More experiments

stevenmu

After reading a link that was posted somewhere here, I've been toying with the idea of creating a computer based PSI test. Below are some of the ideas I've had for different tests

1. The Beep Test:
This would involve the computer making a load beep, and possibly changing the screen colour at random intervals. The user would try to press the spacebar (or some other key) as close as possible to the beep (but before it obviously). The test would be repeated a number of times. Presses after the beep would be discounted, as would the the best and worst times recorded. The result would then be an average of how close the user could come to pressing the spacebar just as the beep was going to happen.

2. Predict A Card Test:
Using images of either zenner cards or playing cards the user would attempt which card would appear next in a random sequence of a set length. The result would be how many the user got right and the probabilty of that.

3. Force A Card Test:
Again using images of either zenner cards or playing cards the user would this time decide which card they want to appear next in the sequence. The result would now be how many times the crd chosen by the user appeared next and the probability of that

4. Remote Viewing Test:
Possibly the hardest one to test. There would be a set of about 20 different potential symbols. I would then randomly select 5 and print them on a sheet of A4 which I would then seal in an envelope. I would do this in such a way that I wouldn't know the symbols either (wouldn't do to have people cheating by reading my mind ). If possible I would then place this envelope in a well known location that everyone would know but noone would have access to. In the test program, users would be given the set of 20 symbols and told to choose 5. I'm not sure about how to handle the results for this yet as a user could conceivebly repeat the test over and over untill they worked out the correct symbols (mastermind style).

Has anybody got any thoughts for other tests or improvements to the ones I've thought up ?
Is anybody be willing to try it out ?
Would people be willing to make a 20+ meg download (for the .Net runtime) or should I put in the extra effort to make it smaller ?

Oryx

Item 2 the zener test is already available in umpteen places on the net, sorry I have no links to hand cos its ages since I was at a site like that.

stevenmu wrote:

Would people be willing to make a 20+ meg download (for the .Net runtime) or should I put in the extra effort to make it smaller ?

I dont even understand this

NotMe

Tests 2 and 3 are basically the same thing aren't they?

Zillah

Psychic tests over computer, what a pile of nonsense, what ever gave you the idea!? :P


But 20+ meg download can kiss my ass, I've got broadband...mmm broadband...

So yeah I like the sound of the last one.

stormkeeper

phlematic wrote:

Psychic tests over computer, what a pile of nonsense, what ever gave you the idea!? :P

Well, it could be done... I'm a little confused as to the "deciding" of a card tho... Surely there'd be a lot of energy involved in making a card come up?

stevenmu

I got the idea for tests 1,2 and 3 from a link Keu (I think) posted a while back, there was an organisation in the US looking for volunteers to come in and do tests similar to these. Tests 2 and 3 are more or less the same, the only difference being the intentions of the user, studies have shown that people are generally more successful with 3 then with 2, altough whether they are actually forcing the card to be chosen or they simply decide to force the card tht happens to be next is unkown.

I realise that these tests (or 1,2 and 3 anyway) are available seperatly in other forms around the net but I thought it might be a good idea to have a single consolidated suite of tests. It would make it easier to compare how different people compare across a range of tests/abilities.

Katiek wrote:

Me wrote:

Would people be willing to make a 20+ meg download (for the .Net runtime) or should I put in the extra effort to make it smaller ?

I dont even understand this

Sorry, I should have explained that a bit better. I'm planning on developing this using Microsofts ".Net" platform. In order for people to be able to run programs developed with this they would have to download (for free of course) Microsofts ".Net Runtime" aswell as the program I create. The size of this is about 23Mb which would only take a minute or two if you have broadband but maybe an hour or two if you use a regular modem.

stevenmu

Kennett wrote:

Well, it could be done... I'm a little confused as to the "deciding" of a card tho... Surely there'd be a lot of energy involved in making a card come up?

I'm not too sure what you mean by this. If you mean that a lot of energy would be required to force a card then it wouldn't really, at least not a computer generated one. What you'd actually be doing is affecting the random number generator in a computer program. Altough I'd imagine that doing it with a real deck of cards would be a different matter altogether.

solas

some links
http://www.staceywolf.com/html/telepathy.html
http://www.getpsychic.com/html/psychic_games.html
-zenner like.

akari no ryu

If someone furnishes me with the Zenner Cards, I can code a very small (within regards to MB) test exactly like what you want for two and three.

akari no ryu

In fact, having just thought this through, it is possible to use databases, php and HTML to create an online system for this.

stevenmu

I'd thought of that, it would make test 1 kind of difficult and there was some other reason against it that I can't remember now. It would be very good for allowing people to compare their results to others and averages etc though. I might rethink it and see if I can remember that other reason.

akari no ryu

I'm a web devellopper in training, I can't do /proper/ programming. I only have basic java so I can't help with 1 or 4.
2 and 3 can be rather simply done, though. I don't think I'll host it on whitedragons.org but I will write it if you (plural) want.

stevenmu

I'm a programmer, so I can handle the coding end of it easily enough, altough some help with general design may be usefull.

I remember now the problem that caused me to rule out a web based system before, it was actually to do with test 1. It would be pretty awkward to write a javascript to play a beep/change the colour of a screen element at random times and also to record a keypress, determine how close it was and send the results back to the server. Probably not impossible but tricky to get right, especially if it needs to be cross-browser compatible. Sure I might give it a go anyway, it'll be an interesting excersise if nothing else.

akari no ryu

I have to say, though, I don't like the .NET system
I don't really like having to pay a $hi7load of money to buy a programming language.

stevenmu

If it's just the cost putting you off you could check out ...

SharpDevelop

or

Eclipse

or for web (ASP.Net development)

Web Matrix

and I think M$s Visual Web Designer 2005 express is free too.

None of them are as good as Visual Studio (I think there's free betas of 2005 still floating around somewhere if you want to google for them) but they're still pretty good.

akari no ryu

For web design, PHP, JavaScript, Perl, mySQL but I refuse to degrade the quality of the internet itself by putting asp onto it.

Yes, I'm a firefox user.

[Jackass]

I'd help by volentering for the last one if you want, seems to be the best1 to try out....only got 56k though

Oryx

I may be able to communicate with ethereal realms, but I cant understand a word of this thread

akari no ryu

We're talking about ways to try and test ESP over the internet, the <tech> was just trying to decide what was the best way to go about it.
I personally believe that avoiding Microsoft at all times is best policy, not just for web related stuff but in general, whereas stevenmu seems to be in favor of using their programming languages.

stormkeeper

KatieK wrote:

I may be able to communicate with ethereal realms, but I cant understand a word of this thread

You know what they say... To each their own.

Zillah

So is this ever gonna get off the ground or are we just gonna talk about it indefinately?

akari no ryu

phlematic wrote:

So is this ever gonna get off the ground or are we just gonna talk about it indefinately?

I'm working on doing it, at the moment, with one person and a php script. Once I have that done, and therefore a basic understanding of how to do it, I will port that to an SQL based system where one person sees the cards and one person guesses.
The problem lies with how to ensure the system can't be rigged and I think allocating the postitions randomly is the best way to do that, but unfortunately, you need numbers to do that.
Either way, phlematic, without having discussed it thoroughly, you end up with a crap system. Software testing should start LONG before the software moves to being written.

On that note:
The way I see it working is this:

There is a "deck" of twenty five "cards", five of each.
The php script "shuffles" said deck.
The person makes all twenty five guesses but gets no feedback until the end.
Wherein they get their results.

I don't know the equation to turn their percentage into the score I kept seeing when I took the test on line (I always score appallingly at zenner tests).

stevenmu

akari no ryu wrote:

We're talking about ways to try and test ESP over the internet, the <tech> was just trying to decide what was the best way to go about it.
I personally believe that avoiding Microsoft at all times is best policy, not just for web related stuff but in general, whereas stevenmu seems to be in favor of using their programming languages.

At the moment, professionally Microsofts .Net and SQL Server make the most sense for me. The last thing I need at the moment is to start learning more new technologies, plus Microsofts platform makes it easy for someone with practically no design talent (like me) to come up with something half decent so I think I'll stick with that.

phlematic wrote:

So is this ever gonna get off the ground or are we just gonna talk about it indefinately?

I was going to have a go at it over the weekend but I screwed up my PC (again, damn M$ :mad: ). I'll try and get it done this weekend and I'll be looking out for a cheap host in the meantime.

DubGuy22 wrote:

I'd help by volentering for the last one if you want, seems to be the best1 to try out....only got 56k though

Hopefully I'll be able to do it using ASP.Net which means it'll all work as a web page and 56k won't matter (too much).

KatieK wrote:

I may be able to communicate with ethereal realms, but I cant understand a word of this thread

You're probably better off, wanna swap ?

stevenmu

akari no ryu wrote:

There is a "deck" of twenty five "cards", five of each.
The php script "shuffles" said deck.
The person makes all twenty five guesses but gets no feedback until the end.
Wherein they get their results.

I don't know the equation to turn their percentage into the score I kept seeing when I took the test on line (I always score appallingly at zenner tests).

The best way to do it is as 25 seperate tests. If you're using 5 different cards then

1. Pick one at random
2. Compare to users choice
3. If correct add 0.2 to score (1 in 5 chance)
4. Repeat

Odds are that on average people will get 5 right out of the 25 tests giving a score of 1, anybody who gets (signifigantly) higher than 1 has beaten the odds. In theory it's also fair to say that anybody who gets (signifigantly) lower than 1 has also beaten the odds and displayed a 'paranormal' ability.

akari no ryu wrote:

Software testing should start LONG before the software moves to being written.

Conventional wisdom would say that you should test software as and after it's been written, but I suppose this is the "Paranormal" board after all