Summary Statistics
From Beiwe Wiki
Domain | Variable Name | Description of Variable | Description of What it Measures |
---|---|---|---|
GPS | |||
Time spent at home | Time spent at home over the course of a day. | Home is the most frequently visited significant location for a person between the hours of 8pm and 8am each day over the course of follow up. Measured in minutes. | |
Distance travelled | Total distance travelled over the course of a day. | Measures in meters. | |
Radius of gyration | Average radius that a person travels from their center over the course of a day. | After converting GPS to xy coordinates, a person's centroid is determined by averaging the xy coordinates for each place visited over the course of a day, with weights proportional to the amount of time spent in the location. Next, the radius of gyration is calculated using a weighted average of the distance between each place and the centroid, where weights are measured in the same way. Measured in meters. | |
Maximum diameter | Largest distance between any two places that a person visited in a day. | Measured in meters. | |
Maximum distance from home | Largest distance between any places that a person visited in a day and their home. | Measured in meters. | |
Number of significant locations | Number of significant visited at any point over the course of a day. | Significant locations are determined using K-means clustering on locations that a patient visits over the course of follow up. Set K=K+1 and repeat clustering until two significant locations are within 100 meters of one another. Then use the results from the previous step (K-1) as the total number of significant locations. | |
Average flight length | Average of the length of all flights (straight line movement) that took place over the course of a day. | GPS is converted into a sequence of flights (straight line movement) and pauses (time spent stationary). The average length of flights (in meters) of the day is reported. | |
Standard deviation of flight length | Standard deviation of the length of all flights (straight line movement) that took place over the course of a day. | GPS is converted into a sequence of flights (straight line movement) and pauses (time spent stationary). The standard deviation of flights (in meters) of the day is reported. | |
Average flight duration | Average of the duration of all flights (straight line movement) that took place over the course of a day. | GPS is converted into a sequence of flights (straight line movement) and pauses (time spent stationary). The average duration of flights (in seconds) over the course of a day is reported. | |
Standard deviation of flight duration | Standard deviation of the duration of all flights (straight line movement) that took place over the course of a day. | GPS is converted into a sequence of flights (straight line movement) and pauses (time spent stationary). The standard deviation of duration of flights (in seconds) over the course of a day is reported. | |
Fraction of the day spent stationary | Fraction of the day spent not moving. | GPS is converted into a sequence of flights (straight line movement) and pauses (time spent stationary). The fraction of the day in a pause is reported. | |
Significant location entropy | Entropy measure based on the proportion of time spent at significant locations over the course of a day. | Letting p_i be the proportion of the day spent at significant location I, significant location entropy is calculated as -\sum_{i} p_i\log(p_i), where the sum occurs over all non-zero p_i for that day. | |
Minutes of GPS data missing | Number of minutes of GPS data missing over the course of a day. | Measured in minutes. | |
Physical circadian rhythm | A continuous measurement of routine in the interval [0,1] that scores a day with 0 if there was a complete break from routine and 1 if the person followed the exact same routine as have in every other day of follow up. | For a detailed description of how this measure is calculated, see Canzian and Musolesi's 2015 paper in the Proceedings of the 2015 ACM International Joint Conference on Pervasive and Ubiquitous Computing, titled "Trajectories of depression: unobtrusive monitoring of depressive states by means of smartphone mobility traces analysis." Their procedure was followed using 30-min increments as a bin size. | |
Physical circadian rhythm stratified | A continuous measurement of routine in the interval [0,1] that scores a day with 0 if there was a complete break from routine and 1 if the person followed the exact same routine as have in every other day of follow up. | Calculated in the same way as Physical circadian rhythm, except the procedure is repeated separately for weekends and weekdays. | |
Texts | |||
Number of outgoing texts | The total number of texts sent by the subject. | Measured in counts. | |
Total outgoing text length | The total number of characters texted by the subject. | Measured in characters. | |
Texting out-degree | The total number of unique subjects to which the subject sent texts. | Measured in unique counts. | |
Number of incoming texts | The total number of texts received by the subject. | Measured in counts. | |
Total incoming text length | The total number of characters of text received by the subject. | Measured in characters. | |
Texting in-degree | The total of unique individuals who texted to the subject. | Measured in unique counts. | |
Texting reciprocity | The total number of times a text is received or sent to a unique person and day without response. | The reciprocity of incoming texts is the number of texts sent from a unique individual to the subject, which is next followed up on the same day with at least one more text sent from that same individual without a response from the subject. The reciprocity of outgoing texts is the number of texts received by the subject from a unique individual, which is next followed up on the same day with at least one more text sent by the subject to that same individual without a response from that individual. Texting reciprocity is the sum of these totals. | |
Texting responsiveness | The mean time before sending a text after a text is received. | Measured in hours. | |
Calls | |||
Number of outgoing calls | The total number of outgoing calls initiated by the subject. | Measured in counts. | |
Total outgoing call duration | The total amount of time spent on calls initiated by the subject. | Measured in minutes. | |
Call out-degree | The total number of unique calls initiated by the subject. | Measured in unique counts. | |
Number of incoming calls | The total number of calls received by the subject. | Measured in counts. | |
Total incoming call durations | The total amount of time spent on calls received by the subject. | Measured in minutes. | |
Call in-degree | The total number of unique calls received by the subject. | Measured in unique counts. | |
Call reciprocity | The total number of times a call is received or sent to a unique person and day without response. | The reciprocity of incoming calls is the number of calls sent from a unique individual to the subject, which is next followed up on the same day with at least one more call sent from that same individual without a response from the subject. The reciprocity of outgoing calls is the number of calls received by the subject from a unique individual, which is next followed up on the same day with at least one more call sent by the subject to that same individual without a response from that individual. Call reciprocity is the sum of these totals. | |
Call responsiveness | The mean time before initiating a call after receiving a call. | Measured in hours. | |
Accelerometer | |||
Accelerometer coverage fraction | Fraction of a participant's follow-up time during which accelerometer measurements have been recorded. | Follow-up time is divided into windows of uniform length (e.g. minutes or hours). The fraction of windows with more than a set minimum number of accelerometer measurements is reported. | |
Accelerometer univariate summaries | A single accelerometer observation consists of three measurements, one for each axis of the device. A univariate summary combines these measurements into a single interpretable number. Univariate summaries include the Signal Magnitude Area (SMA), the Vector Magnitude (VM), and the Sum of Amplitudes (SA). | Given an accelerometer observation (x, y, z), the univariate summaries are calculated as SMA = |x| + |y| + |z|, VM = (x^{2} + y^{2} + z^{2})^{1/2}, and SA = x + y + z. The unit for each of these summaries is meters-per-second-squared (m/s^{2}). | |
Accelerometer signal variability | A measure of the variability of consecutive accelerometer measurements during a given time window. | For the given window of accelerometer observations, univariate summaries are computed and the corresponding sample variance is reported. | |
Acceleration direction (Device orientation) | An estimate of the direction (x, y, z) in which the phone is accelerating during a given time window, with respect to the phone's frame of reference. In many circumstances (e.g. when the phone is at rest), this direction is directly upward in the Earth's frame of reference. | For accelerometer observations {(x_{i}, y_{i}, z_{i})} from the given time window, the normalized vector of (median{x_{i}}, median{y_{i}}, median{z_{i}}) is reported. | |
Device proximity classification | A binary classifciation corresponding to whether the phone is "on-person" or "off-person" during a window of time. A phone is considered "on-person" if the user is carrying it (e.g. in a pocket or handbag) or if the user is physically interacting with it (e.g. making a phone call or playing a game) at any point during the window; otherwise the phone is considered "off-person." | The given window of observations is classified as "on-person" when the accelerometer signal variability exceeds the variability that is expected from sensor noise. The window is classified as "off-person" when the signal variability may be attributed to sensor noise alone. | |
Power State or OS | |||
Total screen events | The total number of times the screen has been turned on during a given observation window. | A record of times when the phone screen has been turned on or off is obtained from the phone's operating system. A "screen event" occurs when the screen is turned on and subsequently turned off; the number of such events during the given window is reported. | |
Total unlock events | The total number of times the phone has been unlocked during a given observation window. | A record of times when the phone has been unlocked is obtained from the phone's operating system; the number of such events during the given window is reported | |
Total power events | The total number of discreet periods that the phone has been connected to a power supply during a given observation window. | A record of times when the phone has been connected or disconnected from a power source is obtained from the phone's operating system. A "power event" occurs when the phone is connected and subsequently disconntected; the number of such events during the given window is reported. | |
Multiple Domains | |||
Sleep onset time | A longitudinal model computes the distribution of sleep onset timing per day according to a parametric model. This model is then combined with daily activity data to compute the most likely time for that day's sleeping onset. | ||
Awake onset time | A longitudinal model computes the distribution of awaking onset timing per day according to a parametric model. This model is then combined with daily activity data to compute the most likely time for that day's waking onset. | ||
Sleep duration | The total amount of time estimated to be asleep. | The sleep awake time subtracted by the sleep onset time. |