I admit that I have little knowledge of how many fields work and probably know the least about academia but I still have to ask (or maybe that's why I have to ask) what research does a business undergrad participate in?
For example, make a game where two people are teamed up. You give person A $10 and tell them they can give as much as they want to person B, and that's it. How much does person A share? Ok, now what if you give instructions explaining to person A that person B leaves empty handed if they don't share? What if you say that they're supposed to share 50%? What if 10% of whatever is given to person B is lost as a "tax"? The idea is to figure out ways to get people to act honestly and fairly in the hopes that it can be applied to business.
Undergrads could help find test subjects. They could help design the tests (they generally involve simple software that administers the games). Some could probably analyze the results in the hopes of finding interesting and unexpected takeaways.
As an aside, I believe my first post and this one could sound like I'm trying to be a dick, I'm honestly not. I am wondering and interested. Business majors and related (like your accounting professor father) always seemed to me to be more practical degrees and not really research fields.
Yes, a lot of b-school research crosses fields. I've worked with b-school profs doing work across sociology, economics, law (to a lesser degree), computer science, and mathematics. The difference between research in those fields and b-school research seems to be that the b-school profs had more of an applied bent. The b-school research was more motivated by practical business problems.
Of course, much like programs within a discipline can be radically different from university to university, I imagine there's also a range of research styles across b-schools.
I'm not familiar with accounting research but my impression is that there's quite a lot of overlap with some areas of microeconomics. After all, accounting is either about modelling and prediction or about accurate reporting and analysis if it's not to be book keeping.
Sometimes the cross-disciplinary/applied thing just means that garbage is published by people who aren't familiar with research in another field that covered the same ground decades before. The worst though is when something is researched, disproved and they keep on teaching it in these professional schools for reasons of ideology or politics. Stuff like Howard Gardner's multiple intelligences is still taught in Ed schools and researchers like Linda Gottfredson have to deal with witch hunts because they do work worthy of a real psychology department while in an Ed school. The perils of doing work on on intelligence in a political faculty.
Listening to a coworker of mine:
Marketing, obviously. This is how industry wide 99% of SUVs will never leave suburban blacktop but most SUV commercials have the vehicle speeding up and down dirt trails in a national park. Sometimes similar things in business come from random convergent individual evolution, but more often from the outside. Somebody released a paper on the topic of Americans being germophobes about 5-10 years ago, predictably we had to sit thru a product cycle where anything that can theoretically include triclosan, had to include triclosan (and other anti-bacterials, etc). You may remember a similar fad a decade earlier for citrus oils.
Generalized, generic version of what they'll do on an individual case basis with numerical metrics in private sector. Talk to buddies in the field to get raw randomized data and turn the data into a report from a business perspective if you pay $x bounty for finding a security bug that results in a supply demand curve shaped exactly as such, with this effect on product development times, and this effect on total cost of production / ownership, sales figures, etc, all theoretically "industry wide" or at least "subindustry wide"
We don't have a free market system, more of a centrally controlled system. So there's a basically infinite collection of govt laws, rules, regs, where you can gather the raw data for game theoretic analysis of what the central controllers have decided for us. Is it better, purely economically, to hire illegals to staff fast food restaurants? Identifying all the economic costs is actually pretty hard. Someone out there is probably researching every law/rule/reg out there... costs of obeying vs disobeying, cost to purchase alternative legislation via election funds / lobbying, and repeat each along the short term / long term axis.
For accounting the rules are often "what rule can you purchase" or "what you can get away with" but there's also real world accounting. You need to keep two sets of books, one with how fast does the IRS allow you to depreciate "XYZ" for tax purposes, and another set of books for how fast "XYZ" actually depreciates in the real world. Then you have to live off the real world books or else you can get in a horrible cash crunch if your cashflow depends on selling a used "XYZ" to make payroll this month or as a downpayment on a new "XYZ".
Finally criminals spend a lot of time inventing new crimes. New forms of control fraud, etc. Basically, classic social engineering without computer involvement.