Religious child

gaynorvader · 20-11-2013 03:13PM

Nerdlingr wrote: »

Your'e talking sh*te. You added all of the 23% of the unsure people on the assumption that they all think teachings arent beneficial. If you want I can do the same and add that 23 to the 46 and come up with an overwhelming majority of 69%...see how that works?!!
The facts are, of those who were surveyed, 46% were in favour, 31% against, and 23% unsure.
So the MAJORITY of those surveyed think the teachings are beneficial to society. Once again : 46>31>23.

Surveys are useless as a source of information, they tend to be biased as even how questions are posed can colour the outcomes. (I know you didn't initially link to it)

The majority were not "in favour". More were in favour than weren't. The 23% unsure tells us nothing. It actually makes the survey even more useless as a data point that they've included it there. You could say "of those who either were in favour or not in favour, the majority were in favour...". However, thanks to the 23% unsure, it's not possible to determine where the majority stands in this group without adding more qualifiers.

smacl · 20-11-2013 04:03PM

gaynorvader wrote: »

Surveys are useless as a source of information, they tend to be biased as even how questions are posed can colour the outcomes. (I know you didn't initially link to it)

I think surveys have some value, biased as they often are. I went for the Iona survey as my assumption is that Catholics would treat the source as credible, and biases would tend to be in favour of the Catholic church. The questions asked, whether or not you consider them loaded, are published with the results, though where they were asked is not. At a thousand, the sample size is small, which weakens the result, as does not knowing what part of the population the sample was drawn from.

What surprised me about the survey was that it was published at all, as regardless how you look at it, it shows a Catholic church in chronic decline. What it suggests to me is that the subject deserves more in-depth investigation, particularly in the context of our Government making decisions on who runs primary school on the basis on a census figure that states we Irish are 84% Catholic. Interestingly in the survey, only 69% of the respondents declared as Catholic, which could either be down to small sample size / demographic used, or that household census information is more likely to be filled in by one person for the entire family.

The majority were not "in favour". More were in favour than weren't. The 23% unsure tells us nothing. It actually makes the survey even more useless as a data point that they've included it there. You could say "of those who either were in favour or not in favour, the majority were in favour...". However, thanks to the 23% unsure, it's not possible to determine where the majority stands in this group without adding more qualifiers.

I actually found the inclusion of don't know / don't care information, along with large groups with an opinion declared as weak valuable. Lack of interest or strong opinion in questions of this nature could also be considered indicative of a drift away from religion as a focus in the lives of most Irish people.

Piliger · 20-11-2013 05:06PM

smacl wrote: »

What surprised me about the survey was that it was published at all, as regardless how you look at it, it shows a Catholic church in chronic decline.

It demonstrates a split in motives between the clergy and the Iona Inst. My guess is that Iona are using it as leverage to criticise the clergy and promote themselves. It fits their holier than thou attitude.

What it suggests to me is that the subject deserves more in-depth investigation, particularly in the context of our Government making decisions on who runs primary school on the basis on a census figure that states we Irish are 84% Catholic. Interestingly in the survey, only 69% of the respondents declared as Catholic, which could either be down to small sample size / demographic used, or that household census information is more likely to be filled in by one person for the entire family

On the contrary I think we don't need any more investigation. The Gov needs to be made to face up to the facts on the ground which are now well established.

I actually found the inclusion of don't know / don't care information, along with large groups with an opinion declared as weak valuable. Lack of interest or strong opinion in questions of this nature could also be considered indicative of a drift away from religion as a focus in the lives of most Irish people.

It is important, imho, to distinguish 'being interesting' with 'being valuable'.

Deeply biased surveys like this have NO VALUE. Worse than that, they seduce the reader into thinking they mean something, even when the reader knows they are biased. The truth is that the numbers are completely and thoroughly corrupted by where the people were questioned, how they were chosen, how the responses were filtered. Hence the results have NO VALUE whatsoever.

In addition .... to be honest ... as an Atheist .. I am losing the will to live even talking about them so much

21-11-2013 01:18AM

smacl wrote: »

At a thousand, the sample size is small, which weakens the result, as does not knowing what part of the population the sample was drawn from.

In fairness, I think you need to rethink some of that, or have a look at some articles around the web about sample sizes, confidence intervals and margins of error.

Not knowing how a sample is drawn could be highly significant, but a sample size of 1,000 is perfectly adequate for most surveys of opinions. Most of the opinion polls in the 2012 U.S. Presidential election had sample sizes of around 1,000 (some had twice that number; others had between 500 and 600). Most of those polls, especially in the closing days of the campaign, suggested a majority for Obama. Likewise, most opinion polls in the 2011 General Election here had sample sizes either side of 1,000. They all suggested big gains for Fine Gael, Labour and Sinn Féin. Not only is a sample size of 1,000 reasonably accurate, it is almost as accurate for a population of 300 million as it is for a population of 60 million, or 4.5 million.

smacl · 21-11-2013 09:44AM

Deleted User wrote: »

In fairness, I think you need to rethink some of that, or have a look at some articles around the web about sample sizes, confidence intervals and margins of error.

A sample of 1000 out of a population of 4.5 million represents just 0.02% of the total. While it might give qualitative results if what you're measuring is known to have a normal distribution across the population, quantitative results are prone to large unknown errors. If the data has an uneven distribution based on geography, you need to know where the sample was drawn from. e.g. Urban vs Rural, outside a shopping center vs outside a church gate. To establish a level of confidence in your measurement, you need to understand the distribution of the variables you're measuring, and the random and systematic errors that they're subject to. The usual method used is to independently repeat the survey with different samples and compare the results, i.e. to investigate error levels and confidence you need redundant independent observations.

There is also the risk that a paid for poll, such as this, might include significant biases to give an outcome that is favorable to the client.

23-11-2013 02:36AM

smacl wrote: »

A sample of 1000 out of a population of 4.5 million represents just 0.02% of the total.

There are a lot of words in there, but the only part that is relevant to what I said is the bit quoted.

You said the sample size weakens the result. That was incorrect. A small sample size does not weaken the result, and decades of statistical analysis proves that a sample size of the order of 1,000 is not at all weak. I'm well aware that 1,000 is a very low percentage of a population of 4.5 million, and I am telling you that this does not matter. In fact, a sample of 1,000 would be regarded as generally adequate to run an opinion survey in the United States, which has a population of about 317 million. You may not believe this, but it is true. There's a Mathematics forum on boards. Go ask there and see what they say. I have no doubt they can explain this better than I can.

You also posted about factors that can weaken survey results. They may very well do, but that's not the point I was addressing at all. What you said was:

smacl wrote: »

At a thousand, the sample size is small, which weakens the result.....

This is not correct.

You can criticise the survey in any way you care to, but criticising it because the sample size is 1,000 is statistically flawed and unscientific.

Sorry for dragging this off topic, but there are plenty of reasons for us to have a pop at theists without needing to resort to bad science while we're at it.

Piliger · 23-11-2013 04:14AM

Deleted User wrote: »

There are a lot of words in there, but the only part that is relevant to what I said is the bit quoted.

Actually no it's not. It's not because smacl also said in the next sentence

smacl wrote:

While it might give qualitative results if what you're measuring is known to have a normal distribution across the population, quantitative results are prone to large unknown errors. If the data has an uneven distribution based on geography, you need to know where the sample was drawn from. e.g. Urban vs Rural, outside a shopping center vs outside a church gate.

And that is wholly correct. The statistical accuracy of the 1,000 is directly dependent on how that sample is selected and ANY bias in that collection will have a distorting affect directly proportionate to that bias.

Further, if one collection point of,say, 50 people were to be taken in a biased way ... such as outside a catholic church ... it would have a FAR bigger proportionate affect on a poll of 1,000 than a poll of 10,000.

So although a sample of 1,000 is adequate and not much less accurate than one of 10,000 - it is far more sensitive to ANY bias in the sampling.

23-11-2013 02:06PM

Piliger wrote: »

Actually no it's not. It's not because smacl also said in the next sentence

And that is wholly correct. The statistical accuracy of the 1,000 is directly dependent on how that sample is selected and ANY bias in that collection will have a distorting affect directly proportionate to that bias.

Further, if one collection point of,say, 50 people were to be taken in a biased way ... such as outside a catholic church ... it would have a FAR bigger proportionate affect on a poll of 1,000 than a poll of 10,000.

So although a sample of 1,000 is adequate and not much less accurate than one of 10,000 - it is far more sensitive to ANY bias in the sampling.

smacl's original point was that a sample size of 1,000 weakened the survey. That's not true, and backsliding, adding to and amending the argument doesn't alter the fact that the original point was incorrect. smacl also tried to argue that the fact that 1,000 is a tiny percentage of Ireland's population is relevant, but it isn't. Any statistician will tell you that.

Crap research is crap research, and a badly selected sample or biased questions are still crap regardless of the sample size. I have no idea where the sample for that survey was found, but if the sample was skewed then the results are skewed as well. However, the argument that a sample of 1,000 is too small of itself just does not stand up to scrutiny. In the example you gave above, the flaw in the research is not the sample size; it is the selection method.

If a sample size of 1,000 was too small, market researchers and pollsters wouldn't use that size of sample. But they do so worldwide every day of the week.

You can try to suggest that you're right all week, and maybe some readers will be naive enough to believe you, but you still won't succeed in overturning the results of decades of statistical research.

Back to the survey in question. I didn't read the survey, and I haven't the remotest interest in what the surveying organisation has to say. You and smacl clearly do, and it seems that you want to challenge the results. What's the basis of your challenge? Have you read the survey methodology and are you challenging the work on the basis of what you read in it? Or are you just assuming the work has been done badly because you don't like the organisation or what it stands for? In other words, is your criticism of the survey based on science or is it based on prejudice?

Piliger · 23-11-2013 04:58PM

Deleted User wrote: »

smacl's original point was that a sample size of 1,000 weakened the survey. That's not true, and backsliding, adding to and amending the argument doesn't alter the fact that the original point was incorrect. smacl also tried to argue that the fact that 1,000 is a tiny percentage of Ireland's population is relevant, but it isn't. Any statistician will tell you that.

Total rubbish.

Turtwig · 23-11-2013 05:01PM

Piliger wrote: »

Total rubbish.

Because. . .

Piliger · 23-11-2013 06:03PM

Jernal wrote: »

Because. . .

Because my post #278 makes it so.

23-11-2013 07:40PM

Jernal wrote: »

Because. . .

It ain't exactly Dawkins, that's for sure.

23-11-2013 07:45PM

Piliger wrote: »

Total rubbish.

You can do the internet equivalent of sticking your fingers in your ears and saying "la la la la la", but it doesn't alter the fact that smacl did say that, and was wrong.

Now, to my question. Have you read the survey methodology and are you challenging the work on the basis of what you read in it? Or are you just assuming the work has been done badly because you don't like the organisation or what it stands for? In other words, is your criticism of the survey based on science or is it based on prejudice?

smacl · 24-11-2013 04:44PM

Deleted User wrote: »

You said the sample size weakens the result. That was incorrect. A small sample size does not weaken the result, and decades of statistical analysis proves that a sample size of the order of 1,000 is not at all weak. I'm well aware that 1,000 is a very low percentage of a population of 4.5 million, and I am telling you that this does not matter. In fact, a sample of 1,000 would be regarded as generally adequate to run an opinion survey in the United States, which has a population of about 317 million. You may not believe this, but it is true. There's a Mathematics forum on boards. Go ask there and see what they say. I have no doubt they can explain this better than I can.

You can criticise the survey in any way you care to, but criticising it because the sample size is 1,000 is statistically flawed and unscientific.

Sorry for dragging this off topic, but there are plenty of reasons for us to have a pop at theists without needing to resort to bad science while we're at it.

To say that a given minimum sample size, e.g. 1000, is adequate to generate valuable quantitative results relating to any variable regardless of population size, distribution of that variable, and measurement error is untrue. I've italicized valuable because a survey can be valuable to different people for different reasons. Typically, in statistics it is a measure of confidence with which we can extrapolate the results derived from our sample to the wider population. If the variables we're looking at have abnormal distribution, e.g. they're subject to clustering, we need to take a much larger sample to gain the same degree of confidence than if they had normal distribution. As per my previous post, to determine a confidence level, you need to prove your observations are independent, and have a sufficient number number of redundant observations to test your assertions.

Your posts say that a sample of 1000 opinions is good enough, which prompts the question, good enough for what exactly? When a question in the survey says when asked to give answer A, B, C, or D to question X, 34% answered B, what exactly does that say about the opinion of the larger population? If you had to bet a substantial amount of money on the outcome of asking the same question to a different 1000 people on the same day, what margin of error would you look for? (1%, 5%, 33%?). What controls would you look for before placing your bet?

Most polls of this nature are also valuable to those that commission them, as they are primarily a PR tool used to throw some numbers at an audience in order to back up a given political or marketing stance with a bit of pseudo science. The veracity of the results in terms of what they say about the wider population is rarely tested, and even if it were the variables being observed (i.e. public opinion) are so dynamic that the tests prove very little. As Hitchens asserts, opinion polls are actually a device for influencing public opinion, and I guess 1000 opinions is perfectly fine for this purpose.

Sarky · 24-11-2013 04:47PM

If the methodology is sound and sample selection is suitably random, 20-30 is often food enough for an accurate prediction.

Could someone post the general gist of the methodology and sample selection so we can get over this argument once and for all?

Piliger · 24-11-2013 05:12PM

smacl wrote: »

To say that a given minimum sample size, e.g. 1000, is adequate to generate valuable quantitative results relating to any variable regardless of population size, distribution of that variable, and measurement error is untrue.

Correct. Plain and simple.

smacl · 24-11-2013 05:29PM

Sarky wrote: »

If the methodology is sound and sample selection is suitably random, 20-30 is often food enough for an accurate prediction.

Could someone post the general gist of the methodology and sample selection so we can get over this argument once and for all?

Survey includes sample size but not how sample was selected or link source data. A statement like '20-30 is often good enough for an accurate prediction' demands rather more context to be meaningful.

I actually don't have a problem with the survey per se. I selected it particularly because it was commissioned by a Catholic organisation, and as such would be unlikely to include systematic bias against the Catholic church. What interested me about it was the degree of dissatisfaction with the church that it indicated, and the huge disparity between nominally Catholic (as per the census) and pro-church Catholic in the poll. My point being that if a survey commissioned by Iona showed up this level of disparity, from an atheist standpoint, further investigation could yield interesting results.

As with any survey where you're interested in what it predicts, you take more observations to see how well those predictions stand up.

Sarky · 24-11-2013 05:52PM

smacl wrote: »

Survey includes sample size but not how sample was selected or link source data.

Then the whole thing is suspect and technically shouldn't really be relied on for valid figures (although see below). Glad that's sorted.

A statement like '20-30 is often good enough for an accurate prediction' demands rather more context to be meaningful.

That's why I qualified it with "as long as the methodology is sound and the sample selection is suitably random..."

As neither of these can be confirmed, the above survey is little better than opinion when it comes down to serious discussion. A sample size of 1,000 does go some way to reducing selective bias; It'd be quite difficult to recruit all 1,000 from one source. Not impossible though, so it's perfectly valid to assume the survey is biased.

Now, all that said however...

I actually don't have a problem with the survey per se. I selected it particularly because it was commissioned by a Catholic organisation, and as such would be unlikely to include systematic bias against the Catholic church. What interested me about it was the degree of dissatisfaction with the church that it indicated, and the huge disparity between nominally Catholic (as per the census) and pro-church Catholic in the poll. My point being that if a survey commissioned by Iona showed up this level of disparity, from an atheist standpoint, further investigation could yield interesting results.

That's the one thing you can probably rely on, any bias is likely in favour of Catholicism. So while one couldn't rely on the survey as an accurate representation of the population, you CAN rely on it to be painting Catholicism in as favourable a light as possible. So you can with a decent degree of confidence say things like "At best, this survey shows the state of Catholicism in Ireland is pretty desperate".

Anything beyond that and we'd really need to have a look at methodology.

As with any survey where you're interested in what it predicts, you take more observations to see how well those predictions stand up.

I'd settle for the details on how the 1,000 were surveyed. If it follows standard good practice for statistics, then the results would be valid. If their methods don't meet the proper standards, we can still tell a great deal from their mistakes, such as what kinds of bias were in sample selection, whether they excluded some data for no good reason, whether the statistical tests themselves were selected to show favourable results...

24-11-2013 07:02PM

smacl wrote: »

To say that a given minimum sample size, e.g. 1000, is adequate to generate valuable quantitative results relating to any variable regardless of population size, distribution of that variable, and measurement error is untrue. I've italicized valuable because a survey can be valuable to different people for different reasons. Typically, in statistics it is a measure of confidence with which we can extrapolate the results derived from our sample to the wider population. If the variables we're looking at have abnormal distribution, e.g. they're subject to clustering, we need to take a much larger sample to gain the same degree of confidence than if they had normal distribution. As per my previous post, to determine a confidence level, you need to prove your observations are independent, and have a sufficient number number of redundant observations to test your assertions.

Your posts say that a sample of 1000 opinions is good enough, which prompts the question, good enough for what exactly? When a question in the survey says when asked to give answer A, B, C, or D to question X, 34% answered B, what exactly does that say about the opinion of the larger population? If you had to bet a substantial amount of money on the outcome of asking the same question to a different 1000 people on the same day, what margin of error would you look for? (1%, 5%, 33%?). What controls would you look for before placing your bet?

Most polls of this nature are also valuable to those that commission them, as they are primarily a PR tool used to throw some numbers at an audience in order to back up a given political or marketing stance with a bit of pseudo science. The veracity of the results in terms of what they say about the wider population is rarely tested, and even if it were the variables being observed (i.e. public opinion) are so dynamic that the tests prove very little. As Hitchens asserts, opinion polls are actually a device for influencing public opinion, and I guess 1000 opinions is perfectly fine for this purpose.

In other words, we agree that a survey based on a sample size of 1,000 will be a sufficiently accurate predictor of the views of the general population from which the sample is drawn. We also agree that a badly-selected sample will produce unreliable results. So what's the problem?

In answer to your quasi-statistical question, the margin of error I would expect from an opinion poll of 1,000 carried out by an independent polling organisation would be of the order of 3 to 3.5 percent, and I would expect that to hold true about 19 out of 20 times. I've already dealt with the controls - crap sampling is crap sampling, and crap questions are crap questions. If the polling body has taken its sample of religious beliefs from outside an Orange hall, or political voting intentions from the parking lot at Democratic Party HQ, they've done a crap survey. But we're not disagreeing about that - what we're disagreeing about is your initial and incorrect assertion that a sample size of 1,000 is too small. It isn't.

Your last paragraph is essentially the kind of thing we see theists write when they're under pressure from the likes of me about positions they've taken in relation to proofs of the existence of their gods. In order for the paragraph to be valid, one first has to accept whatever meaning you've put on the term "polls of this nature". The fact that some people deliberately produce bad surveys, and the fact that you have a problem with the surveying organisation in this particular instance does not alter the fact that 1,000 is a perfectly reasonable sample size for a survey intended to give an accurate prediction of the opinions of the population from whom the sample is taken.

Have you read the survey methodology and are you challenging the work on the basis of what you read in it? Or are you just assuming the work has been done badly because you don't like the organisation or what it stands for? In other words, is your criticism of the survey based on science or is it based on prejudice?

smacl · 25-11-2013 10:55AM

Deleted User wrote: »

In other words, we agree that a survey based on a sample size of 1,000 will be a sufficiently accurate predictor of the views of the general population from which the sample is drawn.

I have said no such thing. When you say sufficiently accurate, sufficiently accurate for what purpose exactly? My original point was that the survey results were interesting as they highlight differences between the number of declared Catholics in the national census versus those who practice in the survey. Further that national policy relating to religious instruction at primary level is tied to the former. While the poll is fine for the purpose of indicating a trend of increased dissatisfaction with the Catholic church, it would be of limited value in an argument to change education policy, for example.

This notion of '1000 observations is enough regardless', without either looking at what the observations refer to, what the data will be used for, or how the results are distributed exhibits a blind faith in statistics not dissimilar to religious faith.

In answer to your quasi-statistical question, the margin of error I would expect from an opinion poll of 1,000 carried out by an independent polling organisation would be of the order of 3 to 3.5 percent, and I would expect that to hold true about 19 out of 20 times. I've already dealt with the controls - crap sampling is crap sampling, and crap questions are crap questions. If the polling body has taken its sample of religious beliefs from outside an Orange hall, or political voting intentions from the parking lot at Democratic Party HQ, they've done a crap survey. But we're not disagreeing about that - what we're disagreeing about is your initial and incorrect assertion that a sample size of 1,000 is too small. It isn't.

So you expect to get a quantitative result that you can extrapolate to a 96.5% degree of confidence, based on the results only from a single sample of 1000 observations to a variable with unknown error or distribution taken from a population of 4.5 million? If you look at analysis of political opinion polls, they'd rarely claim anything like that. They'd also tend to have baseline data, and while they'd often give a good win/lose result, they'd also often fail to do so. Regardless of the sampling method, you have to look at how the results are distributed, as if it is uneven it may be totally unsuitable for extrapolation at any level. (e.g. if I divide my 1000 obs into ten groups of 100 and repeat the analysis, and the results among the smaller groups are significantly different to the principal result, my data does not have normal distribution and will not be reliable for extrapolating information about the wider population).

Your last paragraph is essentially the kind of thing we see theists write when they're under pressure from the likes of me about positions they've taken in relation to proofs of the existence of their gods. In order for the paragraph to be valid, one first has to accept whatever meaning you've put on the term "polls of this nature". The fact that some people deliberately produce bad surveys, and the fact that you have a problem with the surveying organisation in this particular instance does not alter the fact that 1,000 is a perfectly reasonable sample size for a survey intended to give an accurate prediction of the opinions of the population from whom the sample is taken.

The reference regarding use of opinion polls to influence opinion is actually from a noted atheist rather than theist (Hitchens). In terms of generalization you're the one displaying disproportionate faith in the veracity and broader applicability of opinion polls. I tend to be rather more skeptical.

Have you read the survey methodology and are you challenging the work on the basis of what you read in it? Or are you just assuming the work has been done badly because you don't like the organisation or what it stands for? In other words, is your criticism of the survey based on science or is it based on prejudice?

Survey methodology wasn't described, nor was raw data provided, but then that could be simple editorial judgement rather than seeking to obfuscate. As mentioned repeatedly, I selected this survey on the basis that any bias would not favor the atheist position, to avoid prejudice.

Sarky · 25-11-2013 10:56AM

robindch · 25-11-2013 11:59AM

smacl wrote: »

This notion of '1000 observations is enough regardless', without either looking at what the observations refer to, what the data will be used for, or how the results are distributed exhibits a blind faith in statistics not dissimilar to religious faith.

Not really. Given a sampling procedure which selects across the population in a representative fashion, a sample size of 1,000 will produce (with 95% confidence) a result with a margin of error of about 3% (ie, within about 3%) of the actual result you'd get if you sampled everybody in the population:

https://onlinecourses.science.psu.edu/stat100/node/17

smacl · 25-11-2013 12:24PM

robindch wrote: »

Not really. Given a sampling procedure which selects across the population in a representative fashion, a sample size of 1,000 will produce (with 95% confidence) a result with a margin of error of about 3% (ie, within about 3%) of the actual result you'd get if you sampled everybody in the population

Only true if your observations display normal distribution. If you see unusual distribution, e.g. clustering of your observations into distinct groups, trends or spikes, you have to treat your sample differently. While you're trying to take a static snapshot, you could actually be looking at something very dynamic in nature. Taking the case of the Iona survey, it shows a certain percentage level of dissatisfaction in the Catholic church over a thousand people surveyed. This survey will have taken place over a period of time. Say we looked at the observations, and broke them down into groups over them time of the survey, we might note that the level of dissatisfaction changes over the period. In this instance extrapolating the mean value in direct proportion to the larger population is clearly unsafe. We could risk extrapolating based on the trend (albeit with much lower confidence), or we could simply mark the data as unsuitable for this purpose.

Bottom line, working with statistical results without access to the raw data that was used to generate those results is fraught. Where there is any question mark over a data set, taking more independent redundant observations to use as a control set is the easiest and most effective way of verifying confidence levels.

Piliger · 25-11-2013 03:53PM

smacl wrote: »

Only true if your observations display normal distribution. If you see unusual distribution, e.g. clustering of your observations into distinct groups, trends or spikes, you have to treat your sample differently. While you're trying to take a static snapshot, you could actually be looking at something very dynamic in nature. Taking the case of the Iona survey, it shows a certain percentage level of dissatisfaction in the Catholic church over a thousand people surveyed. This survey will have taken place over a period of time.

AND it very well may have been based solely on people at the church gate, or solely people outside community schools.

The nonsense of the claim that accuracy comes from numbers alone is transparent and the necessity of having to even point it out is ridiculous.

26-11-2013 12:00AM

smacl wrote: »

This notion of '1000 observations is enough regardless'....

Ach, at this stage I reckon you're just having me on. It's difficult to see how I can reach any other conclusion. I didn't say "1,000 observations is enough regardless". You said "1,000 observations isn't enough regardless".

You started out by saying something that was in effect an interjection - and as you have now stated it had nothing to do with a statistical critique of the survey. If you want to do that, fire ahead, but if you are just going to attribute something to me because it is an apparent opposite of your interjection, I might as well try to plait the rain rather than take part in an irrational debate about a survey whose results are of no interest to me.

26-11-2013 12:21AM

Piliger wrote: »

The nonsense of the claim that accuracy comes from numbers alone is transparent and the necessity of having to even point it out is ridiculous.

Well said. Now find us anyone in this discussion who has made such claims.

Czarcasm · 26-11-2013 01:36AM

Having read through the last 25 pages of this thread, apart from the fact I'm thinking "What an awful clusterfcuk of a thread! No wonder the OP bailed out early when it went to shìt!", I'm thinking "This isn't about atheism or religion, the problem is staring the OP right in the face" -

Rosina1969 wrote: »

She's been influenced by my parents strong catholic faith.

Rosina1969 wrote: »

My current childminder is a devout catholic and goes to mass every day, so I would think there have been a lot of conversations with her also.

Rosina1969 wrote: »

I think it's like a comfort to her

Rosina1969 wrote: »

Her dad (ex-husband) says we should just say no, you're not doing it now, you can do it when you're older and you understand more about whether it's something you really want to do.

I tried to say to her the other day to wait until she is old enough to go to church herself, and it was a big decision, but she went absolutely mental. Then she started saying she had prayed and prayed that she would be allowed to do it... So if I do say no would it be a lesson that god doesnt answer her prayers?
I'm very confused about what to do.

Your daughter is screaming for your attention OP. When she's not with the childminder she's with her grandparents, and her parents are separated so I imagine she gets very little time with either of you. You have allowed a situation to develop where your child is being influenced by god botherers all round her, then she gets to spend some time with her parents, and what's the most obvious difference between you and the people she regularly interacts with?

(Educate Together schools aren't anti-theist schools, they are in fact so called because they embrace all religions and teach the children about many religions, as part of what they call an Ethical Educational Curriculum).

She's picking up on that one difference that she knows will draw your full attention, and she's maximising the attention she's getting for it by engaging in behaviour that she knows will give you concern for her welfare. Your child isn't really as, erm, "easily influenced" as you think, more to the point in fact it is YOU who is being far too easily influenced by her behaviour. Your daughter is manipulating you and your ex-husband for attention. She's pushing the envelope each time and you're succumbing each time.

My best advice to you is to stop enabling and entertaining her behaviour. Tell her straight up that "No! You will not be baptising her, you will not be confirming her, you will not be entertaining any religious folly!". Let her get mad, let her kick off a stink.

Sure, it'll be hard to watch, she's your daughter and you hate to see her upset. BUT, you have to draw the line somewhere and show her that you are her parent, her primary carer, get your ex on board with the plan and get your grandparents and childminder on board with the plan. They are NOT to be fcuking around with your daughters mental health and well being. She needs support, not having to compete for attention between all of you.

By way of disclosure: I'm Roman Catholic myself, my wife is Atheist, we agreed to bring our child up in the Roman Catholic faith, but since then I have introduced him to many different faiths and belief systems, and indeed explained atheism, science, critical and rational thinking and philosophy, the works of Shakespeare and many classics of English literature, Irish Celtic myths and folklore, Art and Music, History, Geography...

I only mention the above because having read through the thread, some of the posters in here who identify as atheist would actually be an embarrassment to people I know who identify as Atheist, because they are so insecure in their Atheism that they are actually more Anti-theist than simply a lack of belief in a deity!

They seem more concerned with pitting themselves against J C to backslap each other, and when J C was gone they turned on each other and bandied about surveys and statistics in what has to be commended as some of the most fantastical willy waving I've ever seen on Boards, completely forgetting (or ignoring) about the OP's issue.

If this thread was an example of Atheists being rational and logical, well, it read more like a group of immature 10 year olds arguing in the school playground.

Piliger · 26-11-2013 03:40AM

What an appalling piece of amateur psycho babbling nonsense. Blame the parents is all you have to contribute, along with a brutal and ultimately doomed frontal assault on the child.

pauldla · 26-11-2013 03:50AM

Piliger wrote: »

What an appalling piece of amateur psycho babbling nonsense. Blame the parents is all you have to contribute, along with a brutal and ultimately doomed frontal assault on the child.

I quite liked it. A puckish satire of contemporary mores. A droll spoof aimed more at the heart than the head. :pac:

smacl · 26-11-2013 09:30AM

Deleted User wrote: »

Ach, at this stage I reckon you're just having me on. It's difficult to see how I can reach any other conclusion. I didn't say "1,000 observations is enough regardless". You said "1,000 observations isn't enough regardless".

On mature reflection, my point re 1000 obs not being enough for an opinion poll may have been **cough** wrong.

Religious child

Comments