min

15 - 60 min

1 - 4 h

All individuals

P

2.5%

19.4%

39.2%

78.9%

 

P s

1.2%

8.8%

21.8%

65.2%

 

N 1 ' sm - N 1 sm / N 1 sm

47.7%

40.0%

43.3%

50.0%

Degree > 1

P

3.5%

26.3%

48.1%

85.1%

 

P s

1.5%

11.2%

27.0%

69.0%

 

N 1 ' sm - N 1 sm / N 3 sm

42.4%

35.4%

40.0%

39.2%

Degree > 2

P

4.0%

28.9%

50.8%

85.7%

 

P s

1.6%

12.4%

29.0%

73.0%

 

N 1 ' sm - N 1 sm / N 3 sm

39.9%

34.9%

39.5%

45.0%

Degree > 3

P

4.3%

32.0%

55.8%

84.4%

 

P s

1.8%

13.5%

31.5%

73.5%

 

N 1 ' sm - N 1 sm / N 3 sm

40.8%

33.0%

36.4%

51.1%

Degree > 4

P

4.9%

27.7%

56.3%

88.5%

 

P s

1.9%

14.2%

33.5%

75.0%

 

N 1 ' sm - N 1 sm / N 3 sm

36.9%

43.2%

39.1%

39.0%

Smieszek et al. [38]

P

49.0%

81.0%

89.0%

95.2%

  1. Reporting probabilities of (i) survey reporting, P, (ii) reporting probabilities conditional on mote detection, P s , and (iii) proportion of differences in mote and survey data due to underreporting, N 1 ' sm - N 1 sm / N 3 sm . Percentages for P, P s , and N 1 ' sm - N 1 sm / N 3 sm were calculated for four predefined different contact duration categories and for five different networks (all participating individuals included; only participating individuals with a degree of more than 1, 2, 3, or 4 included), all compared to P values calculated from a previous study [38].