Crop Circles
- Thread starter unexplained_mysterie
- Start date
Sciforums cannot spare the requisite bandwidth; alas, a particular--from which one might reasonably generalize--will have to make do:
Where by some, on occasion, and in all seriousness, the name Ayn Rand* is invoked, shits and giggles must necessarily ensue.
* A painful dichotomy: note that she is strangely hawt in this photo.
Where by some, on occasion, and in all seriousness, the name Ayn Rand* is invoked, shits and giggles must necessarily ensue.
* A painful dichotomy: note that she is strangely hawt in this photo.
one would think even a cursory nod towards historical accounts is warranted. ja, hundreds of years rather than a simple "years ago". whether hoaxed or of unknown origin, they are alleged to have occurred. reports have been written. why not acknowledge? is it that easily dismissed? or perhaps one is entirely unaware of these accounts?
watch me similarly theorize.....
the aliens did not have a clue on how make crop circles when they first got here in the 70's. they practiced and over the years they got better. other aliens jumped into their bandwagon and the game was on. the current champions are a species from ngc 345235
/snicker
do you evaluate the technological and scientific prowess of an alien species by its ability to scrawl intricate designs on the ground from orbit or otherwise? why? show me how and why this skill necessarily evolves in tandem with others that allowed them to reach earth
i mean, it seems so totally arbitrary, this conjunction of differing technologies. it is akin to demanding that the moonshots be accompanied by a proficiency in say....cgi?
well make the distinctions. rather than taxing you to produce a list of the alleged manmade crop circles, do the ones you hold to be of unknown origin as they are again alleged to be lesser in number
yikes! i missed this (pure gold)
it appears we have someone knowledgeable about the methodology
stereologist
please elaborate in a precise and factual manner
the paper is here
thanks
I have no idea what you are talking about.
You address me but quote someone else, and then tell me I'm 'fixing on a hypotheses' (surely you mean 'hypothesis'?)
Well yes, I'm mainly addressing the theory that crop circles are messages from aliens. After all, that was what was being talked about when I first posted.
That's not what was being talked about though. Certainly towards the beginning of the thread there was mention of spacemen, but the notions of the esteemed Doc Haselhoff and his colleagues (not the musical ones) are the subject of dispute. Those being more of this variety:
Last edited:
Gustav,
I've heard the quote "The Devil makes use of idle hands", Which I think pretty much refers to people that aren't working could potentially to other things instead (some of those things not being legal).
In the case of a "Devil" portrayed creating a crop circle, you have to take into consideration what the symbolism is. An entity "destroying" (Vandalising) crops (Graffiti) which has no face, just the outcome of it's malice.
Obviously guessing as to who was responsible would of left complete ambiguity, so it would make sense to personify for at least making it feasible to discuss. (Otherwise where would have the circle come from?)
As for where the circle had come? Well the problem is we don't get to Scientifically evaluate the scene from the past, we don't know if the stems were inter-weaved with intricacy or if a giant lurcher dog had a bad case of the worms and decided to drag it's arse round in circles. This is the problem with such empirical evidence, it's absence of any credibility for absence of hard evidence.
I've heard the quote "The Devil makes use of idle hands", Which I think pretty much refers to people that aren't working could potentially to other things instead (some of those things not being legal).
In the case of a "Devil" portrayed creating a crop circle, you have to take into consideration what the symbolism is. An entity "destroying" (Vandalising) crops (Graffiti) which has no face, just the outcome of it's malice.
Obviously guessing as to who was responsible would of left complete ambiguity, so it would make sense to personify for at least making it feasible to discuss. (Otherwise where would have the circle come from?)
As for where the circle had come? Well the problem is we don't get to Scientifically evaluate the scene from the past, we don't know if the stems were inter-weaved with intricacy or if a giant lurcher dog had a bad case of the worms and decided to drag it's arse round in circles. This is the problem with such empirical evidence, it's absence of any credibility for absence of hard evidence.
I was sent an interesting document about crop circles. It is titled, "Opinions and comments on Levengood WC, Talbott NP (1999) Dispersion of energies in worldwide crop formations." Physiol Plant 105:615–624. The author is Eltjo H. Haselhoff.
The paper considers the work of the original authors and then submits some changes.
Basically the short comment on the paper offers what they consider a better fit for the data and justification for that change. The original paper fit the data to one curve and now a new curve is suggested.
There are a number of ways that one can measure how well a curve fits data. In this short comment Haselhoff uses the Pearson's coefficient as a means of showing that the curve is a good fit. The curves in this case are lines so the Pearson coefficient is appropriate.
Haselhoff points out correctly that the Pearson coefficient is better for the believed ball of light crop circle than it is for the known man-made crop circle.
All this paper proposes is that this test may be a means of detecting whether or not a crop circle is a hoax.
That's not very satisfying is it?
1. We notice in the so-called real crop circle that an outlier was excluded from the test. An outlier in the so-called hoax crop circle was not. No explanation is offered.
2. The sampling done in both crop circles was improper.
The authors of the original data sample in a preferential manner and clearly under-sampled. The whole goal of sampling is to avoid the often overwhelming task of measuring everything. The question then is how to obtain a sample to represent the whole. In the case of the crop circles the samples were chosen inside of the crop circles and with preferential orientation to the crop circle and preferential position with respect to the crop circle.
The samples were taken in a systematic fashion. That is actually a good choice. It is provable that systematic sample has a sampling variance no worse than random independent sampling. Usually it is much more efficient. But the sampling that was done was done as if the authors expected to see some form of radially symmetric result. One transect was reported from the center of a disc shaped area to the edge of that area. A group in Italy reports that actually two transects were taken, but only 1 reported.
Why stop at the edge of the crop circle? That is a sampling mistake.
I'll explain more shortly.
The paper considers the work of the original authors and then submits some changes.
Basically the short comment on the paper offers what they consider a better fit for the data and justification for that change. The original paper fit the data to one curve and now a new curve is suggested.
There are a number of ways that one can measure how well a curve fits data. In this short comment Haselhoff uses the Pearson's coefficient as a means of showing that the curve is a good fit. The curves in this case are lines so the Pearson coefficient is appropriate.
Haselhoff points out correctly that the Pearson coefficient is better for the believed ball of light crop circle than it is for the known man-made crop circle.
All this paper proposes is that this test may be a means of detecting whether or not a crop circle is a hoax.
That's not very satisfying is it?
1. We notice in the so-called real crop circle that an outlier was excluded from the test. An outlier in the so-called hoax crop circle was not. No explanation is offered.
2. The sampling done in both crop circles was improper.
The authors of the original data sample in a preferential manner and clearly under-sampled. The whole goal of sampling is to avoid the often overwhelming task of measuring everything. The question then is how to obtain a sample to represent the whole. In the case of the crop circles the samples were chosen inside of the crop circles and with preferential orientation to the crop circle and preferential position with respect to the crop circle.
The samples were taken in a systematic fashion. That is actually a good choice. It is provable that systematic sample has a sampling variance no worse than random independent sampling. Usually it is much more efficient. But the sampling that was done was done as if the authors expected to see some form of radially symmetric result. One transect was reported from the center of a disc shaped area to the edge of that area. A group in Italy reports that actually two transects were taken, but only 1 reported.
Why stop at the edge of the crop circle? That is a sampling mistake.
I'll explain more shortly.
muchos gracias
perhaps of some use.....
the two papers published in Physiologia Plantarum...
Anatomical anomalies in crop formation plants
Dispersion of energies in worldwide crop formations
eltjo's opinion on the latter also published in Physiologia Plantarum
Opinions and comments on Levengood WC, Talbott NP (1999) Dispersion of energies in worldwide crop formations
a critique of all 3 published in Journal of Scientific Exploration
Balls of Light - The Questionable Science of Crop Circles
perhaps of some use.....
the two papers published in Physiologia Plantarum...
Anatomical anomalies in crop formation plants
Dispersion of energies in worldwide crop formations
eltjo's opinion on the latter also published in Physiologia Plantarum
Opinions and comments on Levengood WC, Talbott NP (1999) Dispersion of energies in worldwide crop formations
a critique of all 3 published in Journal of Scientific Exploration
Balls of Light - The Questionable Science of Crop Circles
The author writes the following in the article:
Measurements taken as described are called ocular estimates since the measurements are taken of curved objects placed near a linear scale. There is a bit of guess work here as the observer has to mentally 'straighten the object' for the measurement.
The real problem here is that the objects all have some biological variation. There is often a huge differences between individuals of a biological population. The readers of this post have kidneys. Inside the kidneys are nephrons. On the low end a person might have 125,000. On the high end people have over 1 million. The range in people is 9X! That's a substantial natural range.
The plots we see appear to be nice, but are we seeing data or noise? How much of the measurement difference is due to biological variation and how much to the effects involved? The greater the biological variance, the more sampling that has to be done to see effects.
In the paper the author admits they tried a number of measurements to detect differences, but settled on one. There is no mention of the other measurements other than they were deemed less reliable/consistent.
An indication of the spread of data is the use of SDs. This is a good method of estimating the spread of data. The values in most of the plots are ratios, not the original data. Even in fig 1 the plots are of ratios. This figure suggests that the biological variation of the affected plants is in the range of 10% to 20%. The plot does not tell us anything about the biological variation of the unaffected grass.
Measurements taken as described are called ocular estimates since the measurements are taken of curved objects placed near a linear scale. There is a bit of guess work here as the observer has to mentally 'straighten the object' for the measurement.
The real problem here is that the objects all have some biological variation. There is often a huge differences between individuals of a biological population. The readers of this post have kidneys. Inside the kidneys are nephrons. On the low end a person might have 125,000. On the high end people have over 1 million. The range in people is 9X! That's a substantial natural range.
The plots we see appear to be nice, but are we seeing data or noise? How much of the measurement difference is due to biological variation and how much to the effects involved? The greater the biological variance, the more sampling that has to be done to see effects.
In the paper the author admits they tried a number of measurements to detect differences, but settled on one. There is no mention of the other measurements other than they were deemed less reliable/consistent.
An indication of the spread of data is the use of SDs. This is a good method of estimating the spread of data. The values in most of the plots are ratios, not the original data. Even in fig 1 the plots are of ratios. This figure suggests that the biological variation of the affected plants is in the range of 10% to 20%. The plot does not tell us anything about the biological variation of the unaffected grass.
The next problem with sampling is how the sampling is performed. In the case of these papers a single transect from the center of a circle is taken. The data is collected, manipulated, and plotted. The assumption here is that the data can be collected from a single transect since the data is radially symmetrical.
I do not see any effort to demonstrate the radial symmetry of the data. Instead of making this assumption, wouldn't it be prudent to demonstrate it?
It has been shown that systematic random sampling does no worse than independent random sampling and is often far superior. I refer specifically now to the concept or 'ortrips' or orthogonal triplets. A good way to do sampling of this type is to choose a direction uniformly from the range 0 to 360 degrees. Layout a sampling line in that direction. Next, lay out a second line perpendicular to the randomly selected direction. Now you have 2 orthogonal transects that have been chosen in a systematically uniform random manner.
By selecting the original transect direction at random the observer avoids selecting an orientation dependent on other factors such as:
1. lay of the land
2. the rest of the crop circle
3. difference due to local soil characteristics
4. visual differences in the crops
These and other factors should not affect the choice of transect direction. A comparison between the 4 selected transects, i.e. 4 rays, should indicate whether or not the observed changes suggest being radially symmetrical.
PS The triple in ortrips is from sampling in 3-d.
I do not see any effort to demonstrate the radial symmetry of the data. Instead of making this assumption, wouldn't it be prudent to demonstrate it?
It has been shown that systematic random sampling does no worse than independent random sampling and is often far superior. I refer specifically now to the concept or 'ortrips' or orthogonal triplets. A good way to do sampling of this type is to choose a direction uniformly from the range 0 to 360 degrees. Layout a sampling line in that direction. Next, lay out a second line perpendicular to the randomly selected direction. Now you have 2 orthogonal transects that have been chosen in a systematically uniform random manner.
By selecting the original transect direction at random the observer avoids selecting an orientation dependent on other factors such as:
1. lay of the land
2. the rest of the crop circle
3. difference due to local soil characteristics
4. visual differences in the crops
These and other factors should not affect the choice of transect direction. A comparison between the 4 selected transects, i.e. 4 rays, should indicate whether or not the observed changes suggest being radially symmetrical.
PS The triple in ortrips is from sampling in 3-d.
I'll read the Grassi article next. I wanted to get my own thoughts down before reading the Grassi article.
yes
i might be talking nonsense but is stuff like ...standard deviation/deviation from the mean.... of any relevance here?
That's not nonsense. A good measure of dispersion of a set of data is to use the standard deviation divided by the mean. The reason for using the standard deviation and not the variance is that the SD has the same units as the measurement. In the crop circle papers it is length. variances are better because variances can be added if the variances are independent.
The SD/mean is the relative SD. The value has no units. If the value is 5%, then anyone can think about it. If the SD is 5mm, then that is hard to understand without knowing something about the magnitude of the measurements. A SD of 5mm might be small if the mean if 2m, but big if the mean is 10mm.
One of the issues in sampling is that we need to sample enough to see more than noise. There are ways to estimate the biological variation and the sampling variation. Then we need to work no harder than to get estimates that have a variance that is around half the biological variance.
Think of it this way. If all of the plants were identical, then all you'd have to do is measure one plant at each quadrat (sampling site). Measuring more plants does not add more information since all of the measurements are the same. Remember I just defined the situation to be that way.
But the plants all differ. More plants need to be measured. How many is dependent on the variance of the measurement being taken. That can't be determined from the available information. Is 15 to 20 plants sufficient? No way to know.
One way to know when to stop sampling is to do a running average. As more data is collected the mean begins to settle down. It may be 5, 10, 50, or 100 plants. The only way to know is to collect measurements.