During the study: The personal performance data

This section presents and focuses mainly the descriptive findings regarding the performance data, while the regression models and perceptions are addressed later on.

During the study, a wide variety of data was collected automatically regarding the participant’s daily performance. The personal performance data consists mainly of two types of data, step-related and sleep-related. The step data is an active measure that the participants can actively engage and influence as it is in consciousness, while sleep data is an idle and stagnant activity performed

primarily outside of a conscious state. For example, the step result is a personal performance measurement that the participant usually can actively change by going for an extra walk around the block, getting off the bus a stop earlier or going to the gym. On the other hand, the sleep result is less susceptible to active change.

The participant can opt to go to bed earlier, refrain from caffeine and other similar remedies yet an improved sleep result cannot be consciously and actively pursued in the same manner as the step goal. Beyond this, data on the interaction with the Jawbone UP app was collected, such as how many times the mobile app was checked and whether any additional activities were logged manually by the participants, such as workouts, mood, and food.

The daily step goal for the participants was most commonly 10 000 steps, as recommended by the mobile app. All participants entered and approved the step goal manually in the mobile app upon installing the app on their smartphone.

When doing this, the Jawbone UP app makes a recommendation to set the daily step at 10 000 steps by referencing the National Health Association. The data collection shows that the participants’ step goal ranged between 4000 steps and 13 000 steps, with a mean of 9783 steps as a goal per day. The mode, or the most common goal, was the recommended 10 000 steps, which accounted for 88.6% of the participants. Only 4.5 % of the participants had more than this recommended goal. If a participant chose to have a lower goal, there was a big drop down to 6000 and below, which only constituted 7%. In other words, the ones that lowered the daily step goal chose a considerable decrease from the recommended goal.

Overall, the majority of the participants chose the recommended 10000 steps as the daily step goal and only a handful chose above or below this. When investigated, those that considered themselves to be active were not correlated with a higher step goal. These descriptive statistics indicate that the participant is likely to go with the recommendation provided by the mobile app and this became the standard by which the daily results were compared.

When it came to the results of daily step activity, the participants were active for an average of 1.6 hours per day with a standard deviation of 1.3. The range of steps was originally between a minimum of 112 steps and a maximum of 21 728 steps, with an average of 8869 steps. This suggests that activity levels ranged a great deal between days where the participant was extremely inactive or active a great deal. The great range of the data set suggests that the users do not exhibit the

same patterns of movement every day, but there are variations, which create such outliers. These may be due to the fact that the participant travels or is sick.

However, some participants had continuously low step results because their routine was studying at home, which did not allow for more than a few hundred steps per day to be collected. Another participant had a low step score because of taking local transport and buses, rather than walking or biking. The performance data on steps showed that users have different movement patterns that range from being active to mostly idle on a daily basis.

CATEGORY N Mean SD Min Max STEPS

Step goal 849 9782.82 1248.19 4000 13000 Step result 781 8869 4850.56 112 21728 Active Time (hrs) 800 1.58 1.25 0 10.3 SLEEP

Sleep goal (hrs) 849 7.76 0.41 6.5 8 Sleep result (hrs) 630 7.0 1.74 1 11.6 Deep sleep (hrs) 630 3.15 1.25 0 7.4 Bedtime before 23* 849 0.16 0.36 0 1 ENGAGEMENT

Times app checked 849 1.95 1.95 0 10

Changed goal* 849 0.14 0.14 0 1

SOCIAL

Social connections** 849 0.25 0.25 0 1 Table 15. Descriptive statistics of study 1.

(*binary scale 0=no, 1=yes)

In terms of sleep data, the participant was also prompted by the device to enter a goal as she or he set up the mobile app, similarly to the step goal installation. The in-app recommendation is 8 hours. The sleep goal data indicates a range that varied from 6.5 hours to 8 hours of sleep with a mean of 7.8 hours of sleep and a mode of 8 hours of sleep. There was less range between the participants’ sleep goals than between the different step goals.

The sleep results varied from between 1 hour of sleep to 11.6 hours of sleep.

However, the sleep data is collected differently from the step data by the device.

The sleep data collection is triggered when the participant enters “sleep mode” by activating a button on the device. As soon as “sleep mode” has been entered, the device measures how long it took for the participant to fall asleep, how long the participant slept, and also light versus deep sleep. Therefore, the result of no sleep is only recorded if the participant has forgotten to enter sleep mode prior to actually going to sleep, or ignored the process altogether. This is also representative in the number of logged sleep results (630 observations) compared to the number of sleep goal observations (849). The results show that both the average (M=7.030 hours) and mode (7.0 hours) were close to each other. In terms of bedtime, 84.3% of the observations fell asleep after 23.00. Whilst asleep, the deep sleep lasted an average of 3.2 hours. One participant managed to sleep an impressive 7.4 hours of deep sleep in an 11.6 hour period.

Overall, the female group had a higher mean average both in terms of steps and sleep when compared to the male population. In order to further investigate, independent t-tests were performed to review whether these differences were significant. The female group’s step results were associated with a higher step result (M=9307, p < 0.05) in comparison to the male group, which was associated with a numerically smaller step result (M=8498, p<0.05). In terms of sleep, the female group also got more sleep than the male counterparts. The female group had slept on average M=7.4 (p<0.05), compared to the male group that had numerically smaller sleep of M=6.7 (p<0.05). Thus, there were significant differences for gender between the step and sleep results. The females were also significantly different when it came to deep sleep (M=3.7, p<0.05) compared to the males (M=2.8, p<0.05).