Chronic Sleep Deprivation Among the Poor: A Lab-in-the-field Approach

Systolic, Diastolic. The outcome measurements reported here are those captured after the baseline period (starting on day 9). Each measure is first winsorized at the 5% level and standardized using control group data, to mean 0 and standard deviation 1. The measure is then multiplied by -1 such that higher values represent improvements in health. In order to analyze these outcomes jointly, this measure is next averaged to obtain an overall measure and standardized again. The mean is taken across all participant observations within a group regardless of the time after the baseline period. There are no clinically relevant values and the measure has no max and min.

The investigators measure depression using a scale defined by the Patient Health Questionnaire (PHQ), which is a self-administered version of the PRIME-MD diagnostic instrument for common mental disorders. Each criteria is scored from "0" (not at all, a better outcome) to "3" (nearly every day, a worse outcome). Scores on each question are summed, defining depression severity. The outcome measure was the total summed score captured at endline (day 28). The outcome is thus measured on a scale from 0 to 27, where 0 represents the best outcome and 27 represents the worst outcome.

Illness is measured as the number of days of illness (e.g. cough, fever) in the past week. The outcome measure is the number of days of illness in the past week self-reported at endline (day 28).

Hearts and flowers test -- The outcome used is the payment for the task which is a linear combination of the participant's accuracy (percent correct) and reaction time (measured in milliseconds). Specifically, the payment for this task is determined by the formula: 1.5 * ((percent correct - 0.84) / 0.2 - (reaction time - 526.5) / 95.2) + 15. The outcome measure is the average of the measures collected for each individual after the baseline period (starting on day 9).

Corsi block span - The investigators' measure of memory is the longest correct sequence of blocks lighting in a random order that a participant can reproduce. The range is from a minimum of 0 (worst performance) to a maximum of 9 (best performance). The outcome reported for this task is the participant's payment from the task, which is computed based on their longest span as well as baseline data: 2.29*(maximum span - 5.3) / 1.17 + 13. The outcome measure is the average of the measures collected for each individual after the baseline period (starting on day 9).

Psychomotor Vigilance Task (PVT) - The outcome is payment for the task which was a predetermined linear combination of the inverse reaction time, the number of false positives (response when they participant should not have responded), and the number of lapses (reaction time >500ms). Specifically, the payment was determined by the formula: (inverse reaction time - 3.5)/0.42 - (0.5*(number of false positives-4.7)/6.14) - (0.5*(number of lapses-1.3)/2.58) + 15. Higher payments indicate better performance. Although the theoretical minimum possible score is negative infinity, in practice, the minimum possible payment was censored at zero. The maximum theoretical payment is undefined. The payment reflects overall performance. The outcome measure is the average of the measures collected for each individual after the baseline period (starting on day 9).

The investigators use a simple scale of 0 (no pain) to 10 (the worst pain imaginable) for the participant to rate their pain intensity. The outcome measure was the self-reported pain intensity captured at endline (day 28).

The outcome measure is the number of kilometers participants can cover on a stationary exercise cycle over a period of 30 minutes, as reported on the display of the exercise cycle.The cycle was set to the same level of resistance for all participants.

The investigators use survey questions taken from the commonly used SF-36 Health survey asking participants how much their health has limited them in a certain number of activities. The possible answers range from "they did not limit you at all" (0, the best outcome) to "limited you a lot" (3, the worst outcome). The final scale, which is the sum of the answers, goes from 0 for people who are not limited at all in their daily life by their health to 36 for people who are substantially limited in their daily life by their health. The outcome measure is the 0-36 scale captured at endline (day 28).

The measure of the outcome is based on two survey questions taken from Gallup Survey. In the first question, participants are asked to state a number from 0 to 10, with 0 being a feeling that the participant has the worst possible life while 10 is a feeling that the participant has the best possible life. In the second question, participants are asked to state a number from 0 to 10, with 0 being a feeling that the participant is dissatisfied with life as a whole and 10 being a feeling that the participant is satisfied with life as a whole. The final scale, generated by summing the individual items, is then from 0 for highly dissatisfied people to 20 for highly satisfied people. The outcome measure is the 0-20 scale captured at endline (day 28).

The investigators use a question adapted from a Gallup poll and from the SF-36 survey by RAND that ask participants to rate their happiness on the present day from "Not at all happy" (0, the worst outcome) to "very happy" (4, the best outcome). The outcome measure is the average of the measures collected for each individual after the baseline period (starting on day 9)