to be surprised that a scientist with a doctorate is religious

[Margaritapracataz]

I assumed she was joking, but no she's a very intelligent woman (double first) but she has deeply religious beliefs.

Aibu to think this is a bit strange and to think less of her professionally?

This.

Is there any such theory? At least within science?

Some ideas or theories if you like, such as the idea of God, are not falsifiable, so are outside of science.

What's important is to know the difference and not mix the two.

Lweji I don't think too differently to you so not sure why you're so keen to put me down. I generally don't respond well to stuff like this so perhaps we can declare a moratorium on the sly digs? Otherwise, I'll spend the day being grumpy I'm not remotely interested in relative scientific credentials, or generating a platform for arguing from authority.

Every time you are testing it against a gene you are testing the hypothesis that protein A turns on that gene
I would frame this differently. When you test whether Protein A turns on Gene [insert letter], you are actually trying to disprove the null hypotheses, not support the proposed hypothesis.

Re: imagined hypotheses. OK, so we could make up any hypotheses with zero evidence or plausibility to support the formulation of the hypothesis. Unlikely to persuade RCUK to give a million quid to research it though.... . Appreciate that's a practical rather than philosophical point though.

Re: data publication. I agree that the bias to positive dara is worrying. Even more worrying is the intellectual snobbery towards journals that publish all data, positive and negative, providing the research is technically good.

not support the proposed hypothesis
This should probably start with 'and', not 'not'.

Hypothesis: The Big Bang started the world.

and yes there is evidence for it. scientists in fact do test, data, and conclusion.

In this case to formulate the hypothesis you don't need any evidence apart from the idea that protein A may turn on a gene.

which you already have ample evidence for. We have ample evidence that proteins have affects on genes, and good (or even bad) guesses to which specific ones are connected has good basis in biology in that we have seen it many times before.

these aren't hypotheses out of thin air.

Sorry, MaidOfStars. It's somewhat easy to get too engrossed and late night doesn't help. You are right about the null hypothesis. It actually came to me during the shower this morning. :)

But, about imagined hypotheses, it's more or less what happens if we do a whole genome association study, for example. We test association between a character and thousands of loci, but each test is not based on hard data. We then make a note of which ones show an association that is unlikely to be due to chance. Then we might work on each of those sites specifically to further test if it holds.

It is also a shame that we don't often test imaginary hypotheses, though. And have to wait until someone cocks up in the lab for something unusual to be considered. Which also reinforces the need for blue sky research in addition to translational research. But I digress.

It is also a shame that we don't often test imaginary hypotheses, though. And have to wait until someone cocks up in the lab for something unusual to be considered. Which also reinforces the need for blue sky research in addition to translational research. But I digress.

well, we only have finite resources. and blue sky research is great, but that's different from a shot in the dark.

What would a scientific hypothesis look like in a world in which God actually existed? How do you account for divine intervention in the reaction between protein A and gene B?

So, if we consider that null hypotheses are the ones we are trying to disprove, is the entire argument reframed....?

Hypothesis: God makes A happen.
Null hypothesis: God does not make A happen.
Test: Clever experiments to identify what makes A happen.
Conclusion: High temperature makes A happen (increasing temperature makes A happen, A never happens at low temperatures, lowering temperature makes A stop, etc)
Null hypothesis holds, no evidence to disprove it.

This is slightly different to the premise that contradictory evidence applies to the hypothesis, that it is the hypothesis, not the null, that must be falsifiable? The null hypothesis for any 'God is/does' is 'God isn't/doesn't'. Unless you provide evidence to contradict the null, the null holds - God isn't/doesn't.

Does that follow?

How do you account for divine intervention in the reaction between protein A and gene B?

A long held aphorism in my field is 'Just because something can, doesn't mean it does'. It's a general warning against transferring all the knowledge you generate in the lab into a picture of reality without remembering the potential limits of your experimentation. Of course, it doesn't mean that we can never investigate the 'does' part, and the tools are increasingly available to test 'does'.

To go back to my previous post on null hypotheses (crossed with you), if something both 'can' and 'does', there is no contradiction of the null. Essentially, there is no logical or scientifically-rational reason to evoke God/divine intervention, therefore we don't.

Sorry, that actually doesn't answer your question, I don't think.

Which is fine, because I'm not sure what an answer could be....

If we can measure a system so completely that we understand every principle and mechanism underlying everything we observe about it, the only place for God is 'Well, why?' (as far as I can see). In reality, God gets inserted into any gaps in our knowledge of the system.

I personally think the scientific process is:

Observe phenomenon (eg. lipids spontaneously assemble into vesicles)
Null: This can't be explained by current theory
Hypothesis: This can be explained by current theory

Test by comparing theory to observation

Conclude either theory good so far...or new theory required.

There have been many atheists and many theist posters on this thread who have all agreed that suggesting god in either the Hypothesis or the null would be absurd.

The point at which you would insert god would be when you felt for some reason that not only does the current theory not predict the observation but NO POSSIBLE THEORY COULD. I don't see that happening any time soon and neither have any of the theist commentators.

Atheists have one data element on the existence of god which is that they don't believe it.

Theists have one data element on the existence of god which is that they do believe it.

People are following the scientific method in a very limited way because the evidence is ALL anecdotal...but they are starting with different data sets.

Observe phenomenon (eg. lipids spontaneously assemble into vesicles). Null: This can't be explained by current theory. Hypothesis: This can be explained by current theory

Wrong way round, I think.

Null - fully explicable, given what we know.
Hypothesis - new explanation required, what about this?

Leaving God aside, I'm not sure about these null hypotheses.

Maid of Stars example of 10.04

At no point was the hypothesis "God makes A happen" tested. Showing that changes in temperature make A happen has no relevance. It is an untestable hypothesis as you can't manipulate the variable "presence or absence of God". So there is no null hypothesis to be accepted or rejected.

maid could be...we tend to find there is an assumption that 'biology' is happening...and hence we tend to find ourselves proving the surprise idea that actually you can explain it all with statistical physics...

That's why I had it that way around...the default is that something happens because nature makes it so...or coz evolution init. The new idea is that you don't need to invoke natural wonder or evolution...plain stat mech from 50 odd years ago gets it done....

I'm not sure those H0 and H1 can be formulated in that way. The hypotheses should contain the predictions.

How do you account for divine intervention in the reaction between protein A and gene B?

Null: protein A and gene B remain separated if I pray (time point given)
H1: protein A and gene B get attached if I pray (time point given)

Observation: protein doesn't attach, then we cannot discount null hypothesis and reject H1

Observation: protein does attach, we can discount null hypothesis

Here prayer is used as a proxy measure of God. Because in a normal experiment you can't determine if God was there or not.
But, by definition, god could answer your prayer or not, so it's not a rule and can't actually be tested in this way. It would only require one event of H1 to show god exists, but in science we know there are spurious events, which is why we use statistics, so H1 could happen due to a statistical fluke. And we continue not to know.

In the lipid idea: we see that some lipids assemble
H0: lipids don't assemble in the presence of X
H1: lipids assemble in the presence of X

You test lipids in exactly the same conditions except for the presence of X.
If they don't assemble, you can discount H1. If they do assemble, you discount H0. Again, in a series of repeats.
We are dealing with natural phenomena, which operate in cause - consequence fashion, so we expect it to occur all the time.
But we still wouldn't know if God had decided to make H1 happen all the time in your experiment on a whim.

Imagine that someone has access to your experiments but you can't see what they do to them and you don't get any evidence of their actions. Can you conclude anything regarding whether that person affected the results of your experiments? Unless that person decided to show you what they did or show themselves doing it.

This is why the existence of God can't be tested scientifically.

Not that I really think God is affecting directly our experiments or natural phenomena.
Although, the way some PCRs work or don't work sometimes makes you wonder... :) or just unreliable students

Although, the way some PCRs work or don't work sometimes makes you wonder

My husband (also research scientist) and I always marvel at the idea that the people who developed PCR got a Nobel Prize dinners are a riot of laughs at our house

Wonder technique, my arse.

There's quite a difference between the theory of science - the way it's supposed to work - and the practice of science. Scientists are of course human, so they don't necessarily go about trying to disprove their hypotheses. And they fudge the data, sometimes exaggerate their findings, even lie about their findings. Scientists can be very protective of their own pet theories even when the weight of evidence is against them.

But the real difference about science is that ultimately, the scientific consensus changes when the evidence changes. It may take time, but it happens. It's quite possible that eventually the Big Bang theory (which was hugely contested when it was first proposed) will be displaced by something else.

That doesn't happen with religion, regardless of the weight of evidence. That's the crucial difference. If you look at the claims in, say, The Book of Mormon, most have been conclusively disproved by scientific evidence. That doesn't stop people being Mormons.

Scientists are of course human, so they don't necessarily go about trying to disprove their hypotheses

I disagree. We are required to provide evidence that the results of Experiment XYZ are different to that predicted by the null (which would state that no significant result would be obtained). Every time I do something, measure something, photograph something, I'm trying to find a difference from the null.

I think what you're suggesting is that if I disprove the null, and the data happens to fit my hypothesis, I might be less inclined to seek evidence to contradict that? Sure, that's a reasonable statement in the immediate aftermath of disproving the null. However, in the next experiment, the null will change (to reflect my acquired data) and I will then try to disprove that... And so it goes.

A published piece of work will comprise several sequential null hypotheses.

Yes, you can't publish or gain your degree without trying to disprove your hypothesis. The way we do it is usually by having controls in the experiments or by comparing different models.

Fallen I agree it's nonsensical in real life, but I was just trying to frame the null/hypothesis in such a way that demonstrates that the absence of evidence is no formal justification to maintain a belief that something is real, within the confines of the scientific method (as I think was being suggested).

But we still wouldn't know if God had decided to make H1 happen all the time in your experiment on a whim.

or indeed if he was involved at all.

This is why the existence of God can't be tested scientifically.

Consequently, I claim god isn't a useful concept because he can't be detected at all, nor can any claims about his nature be supported.

It's perfectly unassailable for me to assert that God commanded me to . No one can disprove that God didn't command me to do so. But the reasonable response isn't that your book says god didn't command me to do such a thing. the reasonable response is that I am nuts, and couldn't possibly assert anything about god's nature.

In that case, it's kind of irrelevant why you did an act. The main thing is that you did it and you should be prevented from doing it again or punished for it according to the law.

It's fine that it isn't a useful concept for you.
Other people find it useful.

I don't find it useful for doing science, which is more to the topic of the thread.

Other people find it useful.

I find most other people having the discussion "my god says this" while another group says "my god says this", and neither group having justification for it.

In that case, it's kind of irrelevant why you did an act.

not really. If I say I acted and gave a tangible earthly grievance, people can respond to it.