Impact of Virtual Coworking on Productivity

Participants are medical research faculty recruited through U.S. medical schools' faculty development offices. After assigning participants based on their availability to one of 8 week-long timeslots, we will randomize recruited participants into two groups: 1) control, and 2) virtual coworking treatment. Those who receive the treatment will be further assigned to small groups of 5-7, and participate in a 3-hour-per-day online writing session for 5 consecutive days. Each day, participants will given the session by sharing their intended goals for the day, and end each day with debrief. There is a 5-minute break at each hour mark. Participants are encouraged to turn their cameras on during the writing session. By employing the intention-to-treat (ITT) approach and assuming some extent of dropout at baseline and during the treatment period, we can assess the overall treatment effect, take-up, and the effectiveness of virtual coworking.

2024-07-15

2024-08-09

Productivity, deep work engagement, innovation, collaboration, job satisfaction, writing behavior.

We administer a baseline survey during the pre-intervention registration phase, and collect information on participants' research domain,
research interests, home institution, demographics, rank, and other psychosocial measures. During the intervention, we aim to collect information on participants' task completion, interactions with other participants, length of participation, etc. Upon completion of the intervention, we again administer an endline survey that collects information on participants' self-reported productivity and other psychosocial measures. As part of the endline survey, participants may share their writing samples with the research team, which allows the research team to conduct semantic analyses. The research team will also collect grants submission data from participants' home institutions, and scrape publicly available grants acceptance, paper publication, conference submissions data. To measure program efficacy, we compare differences between the control and treated groups in productivity, innovation, and other psychosocial measures.

Not available

Randomization done on a computer.

Individual

Yes

8 timeslots

200-300

200 in control, 100 in treatment

Given 60% of participants intend to submit a grant, we assume 40% of the overall sample submit a grant. We assume the grant submission rate to be a 30% for the control group and 50% for the treatment group. Assuming we recruit 250 participants, use a sampling ratio of 4:1, and the standard deviation in the grant submission rate is 0.5 (in the overall sample), the power is 66.1%.