Workplace strategists: stop asking people how many days they'd like to work at home. It's a worthless question. Yet two studies I saw today based their headline findings on it. A
explaining why you should avoid it.

Firstly, people's intentions are about as accurate as a New Year's resolution. People can say that they want to work at home, exercise once a day, and give up sugar. But will they?
Secondly, the number of days someone spends in an office is a weird way to conceptualize and measure how people split their time between their home, their office, and other locations.
Sure, some people may follow a rigid schedule where they work in the office 4 days a week and work at home on Fridays. But for many people, it's more fluid. They might do focused work at home and visit the office for meetings.
Or they might work from an office but leave a little early to pick up their kids, resuming working at home in the evening when their children are asleep. How many days a week are they working at home?
Finally, I suspect there is a desirability bias. If you ask someone how often they'd like to work at home and give them 4 options: 3 of which are 'I want to work from home.' People are probably going to say that they want to work from home.
So asking people 'how many days they'd like to work at home' is giving you garbage data because you have people plotting some hypothetical future behavior using a scale that doesn't really fit what you're measuring (and is potentially biased).
If you want to know whether people will return to the office, don't ask them about their plans, instead look at what people are actually doing in regions where the virus is mostly under control, such as Australia.
And if you have to ask people about their future behavior, simplify the scale a bit ('In the future, where would you prefer to work: [home/an office/another location]') and then follow up with better estimates of time.