This Psi Test
Significance Calculator can be used to test for the statistical
significance of data from various types of psi experiment (e.g.,
ESP and PK) provided that the following conditions are met:
- Each trial is
scored on a hit or miss basis.
- There is the
same likelihood of a hit on every trial.
- All trials are
independent of other trials.
You cannot use this calculator
if an identical target sequence is used to test several
people at the same time. However, you can
combine data from more than one run (one person or
several people) provided that a different target sequence
was generated for each run.
To enter input
data
- Select the type
of testing procedure used.
A closed procedure is when there is a fixed number of
each target per run. An example of this is when a deck of
cards is shuffled and cards are selected one by one
without replacing them in the deck.. An open procedure is
when targets are selected randomly every trial. Examples
of this are coin tosses, dice throws, and selecting a
card from a full deck that is reshuffled after each trial
(i.e., replacing the selected card).
- Select the
probability of obtaining a hit on a single trial.
For example if using standard ESP cards, the probability
is 1 in 5 (because there are five different symbols). If
examining how well someone performed in predicting the
toss of a coin, the probability would be 1 in 2 (because
there are two possible outcomes - heads and tails).
- Type in the
TOTAL number of trials.
If combining data from a number of runs, the total number
of trials will be the number of runs multiplied by the
number of trials in each run.
- Type in the
TOTAL number of hits obtained.
- Click the
CALCULATE button.
Interpreting the
Statistical Analysis
The analysis performed is
based on the Critical Ratio procedure (z-test). This uses the
normal curve to approximate the results of the Binomial Test, and
is sufficiently accurate with large samples of data (you will be
prompted if you have insufficient data).
- Expected Hits
This is the number of hits you would most likely get by
chance.
- Actual /
Expected Hits
This is the ratio of actual to expected hits.
- Critical Ratio
(z)
If this value is greater than zero, then this means that
the number of hits obtained was greater than that
expected by chance. In contrast, if this value is a minus
number, it means that fewer hits were obtained than would
be expected. To test whether the results are
statistically significant (i.e., whether they support
psi-hitting or psi-missing) the probability of obtaining
the results by chance are next calculated.
- Calculated
probability
This indicates the two-tailed probability of obtaining
results that are this extreme simply by luck. If this
probability is low (generally less than 0.05, or 1 chance
in 20) then the results are said to be statistically
significant. If the probability is shown in scientific
notation, i.e., with an e- at the end, this means that
you should move the decimal point to the left the number
of times indicated by the number after e-. For examaple
1.2345678e-5 is the same as 0.000012345678.
- Significance
Level
This shows the standard level of significance
(two-tailed) associated with the calculated probability.
- Odds of this
result by chance
This shows how often you would expect to get a result as
extreme as this.
Top of Page
Page contents © Michael Daniels, 2001-2008. All
Rights Reserved.
