Birgit Wallmann-Sperlich followed twelve people out of a conventional office into one built around movement and planting, and asked them again three months and seven months later. Standing time during office hours rose by about 40 minutes a day, at p below 0.01. Nothing else moved. Before the move one person had a height-adjustable desk; afterwards eleven of the twelve did, and of those, five reported using the height adjustment regularly at three months and four at seven, which is 45.5 and 36.4 per cent. Wallmann-Sperlich concludes that the extra standing cannot have come from the desks alone, because most people had stopped using them. It came from the standing meeting table and the counter tables elsewhere in the building. Twelve people is a pilot and the paper calls itself one.
Öhrn tracked a public organisation that split, sending 190 staff to an activity-based flex office and 97 to cell offices, and measured before, at six months and at eighteen. The people heading for the flex office started out enthusiastic: their odds of being satisfied with the office design were 5.32 times the cell office group’s at baseline. At six months that had fallen to 1.12 and at eighteen to 0.86. The cell office group moved the opposite way, to 2.96 and then 4.86. Perceived productivity fell in the flex office and stayed down. Privacy, noise, sitting comfort and posture were all rated worse. General health, cognitive stress and neck, shoulder and back pain showed no difference between the two office types at all. Öhrn separates the staff by what they do: people whose work needs concentration reported lower productivity, people with a lot of teamwork kept rating it high throughout.
Annu Haapakangas put the same questionnaire to 408 people three times, before a move, at three months and at twelve, in four offices that relocated into activity-based workplaces and one that stayed. Satisfaction with communication fell. So did the sense of belonging to a community, and it was still down a year later. Work pace was unaffected. Small negative effects on social support and on quantitative and emotional demands appeared only among the people who had come from private offices, and mostly faded. Haapakangas puts the difficulty of finding a colleague at the top of what needs solving.
Lauren Arundell put accelerometers on 146 employees split between an activity-based workplace and a comparison workplace, measuring before a move and six to nine months after. Satisfaction with the physical environment rose in the activity-based group against the comparison group, at a Cohen’s d of 0.84. Perceived organisational support for being active and the frequency of eating lunch with colleagues also rose significantly. Sedentary time, light activity, moderate-to-vigorous activity, job satisfaction and relationships with colleagues all moved in the right direction without reaching significance, at effect sizes between 0.379 and 0.577. Productivity slipped, at 0.278. The interviews with 21 of the participants were split on whether the work got done better.
Jindo measured the same thing two weeks after a renovation in Tokyo, with 13 workers in the renovated office and 29 in two control offices. Light-intensity activity rose from 130.4 to 150.7 minutes per working hours and sedentary behaviour fell from 346.8 to 321.2. Object detection on video showed where people went: the central aisle and the shared height-adjustable workstations near the entrance and the window, rather than the other spaces.
Bergsten surveyed 699 employees of a Swedish government agency three months before a move, three months after and nine months after, and counted how many preparation sessions each of them had attended. The number of activities someone attended tracked how satisfied they were with knowing how to work in the new office, with the office rules and with the information and support they got. The 44 people who attended all four scored 37 points higher on a scale of 0 to 100 for knowledge than those who attended none. Satisfaction also rose among the people who attended nothing, just not as far.
Öhrn followed 287 people for 18 months and Haapakangas 408 for 12, both with a group that did not move. Bergsten surveyed 699 people about the preparation rather than the office. Arundell measured 146 people with accelerometers. Jindo measured 42 over two weeks. Wallmann-Sperlich followed twelve.