Two items appear below:

1        Survey of Beliefs
2        De-Mystifying Statistics
 
 

SURVEY OF BELIEFS

(Investigator 50, 1996 September)


In a survey of paranormal beliefs the most popular beliefs included belief in life after death, heaven, miracles, ghosts and astrology.

INVESTIGATOR surveyed 88 people. They were asked to tick yes, maybe, or no to each of 37 beliefs.

The results are shown below with the number who answered each question marked on the survey form. Most beliefs have less than 88 answers because most participants left some questions unanswered.

The percentage "yes" for each belief is calculated and beliefs ranked in popularity.

The sample was too small to be truly representative of all Australian adults. Nor was it a random sample but rather a haphazard sample. Bias is suggested by the high proportion of university educated and the low proportion of church-goers.

A proper survey of Australian adults would require about 4,000 participants chosen randomly.

Nevertheless, the results show some agreement with larger surveys. A survey of 1,000 Australians published in the Sydney Morning Herald (July 1988) gave the following percentages of believers:

 

Mental Telepathy    51%
Life After Death      48%
UFOs                       42%
Astrology                 32%
Faith Healing           29%
Reincarnation          24%
Palmistry                  14%
Tarot                           8%


Thorough and consistent skeptics would have answered "no" on the Investigator form to every question except "Evolution".

Eight people answered "yes" to thirty or more items!

The highest scorer, Sandra Geach, said, "I like the idea there is a spiritual side to people. And that gives me comfort. I feel more comfortable with death."

Ms Geach claimed to have had accurate tarot, astrology and palm readings, felt creepy feelings due to ghosts and ouija boards, and experienced "communication of thought" through "emotional vibrations" people emit.

Psychologist Laurie Eddie, speaking informally, attributed most paranormal beliefs to combinations of: peer pressure, defects in vision/perception, outright fraud, selective memory, non-critical acceptance of what's common in society, emotional satisfactions such as reassurance, and attaching significance to chance concurrence of unrelated events.

As an example of the last Mr Eddie stated 5 billion people average five to ten dreams each per night. Some will dream of a plane or car crash. If a crash occurs they might announce a "prophetic dream." The real explanation is that worldwide there are thousands of car crashes every day and some will coincide with a dream about a crash.

Ms Geach's response to skeptics is, "They'd have to prove to me that these things are impossible."

The following shows how many answered what on the survey form:

What Is your age?
Under 20     11
20-29          14
Over 30       50

Education level reached?
Primary School     --
High School        52
University            30

Do you go to Church?
Hardly ever     53
Sometimes     14
Often              12


Popularity
Rank
DO YOU BELIEVE IN YES MAYBE NO
11 Angels 32 = 36% 23 33
8 Astrology 33 = 40% 16 34
29 Atlantis 14 = 18% 23 40
20 Aura Reading 20 = 26% 25 33
6 The Bible 33 = 40% 29 20
10 Clairvoyance 30 = 37% 21 30
2 Creation 42 = 52% 14 25
28 Crystal Power 17 = 21% 19 46
16  Water Divining 23 = 29% 23 32
14  Dreams That Foretell Your Future 26 = 32% 24 31
1 Evolution 55 = 69% 13 12
15 Exorcism 24 = 30% 24 32
34 Fairies 11 = 14% 12 56
18 Flying Saucers 22 = 27% 26 33
19 Fortune Telling 22 = 27% 22 38
7 Ghosts 33 = 40% 20 29
23 Good Luck Charms 20 = 24% 18 45
4 Heaven 39 = 48% 20 22
31  Levitation 13 = 17% 17 45
3 Life After Death 42 51% 22 19
35 Loch Ness Monster 10 = 12% 25 48
25 Lucky Numbers 18 = 23% 19 43
13 Mind Reading 27 33% 26 29
5 Miracles 38 = 46% 14 30
17 Numerology 22 = 28% 14 43
27 Ouija Boards 17 = 22% 17 45
22  Palm Reading 21 = 25% 20 43
12 Psychics 29 = 35% 20 33
26 Reincarnation 18 = 22% 21 42
33 Pyramid Power 12 = 15% 18 50
32 Seances 12 = 15% 20 46
24 Tarot Cards 19 = 24% 17 45
30 Tea Leaf Reading 14 = 18% 15 51
36 Unlucky 13 5 = 6% 9 69
37 Winning Lottery Systems 3 = 4% 17 63
9 Faith Healing 31 = 39% 17 32
21 Witchcraft 21 = 25% 24 39

 
 


DE-MYSTIFYING STATISTICS

Bob Potter

(Investigator 53, 1997 March)


Investigator, September 1996, carried a useful little contribution: Survey of Beliefs. Unfortunately, the writer appears to share the general 'misunderstandings' that are frequently found in the area of inferential statistics, and also fails to comment on some of the very interesting findings from his mini-research project.

First, the theoretical confusion. He tells us that "the sample was too small to be truly representative..." and that "a proper survey of Australian adults would require about 4,000 participants chosen randomly". It is just not possible to make statements of this kind, in the abstract.

The idea behind statistical inference is that if the researcher can obtain a 'representative' sample, he can use the data from that sample to make inferences about the population as a whole. There is a deep irony built into this argument in that
 

(a) we are only able to be confident that our sample is truly 'representative' of the population if we already have the relevant details about the population, hence

(b) if we were, therefore, confident that our sample was truly 'representative', we wouldn't want it anyway (there is no point inferring what one already knows!)


Leaving that point and moving on to sample size.

Inferential statistics is a battle-ground of ideas (as is every field of science), but all statisticians would agree that it is quite meaningless to talk of adequate sample sizes in the abstract. The required size will depend upon the likeness of the members of the population being investigated (for example, if I were researching the importance of the name of Jehovah God among Jehovah's Witnesses, a sample size of one would be quite adequate!) and will also depend upon the kind of hypothesis I am investigating.

(For those who have done some statistics at school, a 'one-tailed' test requires a smaller sample to achieve equal power — if I predict that Witnesses will be less well educated than Quakers, I will need a smaller sample than if I had merely said there will be a difference in educational levels between Witnesses and Quakers).

It is a great tragedy that today's 'person in the street' tends to have a rigid fear of statistics. People feel it is 'too mathematical'. They make a tremendous mistake, because it is impossible to understand even the very basis of modern science without a clear understanding of what we are talking about when we speak of statistical 'probability'.

I have met scores of people who have claimed to be interested in 'hard scientific facts' rather than the 'probabilities' talked about by social scientists. The person who argues in this fashion only displays their ignorance of the manner in which modern science goes about its business.

To re-enforce and illustrate my point, I suggest a glance at the current Penguin Dictionary of Science under the article, Structure of Atom:
 

"…the planetary electrons of the atom were to be thought of as moving in well defined orbits about the nucleus… In the more modern wave mechanics, the electrons are regarded as having a dual wave particle existence which is expressed mathematically ... the precise position of the electron...is...a probability that a particular planetary electron, visualized as a particle, may be found at a particular point."


Having said all that, let's return to the interesting little Survey of Beliefs. It is unfortunate that the reader is not told how the sample was selected. Was the questionnaire just handed out to friends and relatives, or to people walking along King William Street, or individuals sitting in the pub?

There is no reason why the sample should be 'representative' of Australians as a whole, for it to be interesting or important. If the information as to how and where the data was collected was given, the findings would be just as meaningful, in their own specific context, as would findings arising from a national survey. However, this omission is largely ameliorated by the inclusion of information regarding age and education.

What I find particularly interesting is that the three 'beliefs' endorsed by the majority of participants were 'Creation', 'Evolution' and 'Life After Death'.

Already one suspects that a problem with the design of the questionnaire is that the respondents are apparently not interpreting these terms consistently. For example, I find the first two 'beliefs' to be in contradiction.

Again, interestingly, the next strong endorsements go to 'Astrology', 'the Bible', 'Ghosts', 'Heaven' and 'Miracles' — again there is a disparity between these 'beliefs'. (Of the scores of biologists with whom I am acquainted, working in the field of evolution — today evolution is a theory of biology — I know not a single one who could keep a straight face at the mention of 'ghosts' or 'miracles' as those words would he defined in 'religious' terminology.)

Most importantly, the fact that the sample was not 'representative' (in any definition of the term), is totally irrelevant to the more important finding that this 'majority view', in favour of 'life after death', comes from a sample of 'better educated' individuals, nearly a third of whom are university educated.

This fact alone, which I personally, find most disturbing, is justification enough for the continuance of the Investigator through another fifty issues!


Religion; the Supernatural; the Paranormal; the Skeptics.
Over 1700 articles on this website:

http://users.adam.com.au/bstett/

http://ed5015.tripod.com/