“As moral judgments [differ] under certainty and uncertainty, this raises concerns about generalizing empirical results from commonly investigated moral dilemmas, such as the “trolley” cases, in which everything is certain.”  Nadine Fleischhut

Fat Man version of the Trolley Problem

"A trolley is hurtling down a track towards five people. You are on a bridge under which it will pass, and you can stop it by putting something very heavy in front of it. As it happens, there is a very fat man next to you – your only way to stop the trolley is to push him over the bridge and onto the track, killing him to save five. Should you proceed?"

Uncertainty is a state of not knowing for sure what is going to happen. Risk is a felt state of uncertainty where some possible outcomes are undesired. Moral decisions are often risky: we’re trying to do what's right but aren’t sure things will turn out as intended. To act or not is the question. The answer is partly a leap of faith and partly a cool calculation. At least cool calculation in terms of a few hundred milliseconds of brain activity, which we may experience as a instantaneous decision. 

Converting neural impulses to questions, the calculation would look something like this:

How big is the moral good? How big is the moral bad? How likely is the moral good if I do or refrain from doing x? How likely is the moral bad if I do or refrain from doing y? How much risk am I willing to tolerate? Go!

Take for instance the Trolley Problem, which I will boil down to its essence:

Kill one person (Moral Bad), save five people (Moral Good).

Save one person (Moral Good), kill five people (Moral Bad).

But life is never so simple. Each moral choice comes with a probability calculation. And often we don't have time to do a proper calculation because things are moving too fast. So we rely on heuristics and hunches.

If researchers wanted to better capture the moral decision-making process, they should tweak the Trolley Problem to capture the uncertainty that attends moral judgments in general. Here's one possible variation on the Fat Man version:

A trolley is hurtling down a track towards five people. You are on a bridge under which it will pass, and you might be able to stop it by putting something very heavy in front of it. As it happens, there is a very fat man next to you – your only chance to stop the trolley is to push him over the bridge, hopefully onto the track, killing him for sure to possibly save five. Should you proceed? 

Most people would probably say no: too much uncertainty. But then you could tweak the problem further by adding a probability calculation into the story:

...push him over the bridge, hopefully onto the track, killing him for sure to possibly save five. You have a 95% [or 75% or 50% or ...] chance of saving the five people this way.* Should you proceed?

Now that's a study I'd like to see.

--

* Or: "You have a pretty good [or good or excellent] chance of saving the five people."

Reference:

Fleischhut, N. (2013). Moral judgment and decision making under uncertainty, Humboldt-Universität zu Berlin, Mathematisch-Naturwissenschaftliche Fakultät II.