Buttons that do nothing... An
interesting article from the BBC about buttons and other controls that do absolutely nothing, but make people feel better anyway.
This is not new understanding. Back in the early '80s, a customer for my design consulting practice introduced me to the same idea. The customer was a firm in San Diego that built specialty radios, mostly for the commercial maritime trade. These were sturdy devices that internally were quite advanced systems, but were being used by fisherman, cargo ship crews, and the like – not technically advanced folks, typically. They also didn't have the habit of being gentle with their gear :)
I was contracted to build the first radio receiver they'd ever produced with a digital front panel. The company's own engineers were analog folks, so they hired me to design all the digital parts as well as the interface to the analog side. Most of this was very straightforward. One seemingly simple element was the volume control. Back in those days, controlling volume digitally was a bit of a challenge. One couldn't simply go out and buy something like an 8-bit digital-to-analog controller (DAC) at a reasonable price. So we decided to settle for a 3-bit DAC made from discrete components. This provided 8 volume levels, which the company decided was more than sufficient – most users would simply turn it to the max, and leave it there. I designed in an 8 position rotary switch that provided a click-stop at each of the 8 possible volume levels. It worked great.
The company took our prototype and sent it out on a boat with a real customer to get feedback. Mostly the feedback was very positive, to the delight of the entire team. But ... the customer hated the volume control. He claimed that the volume he wanted was always directly in between two of the click stops. I was dismayed, and foresaw a difficult challenge – one that would likely bust the production cost budget. The founder of the company was unperturbed, though. He told me to put in a switch with no click stops, just smooth continuous motion. To do this, we had to switch to a
gray-coded DAC (to avoid glitches at the switch points), but that was relatively easy and cheap. Then we sent the prototype back out to the same customer for testing, with a couple of other minor changes.
The result? The customer was totally pleased, and reported that the volume control now worked exactly as expected. Several follow-on testers reported the same thing. Note that there was no actual change in function – the
only thing we did, functionally, was to remove the click-stops. The smooth motion of the switch now gave the customers the
illusion of analog control, but the reality was they still had just eight discrete volume levels. Customers were happy, and so was the company that employed me, as the new design actually turned out to be slightly cheaper to build.
I've remembered that incident ever since, and I confess that I've employed the same technique several times since then. It worked every time. I'm sure I've been fooled (and satisfied) by non-functional controls myself, many times. We humans are funny beasties, aren't we?
