Disclaimer: I am a big fan of the concept of causality, but not this type. I have concerns on whether these lines of inquiry actually probe causality from a "causal" point of view... I outline a few points below, please change my mind.
(1) In vignette style tasks, you have the usual suspect: language effects. The cheap objection would be "how culturally universal are these patterns"? I won't take the cheap route, but I do point out that there is (I think) a big difference between explicit causal judgements and implicit understanding of causal dynamics. Do you think language, by virtue of being an inherent approximative medium for thoughts, could have distortion effects on individuals' "causal judgements"? In other words, could there be a difference in their "intuitive" causal judgements about X, and their explicit (vocalized, linguistically reformulated) judgements? I believe there could be. Moreover, I believe the very words we use to explain "scenes" might bring connotations upon connotions and play tricks. So, do you think you would find the "abnormal" effect even in non-linguistic causal judgment tasks?
(2) More on delivery. I have some concerns with the vignette style stimuli (in general). Specifically to this study, do you think comprehension of scenes-details could have some explanatory power?
(3) If we translate these verbal scenes into a movie, would you expect similar results?
(4) And what if the movies were shown to "old" children? Keeping the dynamics between "variables" stable but simplifying them, of course.
(5) I have concerns with the potential confounder of "causal ignorance". Lay people (who never read any philosophy, or studied causality) not only have a rough causal understanding, but I do not think that whathever they "causally understand" is a reflection of their inner "causal understanding". There are so many things (e.g., life experience, religious values, cultural norms, a book they read in middle school...) that distorts their "moral" intuitions (for example). In other words, using this set up cannot ( I think) probe into the "human mind's intuitive causal understanding".
(6) I also generally dislike tasks that are explicitly questionaire based (where the participants actually know what the task is). Do you think that the inherent metacognitive aspect of these methods may interfere (e.g. modulating up or down) with their underlying causal intuitions?
(7) More on ignorance/knowledge. How would "mastery of a field" affect the conjunctive/disjunctive differences? In other words, do you expect differences between people who have a solid grasp of the dynamics of scene X (via experience), and those who do not?
I am sure you can dismiss several of these points rather easily, and I am curious how you will break them apart and show my intuitions are wrong. Looking forward.