I’ve blogged elsewhere about a little trick that is embedded within the Morality Play interactive activity.
Very quickly, one of the questions asks whether there is a moral obligation to help a person who is in severe need.
You see a charity advertisement in a newspaper about a person in severe need in India/Australia. There is no state welfare available to this person, but you can help them at little cost to yourself. You have good reason to believe that any help you offer will make a difference. Are you morally obliged to help the person?
Half the people undertaking the activity are told that the person lives in India; the other half that the person lives in Australia. They are then asked to state whether they think we are “Strongly Obliged”, “Weakly Obliged” or “Not Obliged” to help the person.
After nearly 1000 responses, this is what the results are showing us.
The thing that has really caught my attention is the results for people who self-identify as Christians and atheists, respectively (more precisely, the atheist group self-identify as having “No Religion”, so they could be agnostics, or perhaps even deists of some sort, but for the sake of convenience, I’m going to call them atheists).
The headline news is that atheists are twice as likely as Christians to think we’re “Not Obliged” to help the person in need in India (currently, 43% as opposed to 21%).
I actually find that quite shocking. But perhaps even more shocking is the fact the atheist group are much less likely to respond that way when asked about the person in Australia. Here (only) 35% think we’re not morally obliged to help. There are two further points here: (1) this gap is four times as large as the average gap across all respondents (and it’s easily statistically significant – I checked!); and (2) if you look at the Christian group, in complete contrast to the atheist group, you find that they are more likely to think we’re not obliged to help the person in Australia.
My first reaction to these figures was to think I had messed up the programming somewhere. But I have double and triple-checked, and I’m almost certain that I haven’t. Plus, I’ve checked the numbers manually (so to speak); and the figures in the charts correctly add up to 100, so I think this really is what the numbers are saying.
My second reaction, of course, was to think about confounding variables and systematic biases. (Note to any stray new atheists reading this: I am fully aware of the dangers of a non-randomised, self-selecting sample, and that it is not possible to generalize these results, but the fact remains that these results are curious, and rather shocking, in and of themselves – we’re not talking about tiny numbers of people here).
So what’s going on? I don’t really know, but if I had to guess, I’d say it’s possible there is some correlation between youth and irreligiosity specific to these activities (because they tend to get picked up by European schools and colleges), and that it might be that young people are less likely to think in terms of moral obligation than older people; it also seems possible that various stripes of moral nihilism might result in non-religious people denying that one is morally obliged to help others (even if they would in fact help others).
But the difference between the atheist response to the India and Australia conditions is… well, harder to explain (and, as I said, it’s a little disturbing). Anybody got any ideas?
Maybe it’s the use of the term
“obliged”.
I’m an atheist, and I generally don’t use the word “obliged” when I speak of morality.
I speak of morality towards others more in terms of compassion, of concern, of being moved, of empathy, of solidarity.
Religious people are more likely to see morality in terms of obligation, I guess.
I might be concerned enough about poverty in India to give money (if I had surplus funds), but I would
never see it as an obligation.
Of course, I’m not denying that I have legal or social-contractual obligations.
Sure, I think that might (partly) explain why the atheist group are less inclined to seem themselves as being morally obliged to help.
But it doesn’t explain why the difference between the way that atheists respond to the India and Australia version is much greater than the average difference.
In other words, atheists are less likely to respond that we are “Not Obliged” to help if the person is living in India than if they are living in Australia, which isn’t true of Christians, and it’s much less true overall (i.e., it remains true overall, but to a much lesser degree).
As I stated above, I not feel morally obliged to help people in either in Australia or in India.
However, one might guess (as you undoubtedly already have done) that those who respond to your tests are often from Australia or being English-speakers, identify with those from Australia and thus, feel more “obliged” to help those in Australia than in India.
If that is the case and if they are consistent in their replies, then, unlike me, those atheists feel moral obligation, but only towards fellow-citizens or fellow-Native-English-speakers.
I hope that my concern for those in Australia or in India would depend on the situation, not
on the nationality (or race?) of the person in need, but undoubtedly, I have my share of unconscious biases, which might influence my choice.
Hi Amos and Jeremy,
I’m with Amos on the first point I want to make. Morally obliged? To whom? I suppose ones self. Anyway, I had to work at understanding the scope of what the expression implied and decided it must be some sort of philosophical given.
Judging from what you said in your post, it seems you had some preconceived idea what the results were going to be. Most everyone who does such studies usually do. About the only study of the sort I ever did formally: I tested the correlation of student’s grade point averages with the evaluations they gave a teacher. I expected a close correlation, but what I got was a “perfect” scatter diagram. I had no idea what to make of it but had to read it as it stood. I don’t think it’s generally convenient or honest or (something else) to create a test when you have a strong bias in favor of one of the possible outcomes.
But, now let’s try to make sense of the result you got. My own bias would lead me to believe the atheists felt freer in giving an honest answer whereas the Christian was expected to have “Christian charity” and acted accordingly. If I (an atheist) were to answer your questionnaire a lot of things would have drawn me to the negative answer though I would feel compassion for the poor persons. I probably would have ended up feeling guilty. Of course, your respondents might have had trouble, as Amos and I did, with the “obliged” part.
The last part, I would say that atheists feel more compassion for people of their own race (lots of assumptions here.) If they’re being honest, I don’t see what there is to be very surprised about. You want to believe that if a person is a rational and flexible thinking person in one area, it must be so in all areas. Obviously, not so.
Gra
Nothing serious really about my comment here but it is interesting to me that you seem to be saying that Christians are more likely to lie.
Bill,
I tried, obviously not hard enough, to avoid giving that impression. I think a lot of people who answer the question posed, after admitting to a certain belief or other, probably want, perhaps even subconsciously, their answers to conform to what they believe in. In the case of a Christian it would be charity. In the case of an atheist there’s nothing along those lines, consciously or not, to muddy up their answer. As I indicated, it’s my own bias idea.
GRA,
I suspected that you didn’t intend to give that impression.
I’m not sure why Atheists would not feel pressures of morality or ethics much the same as Christians do. I wonder if there really is a difference and how that could be measured.
Bill:
I think that atheists simply justify their ethics in different ways than Christians do or that they tell themselves different stories about what they are doing when they do good (or bad) deeds.
For example, if I (an atheist) give a dollar to a homeless person, I’ll tell myself that his or her situation moves me.
A Christian might give exactly the same dollar and tell him or herself that he or she is following Christ’s teaching or something like that.
Amos,
I appreciate your insight.
I think we need to remember that Christianity is voluntary, and sins are forgiven. I don’t think that there is extra pressure on Chritians. If they do something good it is probably just because they want to do it. Christians aren’t required to live in fear.
I would suggest it might be because Christians feel the strongest pull towards charity to the neediest, whereas non-Christians feel the strongest pull towards sharing with those who are most like themselves, which is a perfectly rational thing to do (reciprocal altruism).
Jeremy, your data charts don’t display on my Mac, which I guess is a due to Apple vs. Flash Wars. Would it be possible to put them up as images?
Perhaps the Christians think in terms of a moral obligation to help the poor (simpliciter) so reference to a country such as India pushes their buttons.
Whereas atheists, etc., may think in terms of an obligation to help those who have fallen behind in their own society. Morality is more like a social contract. There may then be a feeling that the English-speaking industrialised countries – UK, Australia, US, Canada, NZ – are for many purposes one big society.
Or maybe there’s some more sinister explanation. I dunno.
Seems to me that there would probably be ways of testing the above – not in a scientifically exact sense, but perhaps in a sense that could make it seem more or less plausible. E.g. identify the countries of Europe that tend to vote for each other in Eurovision. Then ask people from those countries the same questions. But instead of saying “Australia” say one of the countries in the same Eurovision voting block.
It might also be interesting to see what would happen if pretty much the existing demographic were exposed to the question with “Switzerland” or some other country that is industrialised and wealthy, but not so intuitively part of the “same big society” as is Australia (from the point of view of British and American respondents).
I guess a lot of variations like this could be tried – what about France? what about a “white” but seemingly “exotic” country like Belarus? – though I suppose the hassle and expense might be prohibitive.
(This is purely speculative, an attempt to see why I might feel more obliged to help an Australian than an Indian)
It’s not going to be popular, but…
There are two factors:
1. I can relate to the Australian person more than I can relate to the Indian person.
We feel closer to those who are like ourselves. We are more likely to make friends with, surround ourselves with, fall in love with, empathise with, help those who are like ourselves.
As most respondents would most likely recognise more of themselves in the Australian than the Indian, this is likely one of the reasons.
If you think this means that white atheists care less about foreign non-white atheists, you could be right. Others have suggested it’s about a wider ‘community’ that we belong to – it would be interesting to see if the same difference happened with a person living in Australia but of Indian descent. I have no idea how you would construct this experiment (photos to illustrate?)
2. There are more poor people in India compared to Australia. Despite your careful wording indicating that each payment would help, I suspect most people have a subconscious belief that trying to help a poor person in India is less effective than trying to help a poor person in Australia. As the Australian has more in the way of infrastructure, their money goes further; helping an Indian person wouldn’t improve the infrastructure so is the beginning of a much larger task. Once you begin, you must keep going – it’s a much larger project and will demand more of your resources. We don’t want to admit we have an obligation to help poor Indians because that means we have a massive obligation! Poor Australians are fewer in numbers so we can probably manage that without having to give up our smartphones.
Or not. I don’t know and we probably never will.
Thanks guys.
@Gra – I have no idea why you think it’s not honest or convenient or whatever to create a test if you have a clear bias in terms of what you think the outcome will be. Indeed, I think I’d want to argue that creating a test in such circumstances is a mark of intellectual integrity, since it shows that one does not simply assume one’s biases are true. Moreover, there is a sense in which all scientific experiments are precisely testing some preconceived idea or other (hence the concepts of hypothesis, null hypothesis, etc).
@Russell – Actually, it wouldn’t be too much hassle to substitute other countries in for Australia (just a matter of resetting the database and changing the wording of the question).
My hunch is that there’s something to the idea that Australia is seen as being part of “the same big society”, and that this might explain the difference in the way that atheists respond to the India and Australia questions. (Though, having said that, I’m pretty certain that at least some of what’s going on here has to do with confounding variables.)
@Tom – Thanks. I think your 1. is likely part of the story (and, of course, it might also be part of the story for Xians as well, it’s just that it gets swamped by countervailing factors). And yes, your point 2. is almost certainly correct. You have correctly surmised that the question was deliberately worded to discourage people from that way of thinking, but… well, you’re not going to stop people thinking that way just because you assert something to the opposite effect.
Have you checked the IP-adresses for any signs of foul play? Online polls that potentially reveal moral differences between religious or political (or ethnic or geographical) groups are liable to ‘activism’ to skew the results. Though this might be paranoid of me, it would be good practice to try to control for such elements.
I obviously have a strong bias for asking such a question, as I am myself not a Christian (I wonder what the results are for other religious groups?). In addition to this bias, the results also go against my prejudice that Christians on average tend to be further right on the political spectrum than non-religious, and that people on the left side of the political spectrum tend to have wider circles of empathy (caring more for people unlike themselves) than people on the right.
I suspect that many (though not all atheists of course) are motivated by a strong sense of personal freedom, a rebellion from authority.
This may be simply in terms of freedom from worship obligations of an entity that they can not be persuaded (by any rational argument) exists, and/or that the argument of theodicy (answering the question (“why there is evil in the world?”) are sufficiently compelling. This implies that suffering in the world really has no purpose and coming to terms with this results in a freedom from responsibility to others(a key argument in theodicy is that evil exists so that man has a free choice to fulfill God’s will so that man loves his neighbor charitiably).
If this is so there may be a reaction to the survey to reject being morally obliged for the sake of personal freedom from social obligations – including seeking to selflessly support charity.
This is why I personally think it is important for socially minded atheists to embrace a secular humanist perspective. This means giving up some personal freedoms for the sake of building a better society, where freedom from suffering is maximised.
Please excuse typos
@Kristoffer – These activities have multiple checks for “cheating” built into them. I’m not going to say what they are, or how they work (for obvious reasons), but it would take a lot of effort and deviousness for any activist group to distort the results (and even then I’d probably notice something amiss because I tend to monitor the traffic in real-time – well, at least, I have a laptop to my right which is doing so).
@Martin
“This implies that suffering in the world really has no purpose and coming to terms with this results in a freedom from responsibility to others”
If you’re right, then it is disastrous for atheism/secularism. I’m not sure you’re right, but there might be something in the idea that atheism tends to be associated with a sort of radical rejection of all authority (including moral authority). But, of course, that doesn’t explain why atheists are more likely to say they’re not obliged to help somebody in India than somebody in Australia.
Kristoffer,
You ‘obviously have a strong bias’ only because you don’t like the results. You wouldn’t question the methodology if the results suggested atheists were more charitable and less prone to seeing a significant moral difference between helping an Australian and an Indian (to the Australian’s advantage.) That Christians are further to the right, and that those on the left tend to have wider circles of empathy (caring more for people unlike themselves) than people on the right’ is, as you say, mere prejudice. (I say this as an atheist who doesn’t think of himself as being politically on ‘the right’)
-
To self-identify as ‘non-religious’ doesn’t say much about you other than that you reject (or do not affirm) religious belief. It doesn’t suggest increased rationality (just check out the blogosphere), better morality (ditto) or anything else of interest.
Identifying yourself as a ‘Christian’ could mean a number of things – there are a wide diversity of creeds – but mainstream moderate Christians seem to have the idea that they have obligations to help the poor and needy of whatever creed as part of their core set of values.
Atheists can, of course, be just as morally concerned – and indeed freed from certain religious dogma they can be willing to offer help to the poor that actually works (I’m thinking in terms of contraception here) – but there is nothing about being an atheist, as such, that suggests you are a strong humanitarian and not a bigot.
@Jeremy,
I think there is something of credit to the argument that altruism within a kin group is a behavior selected genetically as it promotes survival of the kin group. If so it is not unreasonable for an atheist/theist/whatever to naturally be inclined to charitable acts preferentially to their closest genetic kin groups. This might explain the test skew for a poll of Europeans/Americans towards Australians more so than Indians (given majority racial distributions in those kin groups). If this hypothesis is true then you might find the same relative skew for a poll of Koreans say towards an Asian charity as opposed to say an African. That might be tested.
For Christians who believe that “neighbors” are not to be recognized by racial division this skew might not be as dominant (as per the results). Perhaps because Christians are invited by their faith to “radical altruism” (charity outside of the kin group) as per parables of “The Good Samaritan” and their admonition to “Love their Enemy”.
Crucially, to be fair, the tests should differentiate between “Secular Humanist Atheists” and those who choose to identify with merely with the term “Atheists”. If my speculation is valid you would not see difference in the results of Christians and Secular Humanist Atheists. Indeed Secular Humanist Atheists may be the most charitable group in practice – hard to say without the data!
Jeremy,
Thanks for your response to my comment.
I agree with you that what I wrote “I don’t think it’s generally convenient or honest or (something else) to create a test when you have a strong bias in favor of one of the possible outcomes”, is questionable as written, but I do have an idea in mind; I went to Google to find support for it and struck out. So, without anything to back me up I don’t have a leg to stand on.
The type of thing I was trying to express was the danger of having too much invested in the results of the test which can and does end up in, as an example, throwing outliers from generated data which might be perfectly proper pieces of data. This procedure is used all the time, and when someone is looking for a particular result, overused.
I think I can argue with you (but let’s not) on your statement “a test in such circumstances is a mark of intellectual integrity, since it shows that one does not simply assume one’s biases are true.” History has proved that integrity is often a victim when it comes to statistical testing.
Lest I be charged with having been mealy-mouthed with respect to my poorly constructed comment, I hope you don’t think I was questioning your honesty or integrity; I wasn’t and I don’t.
As a side issue, being more likely to think you’re not obliged to help the person in Australia seems explicable in terms of thinking Australians have obligations and, importantly, the means to look after their own.
@Gra – Ah right. Well you’re absolutely right that bias can play a part in this stuff. So, for example, as I’m sure you know, double-blind techniques are utilised as a mechanism to minimize the possibility of experimenter bias influencing results when dealing with human subjects, etc.
Also, you’ll find a big chunk about experimenter bias here.
http://en.wikipedia.org/wiki/Experimenter%27s_bias
But while it’s true that one always has to look out for these things, it doesn’t follow that one shouldn’t do research if one has strong biases in a particular direction. One just needs to be extra careful, employ proper techniques, rely on peer scrutiny, etc.
@Jim
Yes, your point about Australians being able to look after their own is well taken. It’s why I have that bit about there being no state welfare available. But, as you suggest, it’s entirely possible that people will be thinking in those sorts of terms, anyway (though, having said that, it’s interesting that Christianity seems to be a countervailing factor in that respect).
I would be interested to know if the respondents subscribe to a more consequentialist or deontological world-view. Perhaps it does not per se have to do with being “religious versus non-religious”. Perhaps it has more to do with whether someone feels that there are universal, fixed and prior dictums for good behavior or whether someone believes that good behavior is to be judged solely by the potential and actual overall outcome.
Of course, perhaps such views are implied in “religious versus non-religious”. But having met many deontologists who do not think of themselves as religious and, in contrast, religious people who contextualize and equivocate on their “divine commandments”, such an assumption would not seem wise to me. However, I would assume (perhaps incorrectly) that there are more consequentialist members in the non-religious group and more deontologists in the religious group. If the latter is true, then it would appear that having no religion might cause you to respond differently depending on the nationality of the needy. But it could be that, as has been suggested, the “cause” is a bias about the overall effectiveness of helping one needy person in India versus one needy person in Australia. Perhaps “western” consequentialists assume that helping one needy person in Australia safeguards the nation against decline, but helping one needy person in India will do nothing for improving the nation’s structural problems.
Did the survey include any questions from which such positions could be extrapolated?
The fact that many atheists in Latin America and, I imagine, in the ex-Warsaw block countries as well as various third world countries are Marxists or ex-Marxists does not account for the fact that self-declared atheists are more willing to give charity to the poor in Australia than in India, but it might explain why they are less willing to give alms to the poor than are religious believers.
Many Marxists believe that charity is a palliative that prevents radical social change: instead of giving charity to the poor, you should, according to Marxists, organize them to overthrow the system and institute a more just order of things.
Thus, the Marxists are as concerned about poverty as are those who give alms and as well intentioned towards the poor as the alms-givers, but their vision of how a good person acts towards poverty is very different than that of a bourgeois liberal.
As far as I can see, most online accounts of atheism fail to take into account that in Latin America at least most people become atheists because they are Marxists and that Marxists or ex-Marxists constitute a sizeable part of the population in lots of Latin American countries, especially among the educated sectors.
Re: Australians being able to look after their own… I suppose the points stands whether there were state welfare there or not, its a wealthy place and charity begins at home..
But it leaves unanswered the question of why Christians might think in those terms and not atheists yes.
Christians might be more aware of the practical assistance offered by churches and other voluntary organisations to the local poor. Its still church-related groups that seem to be running soup kitchens.
And then, possibly there is some element of truth to Kristoffer’s association of more ‘Right-wing’ views with Christianity – especially in the States. ‘God helps those who help themselves’. And if you are in dire staits in Australia (or the US) as opposed to India its perhaps more likely it will be viewed as being your own fault?
nb Hello Andreas, nice to see you about j
@Jim – Yes, the point still stands, but as you say it goes against what the data is showing for atheists; namely, that they’re less likely to say that we’re not obliged to help somebody if they’re in Australia (compared to India).
And really it doesn’t show it for Xians either; partly because they’re much less likely than average to respond that there is no moral obligation (for both the India and Australia conditions); and partly because the difference between their responses to the India and Australia conditions, where they’re slightly more likely to say that we not obliged to help the person in Australia, is probably not statistically significant.
I think I have an idea of what’s going on here – at least partly.
A whole load of students from Rockville Centre School District, NYC have spent the day on the Philosophy Experiments web site (presumably because their teacher reckoned s/he wasn’t going to get much work out of them today!).
Anyway, it’s shifted the data in a fairly suggestive way. Basically, the combined effects of their efforts has been to shift the Christian responses in the direction of the atheist responses, which provides evidence that youth rather than religiosity/irreligiosity is the key variable here – or at least a key variable (along with religiosity/irreligiosity).
So my conjecture is:
1. The activity had been picked up by a lot of British/European schools in its early stages. Pupils there are likely to self-identify as having no religion. They tend not to think in terms of moral obligation, but to the extent that they do, they are likely to favour people in Australia over those from India (not to a huge degree, but to some degree).
2. At that stage, there was nothing to isolate their irreligiosity from their youth.
3. The students from the New York school help in this respect, since they’re likely to be predominantly Christian, and they have shifted the Christian group (which presumably up until their arrival had included a much more representative range of ages than the atheist group) in the direction of the atheist group.
4. Therefore, it follows that youth is the confounding variable here. It messed up the data because it correlates with both the independent variable (level of religiosity) and the dependent variable (tendency to think in terms of moral obligation).
Phew. I think that largely lets the atheists off the hook. It’s young people who are to blame!
Jeremy ,
The Web page appearance has changed; the preview button has disappeared, etc. What happened?
I updated the theme. But we have a new Preview button!
Jeremy,
You haven’t been threatended by the Dawkins Foundation have you? I suspect a cover-up.
btw I like what you’ve done with the place (and I don’t usually like change)
Ha! I was tempted to keep quiet about the systematic bias, but… well, the truth must out and all that.
Glad you like the new look. I think it’s a little cleaner…
After the theme update I tried posting a comment for the “Religion and science” post but my comment never appeared. Trying to post it again results in a “this appears to be a duplicate comment” error. Just glitch or has something changed with the commenting process?
Hi – It got tagged as Spam. I have no real idea why – it might have been your use of capital letters. The theme update was just a coincidence!
You don’t give the actual counts, but even if you had 1400 atheists (700 India, 700 Australia), the differences in proportions between the two country conditions are not statistically significant. This is equally true for the country comparison in the Christian group. The differences between Christians and irreligious should be statistically significant. You might consider consulting a local statistician.
[In the R stats package,
chisq.test(cbind(round(700*c(.23,.36, .40)), round(700*c(.22,.42,.36))))
chisq.test(cbind(round(700*c(.30,.42, .27)), round(700*c(.31,.45,.24))))]
You should also consider multiple questions with different wording addressing the same issue, and see if there are marked effects eg “morally obliged to help” versus “will you help”, or move to the designs used by economists to measure utility etc.
@David – So what you’re saying is that when I said above (Dec 20, 5.51pm) that the difference between countries within the Christian group probably wasn’t statistically significant, I was right!?
I’ve done statistical testing on the results. The straight country comparison isn’t statistically significant (though it was earlier on: Type 1 error, though actually a function of a systematic bias in the sample).
The difference between the Christian and irreligious is statistically significant (though again I think there is some systematic bias in the data).
I have consulted a local statistician about the data from these experiments, which is why I ended up programming a Z-test calculator:
http://www.philosophyexperiments.com/statistics/
Maybe I’ll have a go at a Chi-Squared test next!
The term “Z test” can apply to a lot of things, but it looks like yours does test for equality of two proportions, without a continuity correction. Since you have some items with ordinal responses, you will also need to program a trend test (eg Cochran-Armitage test). Alternatively, you could just use existing software: this will allow you to jump straight to correctly modelling covariates such as age of respondent and country of origin.
Your results for this particular item on altruism are in keeping with other published research:
Furrow, J. L., King, P. E., & White, K. (2004). Religion and positive youth development: Identity, meaning, and prosocial concerns. Applied Developmental Science, 8, 17–26.
Kendler, K. S., Gardner, C. O., & Prescott, C. A. (1999). Clarifying the relationship between religiosity and psychiatric illness: The impact of covariates and
the specificity of buffering effects. Twin Research, 2, 137–144.
Koenig LB, McGue M, Krueger RF, Bouchard, TJ. Religiousness, Antisocial Behavior, and Altruism:
Genetic and Environmental Mediation. Journal of Personality 75, 265–290, April 2007
Morgan MM, Goddard HW, Givens SN (1997). Factors that influence willingness to help the homeless. Journal of Social Distress and the Homeless, 6,
45–56.
Morgan, SP (1983). A research note on religion and morality: Are religious people nice people? Social Forces, 61, 683–692.
The Koenig et al paper suggests altruistic behaviour (as they measured it) has contributions from shared family environment (30% of variance), while antisocial behaviour is more consistent with genetic influence (~40%). The correlation between “retrospective religiousness” (hoping to measure childhood effects) and altruism was “modest” (r=0.24).
Hi David
Thanks for this. My stats expert told me that given certain assumptions I could use the Z-test (as programmed) for ordinal data (yes, it tests for equality of two proportions, or at least it’s supposed to – I’m slightly nervous I might have botched the programming!). He wrote a long technical explanation, which I couldn’t really follow, to explain why it was okay. Of course, I’m not in the least bit qualified to say whether he is right (but his credentials are first class).
Thanks for the references. I’ll follow them up.
I’d love to use some existing software, SPSS, in particular, but it’s expensive, and I’m not attached to an institution so I can’t get it on a student or faculty license.
There is a free SPSS project, but given you program, I would learn about R, which offers several web interfaces (R FAQ, Rpad)
Finally, looking at your trolley cars etc, the area of human endeavour that regularly weighs up numbers of deaths as a means to an ends is the military. There are plenty of decisions made using imperfect information there (viz another recent thread). You might consider the “bomb Hiroshima” exercise next, recalling 150000 people died during the invasion of Okinawa…
David
*Many* thanks. I’m downloading R as I type.
The “bomb Hiroshima” is an interesting idea. I’ll have a think about it.
Thanks again.