I find these results interesting:
Trolley problem (five straight ahead, one on side track, turn requires switching, what ought one do?): switch or don't switch?
Accept or lean toward: switch 635 / 931 (68.2%)
Other 225 / 931 (24.1%)
Accept or lean toward: don't switch 71 / 931 (7.6%)
In Fiery Cushman's and my survey of philosophers' attitudes about moral dilemmas, we asked about this case and our results don't line up very well with yours. Here's the prompt:
You are standing by the railroad tracks when you notice an empty boxcar rolling out of control. It is moving so fast that anyone it hits will die. Ahead on the main track are five people. There is one person standing on a side track that doesn't rejoin the main track. If you do nothing, the boxcar will hit the five people on the main track, but not the one person on the side track. If you flip a switch next to you, it will divert the boxcar to the side track where it will hit the one person, and not hit the five people on the main track.
The response scale ran from 1, labeled "extremely morally good", to 7, labeled "extremely morally bad", with 4 labeled "neither good nor bad".
Looking just at the responses when this was the first question in our survey (and thus not subject to order effects -- which we did find, even for philosophers with PhDs and an AOS or AOC in ethics) we found the following:
Rating toward the "good" end of the scale: 57.5% (96/167)
Rating "neither good nor bad": 19.2% (32/167)
Rating toward the "bad" end of the scale: 23.4% (39/167)
"Don't switch" is obviously the big difference here (chi-square p < .001).
One possibility is that the response options just don't map very well onto each other: "Don't switch" isn't quite the same as saying that it would be "morally bad" to switch. Another possibility is that our respondents are demographically different in some key respect. Still another possibility -- what I lean toward -- is that the more vivid the presentation of the case, the more one finds don't / bad responses. Your presentation is minimalist; ours is bare but has at least some narrative structure. Imagine a Hollywood movie with the hero standing by the switch and close-ups of the victims on the tracks. Conjecture: A large minority of viewers, maybe even a majority, will think the hero is doing something wrong if she flips the switch. This might also fit with Greene's findings that people with more activation in emotion-related brain areas tend to give more deontological judgments about moral dilemmas.