Peter Rubic and the study?

[PeterIMA]

Hy,
Do you think the increase was exponential or assymptotic?
Pun intended,
Peter

 

[hdtran]

Well, if it were an exponential increase, in a couple of years, you'd have greater than 100% of the aquarium fish being caught with cyanide. This is very probably impossible. Think about the process of elimination. That should answer your question (smilie here).

Hy

 

[Piero]

I'm often lurking here, but very rarely post my opinion. Sometimes though, when i miss breakfast...restraint eludes me...lol

DISCLAIMER: I have not read the study, and I'm no expert in anything...

But, it appears obvious to me, just based on observation and a rudimentary understanding of psychology, that Kalk is manipulating facts, and not open to any opinion but his own. As usual, he had a prefabricated conclusion at the beginning of the post, and he will never concede to anything else until someone supports his "secret" conclusion, which he then plans on dropping like a bomb. When the facts don't go his way...he's silent, if not evasive.

"I could be wrong" is an underlying necessity of scientific observation and healthy debate, but Kalk is oblivious. If scientific method was so blind, we'd probably still be rolling on square wheels.

It's sad, really. Because despite an obvious love for the ocean and its creatures, and potentially good intentions, Kalk has - over time - proven to be a detriment to any discussion of industry reform, and an obstacle to healthy debate. I just fail to understand why anyone bothers responding to the baited and manipulative posts...

Just my .02...callin it as i see it. I could be wrong...
_________________
what is global recession

 

[PeterIMA]

Piero, Unfortunately, I think you are right about Kalk and those in the trade (primarily exporters and importers) who wish to stonewall, obfuscate, evade, and continue to support destructive fishing methods for their own benefit (profits) to the detriment of the coral reefs, the fish sold to hobbyists (which die), the fishermen (outside the aquarium trade because of the reduction in reef productivity), and the fish collectors using cyanide (exploited and putting their health at risk).

Peter

 

[Kalkbreath]

Hy, Somewhere earlier in the Fuzzy Numbers thread I rebutted Kalks assertion that cyanide use had gone down by noting it had gone up from 1999 to 2000 (my results previously discussed) and was even higher in fall 2003 (49%) as reported by the Philippine Council For Sustainable Development (PCSD) web site. So, yes there was a change. Cyanide use has gone up (not down). You were correct in your extrapolation, except it was not a simple linear (straight line) increase.

Peter

Peter

Peter , do you really think there is less pressures on cyanide users today in 2004 then in 1998? The alternative sources for fish alone {like Bali ,Srilanka , Veitnam, Palau....} have forced The Philippines to have a better product . Then we have MAC bringing attention to MO like never before . As well as much greater attention by local media on the effects of cyanide .{how many links has John provided on this board during the last year} The collectors had no problem collecting 4 million fish in 1999 without much cyanide. Why is it out of the question today? Also how about coughing up some of the data for 1998 1999 2000? Were more cyanide prone fish included in the data in 2000? What were the ratios ? were more damsels included in 1996 then in 1999? You realize it would effect the out come?

 

[Kalkbreath]

But more importantly .............you have been avoiding the question. Explain how so few fish {those fish we as a hobby import} can explain anything ? How many blue damsels [Chrysiptera] the number three fish in most collected list. I count less then five per year? 111 species and 1800 fish by five years..........? Why would you conclude anything on such limited data?

 

[Kalkbreath]

I'm often lurking here, but very rarely post my opinion. Sometimes though, when i miss breakfast...restraint eludes me...lol

DISCLAIMER: I have not read the study, and I'm no expert in anything...

But, it appears obvious to me, just based on observation and a rudimentary understanding of psychology, that Kalk is manipulating facts, and not open to any opinion but his own. As usual, he had a prefabricated conclusion at the beginning of the post, and he will never concede to anything else until someone supports his "secret" conclusion, which he then plans on dropping like a bomb. When the facts don't go his way...he's silent, if not evasive.

"I could be wrong" is an underlying necessity of scientific observation and healthy debate, but Kalk is oblivious. If scientific method was so blind, we'd probably still be rolling on square wheels.

It's sad, really. Because despite an obvious love for the ocean and its creatures, and potentially good intentions, Kalk has - over time - proven to be a detriment to any discussion of industry reform, and an obstacle to healthy debate. I just fail to understand why anyone bothers responding to the baited and manipulative posts...

Just my .02...callin it as i see it. I could be wrong...

What facts have not gone my way? This study includes too few fish and in the wrong ratios .........Fifty percent of the data is missing ......The year by year data is not availible nor is which fish were tested which year ........ and its the only study of its kind !

 

[mkirda]

But more importantly .............you have been avoiding the question. Explain how so few fish {those fish we as a hobby import} can explain anything ? How many blue damsels [Chrysiptera] the number three fish in most collected list. I count less then five per year? 111 species and 1800 fish by five years..........? Why would you conclude anything on such limited data?

Kalk,

See the Statistics 101 thread.

The answers to all your questions here can be found in that thread.

Regards.
Mike Kirda

 

[mkirda]

What facts have not gone my way? This study includes too few fish and in the wrong ratios .........Fifty percent of the data is missing ......The year by year data is not availible nor is which fish were tested which year ........ and its the only study of its kind !

Um, ALL OF THEM.

Do you find fault in my analysis?
USING THE NUMBERS THAT YOU PROVIDED, I went ahead to show statistically that the sample sizes were large enough to account for Peter's conclusions at a 99% confidence level, with error bars of less than 5%.

The important thing to understand here is that most science is based on LESS THAN THIS! Most scientists are quite happy to have 95% confidence levels with error bars of 5%. I've shown you that the data set supports GREATER PRECISION than what most scientists are satisfied with.

What more could you possibly want?
99% confidence with less than 5% error bars not good enough?
What degree of precision do you want?

Regards.
Mike Kirda

 

[mkirda]

Peter , do you really think there is less pressures on cyanide users today in 2004 then in 1998? The alternative sources for fish alone {like Bali ,Srilanka , Veitnam, Palau....} have forced The Philippines to have a better product . Then we have MAC bringing attention to MO like never before . As well as much greater attention by local media on the effects of cyanide .{how many links has John provided on this board during the last year} The collectors had no problem collecting 4 million fish in 1999 without much cyanide. Why is it out of the question today? Also how about coughing up some of the data for 1998 1999 2000? Were more cyanide prone fish included in the data in 2000? What were the ratios ? were more damsels included in 1996 then in 1999? You realize it would effect the out come?

I'm not Peter, but let me address this anyway...

I think there are less pressures today than in 1998.
I outlined some data from a draft paper by Dante Dalabajan that talked about the economics of enforcement in one area of the Philippines in a previous post. Essentially, out of an estimated 600,000 violations of blast fishing and cyanide fishing (combined MO and LRFF trade), forty arrests had been made, 34 cases made it to court, and four convictions made over three or four years. Fines were less than could be earned in a few hours of cyanide fishing. In other words, there is no disincentive.

Philippine law makes it illegal to sell fish caught with cyanide.
Under that law, just about every single exporter in the Philippines should be in jail- with a penalty of not less than five years.

Back in 1997/8, a big fuss was made about how things needed to change.
Maybe less cyanide fishing was going on. Maybe the exporters just cheated by sending in known net-caught fish. After a couple of years with zero arrests, zero fines, the exporters wisened up- There is no need for them to worry as they were still free and had no case filed against them.

At that point, they ceased caring, and went about business as usual.

Occam's razor suggests that this simple explanation is the truth.
And the CDT data that has been released about more recent years all suggests that this is true.

Finally, contrary to your suggestion, I doubt more samples would do much to affect the outcome all that much. It would only increase the already more than adequate precision....

Regards.
Mike Kirda

 

[PeterIMA]

Mike, I appreciate your comments. I could not say it better.

Peter

 

[hdtran]

Mike,

Let me rephrase your statement about 99% and 5%, if I may...

Given the number of fish reported tested, let's say that 25.3% of the fish tested were found positive for cyanide.

Based on the total number tested, the likelihood that the actual percentage* of fish being positive as 25.3% + or - 5% (e.g. between 20.3 and 30.3%) is greater than 99%.**

Maybe rephrased in this fashion, Kalk would not object?

*Footnote: The actual percentage is the actual percentage out of the 14 million damsels, or whatever number is being used today. It's a number that you cannot know exactly, unless you choose to sacrifice all the fish for testing. The sampling and testing is a process of estimating the actual percentage.

**Second footnote: If I recall correctly, estimating the mean of a population has to use the Student T distribution, rather than the Gaussian. But as the degrees of freedom get large (math-speak), the Student T converges to the Gaussian.

***Last footnote: Maybe this was discussed in an earlier thread, but I'm completely lost.

 

[mkirda]

Based on the total number tested, the likelihood that the actual percentage* of fish being positive as 25.3% + or - 5% (e.g. between 20.3 and 30.3%) is greater than 99%.**

Actually, going back to that post, you will find that the error bar is closer to 2.5%, based on sample size. So, we can rephrase it as:

Based on the total number tested, the likelihood that the actual percentage of fish being positive as 25.3% + or -2.5% (e.g. between 22.8 and 27.8%) is greater than 99%.

But I am quibbling over trifles here.

Regards.
Mike Kirda

 

[Anonymous]

(in my best Homer Simpson voice) MMMM, truffles.

 

[Kalkbreath]

So what your stating is as a group you find nothing wrong with this study? ...............niether the vary limited number of fish included in the data{120 a year} or the non real world ratios{ like three puffers} is fine and dandy and the reslts a sure fire way at discovering the level of cyanide fishing ............So lets sea how quickly you all back off of this position and your real motives are revealed........................................................

Seven hundred fish sampled from 1996 to 2000 ............... seven hundred fish were used to gauge the other 15 million fish collected during the five years....

Really ?........................................................................................................... What if I told you MAC has decided to use this study as a Benchmark to test their fish for cyanide? Youwill really have no problems with MAC testing a mire 70 fish every five years? Bull*** MAC collects about 400,000 fish a year, about one-tenth that was collected during Peters testing data .......so by your collective INDUSTRY STANDARDS ........MAC would only need to only test the same ratio? Yes? That would work out to about TWELVE FISH A YEAR!!! You all are so full of it....... it smells . Now try and wiggle your way out of your self imposed predicament :lol:

 

[mkirda]

So what your stating is as a group you find nothing wrong with this study? ...............niether the vary limited number of fish included in the data{120 a year} or the non real world ratios{ like three puffers} is fine and dandy and the reslts a sure fire way at discovering the level of cyanide fishing ............So lets sea how quickly you all back off of this position and your real motives are revealed........................................................

Seven hundred fish sampled from 1996 to 2000 ............... seven hundred fish were used to gauge the other 15 million fish collected during the five years....

Really ?........................................................................................................... What if I told you MAC has decided to use this study as a Benchmark to test their fish for cyanide? Youwill really have no problems with MAC testing a mire 70 fish every five years? Bull*** MAC collects about 400,000 fish a year, about one-tenth that was collected during Peters testing data .......so by your collective INDUSTRY STANDARDS ........MAC would only need to only test the same ratio? Yes? That would work out to about TWELVE FISH A YEAR!!! You all are so full of it....... it smells . Now try and wiggle your way out of your self imposed predicament :lol:

Kalk,

It doesn't work that way.

Go back to the thread and take a look at the numbers.

Even you can see that the numbers do not increase in a linear fashion- In other words, You cannot use the ratio as you did above.

This is a math error on your part.

If I use a population sample of 400,000 per year, and I go for an acceptable level for scientific publication (95% confidence level with 5% error bars), the minimum sample size needed is 384.

Never did I claim that 12 would be acceptable- Those are your claims, and I expect an apology for your lying about it.

Regards.
Mike Kirda

 

[Kalkbreath]

So if they test one fish a day .........you will have no problems? Why can the total number be one tenth as large ......... yet the required sample only be cut in half? Actually MAC only exported about 100,000 fish last year ........How many fish should they have sampled?

 

[clarionreef]

Piero,
That was an astute observation a few posts back.
The diversion into a cyclical statistics debate has pushed most everything else off the table.
Excellent strategy or just recreational banter? Who knows?

Looks to me kinda like endless arguing of non-combatants 100 miles from the front lines in a battlefield.
Meanwhile the battle rages.
Steve

 

[mkirda]

So if they test one fish a day .........you will have no problems? Why can the total number be one tenth as large ......... yet the required sample only be cut in half? Actually MAC only exported about 100,000 fish last year ........How many fish should they have sampled?

Jesus, Kalk. READ THE STATISTICS 101 THREAD.

EVERY QUESTION YOU ASK HERE IS ANSWERED THERE.

If you really want to understand how sample size relates to Population size, I would again refer you to any decent Statistics book. It will explain it in far greater detail than I ever could.

Until you can show me that I made some mathematical error in my chart, this debate is OVER.

You lost.

Get over it.

 

[Anonymous]

never argue with an idiot, he'll drag you down to his level, and beat you with experience :?