X-Message-Number: 30343
From: David Stodolsky <>
Subject: Re: 1998-2007 Survey Comparison 
Date: Tue, 22 Jan 2008 10:25:41 +0100
References: <>

On 12 Dec 2007, at 04:58, Kennita Watson wrote:

> I only worked with OpenOffice Calc and my scanty statistics
> knowledge, so I wasn't able to get p-values for differences between
> males and females; I made do with deciding that a difference in
> values of 0.3 or more might be interesting, if not statistically
> significant.  Values with slashes are AVERAGE/STDEV.
>
> Number of respondents:
> F: 55

56

> M: 35
>
> I could accomplish much more with my life if it were significantly
> extended.
> F: 2.65/1.31
> M: 2.12/1.2
> (corroborates 1998 diff)

This is on the border of being significant at the p<.05 level.


>
> Dealing with wills, insurance policies, and other legal matters is
> too much trouble to make Cryonics worthwhile.
> F: 2.9/1.17
> M: 3.53/0.98

Significant (p< .018)


>
> Cryonics is a bad idea because it would lead to an overpopulation
> problem.
> F: 2.96/1.29
> M: 3.26/1.22

Not Significant


>
> Cryonically preserving me would be too hard/weird for my family/
> friends to handle.
> F: 2.65/1.32
> M: 2.97/1.22

NS

>
> I would not want to wake up in a future time without my family or
> friends around.
> F: 2.28/1.25
> M: 2.97/1.21
> (corroborates 1998 diff)
>

Highly significant.


> I believe that Cryonics is an exciting idea an intend on looking into
> it further.
> F: 3.16/1.30
> M: 2.69/1.28
> (corroborates 1998 diff)


Not Significant (p=.4)

So, three of six "findings" were significant. The moral to this story  
is that this type of approximation does not do it.


>
> In the 2007 survey, women again perceived cryonics as less affordable
> than men, but this time not by much (2.51 to 2.48).
>
> Men again were more excited about a young, healthy body, but only by
> a 2.65 to 2.84 margin.
>
> This time, women and men had almost identical attitudes about
> cryonics being immoral, 3.35 to 3.39.

This is the question about cryonics being too selfish:
Extending one's life span through Cryonics is unnatural, selfish, and  
immoral.

No conclusion can be drawn from a non-significant result.
So, we can't say they have almost identical attitudes.


>
> Unlike on the previous survey, women seemed slightly more inclined to
> disagree that people in the future will have no interest in reviving
> frozen bodies (3.47 to 3.33).

No conclusion can be drawn from non-significant results. It could also  
result from bad data or a bad question.


>
> While respondents in age ranges 25-34 and >64 did tend to have
> favorable attitudes towards cryonics (to agree more with more of the
> negative statements, and disagree more with more of the positive
> statements), in my analysis those in the 45-54 age range expressed
> those sentiments more strongly, and those under 24 less so.

I can't get significance with an overall test, which included all  
attitude statements and the motivation statement "looking into it" and  
the statement about being "able to revive a human in 100 years",  
either with all age group comparison or comparison to the 45-54 yr.  
old group, which means that the individual tests are suspect. However,  
the following were significant when comparing 45-54 yr. olds and  
others with a t-test:

no loss due to ageing
young again
too young
looking into it further


I repeated the age related tests after log transforming the data,  
since some of the data clearly was non-normal.

Overall test yielded significant multivariate results for age groups,  
but not for the 45-54 age group.

An ordinal logistic fit predicting age groups from responses showed  
the following were significant:

no guarantee
accomplish more
optimistic future
young again
too costly
overpopulation
too weird
too young [sig. dif. comparing 45-54 (Age group 4) vs 25-34 (Age group  
2)]
without friends [sig. dif. comparing <24 (1) vs 35-44 (3)]


It certainly is no surprise that the response 'too young' was  
different with different age groups. The overall trend is that the  
young tend to agree that they are "too young."

Both those under 35 (1+2) and those over 64 (6) tended to agree that  
waking up in the future without friends would be unacceptable. This  
supports the previous findings that middle age is when people come to  
terms with existential questions (these people would be more concerned  
about personal survival than continuing social relationships as  
compared to other age groups).

These overall significant results in comparing age groups support the  
quality of the data as being adequate for analysis.


However, the data set is noisy. For example, the question:

I believe that Cryonics is an exciting idea an (sic) intend on looking  
into it further.

was repeated, separated by three questions, and the first accounted  
for only 43% of the variation in the second (non-parametric r-squared)  
(includes 'Don't know' responses). Cronback's Alpha on the data  
without the Don't Knows was .76. However, after deleting 5 outliers on  
the two question comparison, the value of r-squared is .81 and the  
Alpha is about the same. Two of these saboteurs (who indicated agree  
on one question and disagree on the other) requested membership  
literature from all organizations, out of a total of 6 requests for  
membership info. So, it looks like a data accuracy prize is needed at  
this venue.

There was an error made in constructing the response scale (a central  
'neutral' value, however, this is mitigated by use of an explicit  
"Don't Know"), so there is plenty of room for improvement.

Anyway, I'd like to thank Kennita Watson for the data. And since the  
data allows comparison with the Badger 1998 work, we are heading in  
the right direction.



dss


PS: I once again urge anyone collecting survey data to consult with  
someone who has an earned research degree within the social sciences  
from an accredited university *before* starting to administer it. The  
earlier in the process you get assistance, the more likely the results  
will be useful. Getting survey data is not rocket science, but getting  
any useful results with a survey is!


David Stodolsky    Skype: davidstodolsky

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=30343