Apr 16 2009

Homeopathy for Cancer Treatment Side Effects

Published by under Uncategorized
Comments: 80

The Cochrane Collaboration, an organization dedicated to evidence-based medicine, has published a review of studies of homeopathic treatments for side effects of radiation therapy and chemotherapy for cancer.  The results are unimpressive – consistent with the null-hypothesis that homeopathic remedies have no effect. And yet the review is being distorted to promote a very misleading bottom-line to the press – that homeopathic remedies have a role to play in cancer therapy.

One point has been made clear – the treatments under study are not for cancer itself, but for the side effects of standard cancer therapies: radiation and chemotherapy. However, the results are being presented as if they support the efficacy of homeopathic remedies, when they do not.

Homeopathy

Some quick background on homeopathy: This is a pre-scientific philosophy of medicine invented by Samuel Hahnemann around 1792. The principles of homeopathy are akin to sympathetic magical rituals. They include the notion that like cures like, or that a small amount of a substance that causes a symptom can be used to cure it. However, he then combined this principle with his “law of infinitessimals” which states that substances will become more potent as they are diluted, even if they are diluted beyond there point where any active ingredient remains. And finally he came up with the notion of succussion – that homeopathic remedies are given their power by shaking them 10 times in each spacial plane.

There, of course, are many more details to discuss, but that is the bottom line. There is no scientific basis to any of the ideas of homeopathy. This nonsense was rightly ridiculed in its day, but it has somehow survived in the culture, especially in Europe. It has made somewhat of a resurgence, riding the coattails of the broader CAM movement, but there may be signs recently that it is once again waning.

Two centuries of science after Hahnemann has not only failed to validate any of his principles, but has moved even farther away. Modern homeopaths now appeal to the notion that water can somehow have “memory” of the chemicals (but only the desired ones) that were previously diluted in it, and that this memory can survive the preparation of the homeopathic remedy, storage, later ingestion, absorption into the blood and tissues, transportation to its active site (whatever that is) and still retain its properties to have some physiological effect on the body. This is beyond fanciful, beyond science fiction – it is pure fantasy.

If that were not enough, clinical trials of the efficacy of homeopathy have been essentially negative (when looked at as a body of evidence). We see the typical distribution of results for ineffective treatments – a scattering of positive and negative, with a clear trend that the better designed a study the more likely it is to be negative. There is also the absence of replicability. To date no one has been able to establish with high quality replicable studies that any homeopathic remedy has any effect.

The Review

The review looked at 8 clinical trials:

Eight controlled trials (seven placebo controlled and one trial against an active treatment) with a total of 664 participants met the inclusion criteria. Three studied adverse effects of radiotherapy, three studied adverse effects of chemotherapy and two studied menopausal symptoms associated with breast cancer treatment.

So really this is three separate reviews, two of three trials, and one of two trials.  Those are extremely small numbers of trials to review, and perhaps that is why they were combined into one publication. Even 8 trials is a small number, and only three of the studies were considered of high quality. That’s slim pickings.

The results overall reveal that four studies were negative and four were positive. This is the kind of distribution we expect for treatments that have no physiological effect.

Here are the author’s conclusions:

This review found preliminary data in support of the efficacy of topical calendula for prophylaxis of acute dermatitis during radiotherapy and Traumeel S mouthwash in the treatment of chemotherapy-induced stomatitis. These trials need replicating. There is no convincing evidence for the efficacy of homeopathic medicines for other adverse effects of cancer treatments. Further research is required.

Given the slight data, this is not, in my opinion, a responsible characterization of the results. For other (non-CAM) questions Cochrane reviews, in my experience, would more likely simply say that there is insufficient data to draw any conclusions – not that there is preliminary data to support efficacy.

The notion that further research is required is also problematic. This is true in that the data available so far (with regard to these specific indications) is very slim. However it assumes that the clinical question is reasonable, when it isn’t. Homeopathy is without prior plausibility, and therefore one could argue that it does not deserve any research. Further, homeopathy overall has already been studied sufficiently to conclude on empirical grounds alone that it is without efficacy.

There are endless specific indications that could be studied and if you focus only on the specific indication we could be in an endless cycle of studying various homeopathic preparations for numerous indications. In fact, we are in this endless cycle. Frankly it is long past time that we should get off this treadmill and conclude that homeopathy as a scientific notion is dead and should be discarded.

But the homeopathy industry thrives on the phrase – “more research is needed.”

Further, many of the press reports of this review declare Homeopathic Meds Can Coexist With Conventional Cancer Treatment, or something to that effect. They are referring to the fact that the review found few side effects from the homeopathic remedies. I acknowledge that truly homeopathic remedies have no direct side effects – because they have no effects.

By that reasoning tap water is also compatible with conventional cancer treatment. This is an extremely misleading headline as it makes it seem as if there is some utility to combining homeopathic treatments with conventional cancer treatment, but even this friendly review did not show that.

You Got Drugs In My CAM

I also have not yet discussed the most deceptive aspect of this review. When interviewed for the BBC about this review, Edzard Ernst, professor of complementary medicine at the Peninsula Medical School, said:

Second, nobody doubts that undiluted remedies can have effects; and interestingly, the positive studies here seem to be on such medicines rather than on the highly diluted treatments which are a hallmark of homeopathy.

In fact, the calendula cream found to be effective in one study is not diluted at all and thus it cannot, to all intents and purposes, be considered to be a typical homeopathic remedy.

That is what we euphemistically call “cheating.”  It is highly misleading to include a study of an undiluted herbal drug and call it homeopathy. This was the single most positive study in the review, and it wasn’t even a study of true homeopathy.

This is a broader problem in CAM treatment and research. For example, studies are often promoted as showing that acupuncture works, when in fact they are not studies of acupuncture but of electrical stimulation through acupuncture needles. This says nothing about the principles of acupuncture, or whether or not sticking needles in the skin does anything physiological. It only confirms what we already know, that electrical stimulation can have certain effects, including pain relief.

Likewise homeopathic and even herbal remedies are sometimes adulterated with actual pharmaceuticals. (See here, and here.)  Most clinical trials of these products do not do purity testing to eliminate the potential for adulteration to affect the outcome of the study.

In this case the product was not adulterated with a drug, it was a drug and was just mischaracterized as “homeopathy.”

Evidence-Based vs Science-Based Medicine

The bigger issue here is what I and others perceive to be a critical flaw in the concept of evidence-based medicine (EBM) embodies in the Cochrane Collaboration. In practice they focus on clinical evidence almost to the exclusion of scientific plausibility (which, I should add, is itself based upon evidence). This was done specifically to level the playing field, so that treatments and practices that made sense were not given a pass on the evidence. Even treatments that make scientific sense and seem to work need to have evidence to back them up, because they could still be wrong, as history clearly indicates.

While I agree with this, they neglected to consider the other side of the coin – what about treatments that are extremely scientifically implausible. My guess is that EBM innovators simply did not consider this. They assumed (and this is not a bad assumption, it just came at the wrong historical time) that treatments under consideration would make some scientific sense. We’re talking about medicine not witchcraft, right?

But then along came alternative medicine, which is witchcraft desperately trying to conceal itself in the garb of modern medicine. By leveling the playing field EBM also gave a leg up to therapies with extreme scientific and biological implausibility.

And that is how we end up with a homeopath leading a Cochrane review of homeopathic remedies (i.e. water) and writing a distorted review that is spun to the press that homeopathic is useful when it isn’t.

This is also precisely why we need to replace EBM with SBM (science-based medicine). Prior plausibility needs to be factored back into the scientific process of evaluating potential medical modalities and claims.

Share

80 responses so far

80 Responses to “Homeopathy for Cancer Treatment Side Effects”

  1. MarkWon 16 Apr 2009 at 11:57 am

    I’m reminded of Billy Connolly’s joke:

    Imagine there’s a big crash and everyone is screaming and there’s a lorry driver lying screaming in agony – his arm’s been blown off and there’s panic and screaming and mayhem.

    “Let Me Through! I’m an Aromatherapist!”

    “Lay still! Oh aye this looks nasty. Here I’ll just dab a wee bit of tea-tree oil, you’ll be right as rain.”

  2. HHCon 16 Apr 2009 at 11:59 am

    After reading your blog and links, I wonder whether these cancer patients could not have effectively handled their side effects by asking the licensed professionals, pharmacists, dermatologists or dentists for an effective product. I know this takes some initiative on the patient’s part, but they are the key to their own healing process. If the homeopathic treatments in these studies were indeed inert, then the patients would have healed from their side effects within the same timeframe without spending the time and effort to take the homeopathic treatment.
    When I think about the concept of dilutes, I am curious if this does not take us back to genetics and breeding. For example, a dilute of a seal point Himalayan cat is a chocolate point cat. A dilute of a tortie point Himalaya cat is a lilac point cat. This works with breeding cats, live biological organisms. I can not conceive of using water for dilution purposes at Avogadro’s number to obtain any purposeful results.

  3. pecon 16 Apr 2009 at 12:56 pm

    “Homeopathy is without prior plausibility, and therefore one could argue that it does not deserve any research.”

    So … one could argue that nothing deserves research unless you already believe in it?

  4. straightgodlesson 16 Apr 2009 at 1:24 pm

    Good job in keeping the record straight on homeopathy.

    Steve you have often compared the history of research into homeopathy with that of parapsychology. Recently I was listening to your interview on the Skeptiko podcast and the interviewer mentioned that the presentiment (parapsychological) effect has been replicated 9 times in three university laboratories. I inquired about your knowledge of the presentiment research a few blog entries ago and your reply was that this effect “has not been adequately replicated.” I believe these are the nine replications:

    Bierman, 2000
    Bierman & Radin, 1997,1998
    Norfolk,1999;
    Parkhomtchouk et al. 2002;
    Spottiswoode & May, 2003;
    Wildey, 2001
    Bierman & Scholte, 2002
    McCraty, 2002

    Are you aware of these nine “replications?” I think I remember Tsakiris saying the effect has actually been replicated about 13 times in an interview with Lynne Kelly. Have you reviewed these replications listed here? Thoughts?

  5. superdaveon 16 Apr 2009 at 1:27 pm

    Also it should be noted that the calendula is a topical agent and it was applied for skin irritation. It is highly likely that rubbing ANY balm or salve will be helpful in reducing irritation provided there is nothing in the balm that would specifically increase the irritation.
    This also goes hand in hand with the fact that pain reduction is inherently subjective.

  6. Steven Novellaon 16 Apr 2009 at 3:45 pm

    “So … one could argue that nothing deserves research unless you already believe in it?”

    Only if you don’t understand the difference between plausibility and belief (more accurately “scientific acceptance”). Things could be plausible but unknown.

    Also there are notions that we can say from existing evidence are highly implausible. We are not starting from square one with each idea or claim. To believe that is to deny hundreds of years of scientific research.

  7. CKavaon 16 Apr 2009 at 3:47 pm

    Is there anyway that you can provide feedback to the actual reviewers or contest the inclusion of certain trials? I don’t really know much about how the cochrane collaboration works but I was under the impression that it was open to responses from qualified clinicians?

    I would imagine that, at the very least, the non-homeopathic trial’s inclusion in the review has very good grounds for being challenged?

  8. Steven Novellaon 16 Apr 2009 at 4:15 pm

    straightgodless – this is way off topic, and will take some time to specifically comment on each claimed replication.

    But here’s a quicky – First, Radin himself could not replicate his results. Two of his replications were negative, one “positive”, and he had to do some creative meta-analysis to rescue his results.

    The “positive” tests are a joke, he has a clear response after the pictures are shown only. But then he claims that there are tiny blips before the pictures are shown. This is just noise in the data – half the time he could pull it out, half the time he could not.

    The replications are similar. Also, they often used very different techniques and are not really replications. Many of them were also not independent. And I cannot even verify that they all showed results.

    For example, the Parkhomtchouk study, as far as I can tell from the translations, is negative – “Japanese do not show expected responses.” They hypothesis that Japanese require different stimuli (this is not replication, it’s anomaly hunting). http://wwwsoc.nii.ac.jp/iri/iri_eng/IRI-CurrentStudy01E.html#presentiment

    Take a look at Radin’s data. http://www.scientificexploration.org/journal/jse_18_2_radin.pdf It takes a huge leap of faith and a good imagination to think that there is any signal there.

  9. daedalus2uon 16 Apr 2009 at 4:35 pm

    If a study with an herbal product that is not diluted to non-existence is called “homeopathy”, then that is scientific misconduct on the part of the authors of the study. That is not a close call. The study should be retracted and the authors sanctioned.

  10. PalMDon 16 Apr 2009 at 4:52 pm

    The Traumeel study had many, many flaws. It was quite small, and dealt with a naturally self-limiting condition.

    Oh, and it studied a biologically inert substance (unless you count the saline).

  11. DarwynJacksonon 16 Apr 2009 at 5:20 pm

    It seems disingenuous to make the claim that homeopathic remedies have NO efficacy in any condition. If pharmacological substances or herbal preparations are diluted in water to values well below Alvagrado’s number then we should admit to ourselves that these remedies are highly effective in combating dehydration. Most such remedies, if accurately prepared, would be more beneficial than the vast majority of sports drinks on the market today.

  12. tmac57on 16 Apr 2009 at 5:58 pm

    I wonder how effective homeopathic ‘medicines’ would be if the people taking them were aware of how they are concocted. Some would probably be ‘enchanted’ by the rituals , but I suspect that the more grounded individuals would realize that there is no ‘there’ there, and would see on benefit.

  13. pecon 16 Apr 2009 at 7:17 pm

    [The “positive” tests are a joke, he has a clear response after the pictures are shown only.]

    I only read the abstract so far. He said the difference in EDA BEFORE seeing the pictures was significant (p = .002). So you’re saying he was lying?
    I’ll read the whole article. I can’t believe Radin just lied and hoped no one would notice.
    And he said the correlation across 4 experiments was significant. I don’t understand why he would lie or distort in the abstract.
    I will read the whole thing. I don’t think he has any logical reason for making up all that stuff, knowing it will scrutinized.

  14. pecon 16 Apr 2009 at 7:28 pm

    Sorry I read the abstract too fast. The correlation was for ratings of the pictures. He says all 4 studies were positive, or seems to say that. I will be back after reading the whole thing. Either he lied (or distorted) in the abstract, or someone has misrepresented this research, for some reason.

  15. pecon 16 Apr 2009 at 8:21 pm

    [The “positive” tests are a joke, he has a clear response after the pictures are shown only. But then he claims that there are tiny blips before the pictures are shown. ]

    It seems that you misunderstood Radin’s article. Of course there was a big response after the pictures were shown! Radin assumes that his readers know that was not the point of the experiment. But showing that the post-stimulus response correlated with the degree of emotion (which were rated by independent judges) was a necessary step in the process.

    This effect was originally discovered by accident, I believe, in an experiment that was simply testing physiological responses. Radin developed the idea further, with an extremely carefully controlled design.

    The point is to investigate the hypothesis that subconscious awareness is not limited to the past and present, but may sometimes extend a bit into the future. This idea may seem strange, but is not at odds with what is currently understood in theoretical physics.

    The pre-stimulus contrast in response between emotional and non-emotional pictures is what all those statistics are trying to get at. The reason the differences on the graphs look small is because they are small! But that is to be expected!

    All 4 experiments showed results in the predicted direction, but only two reached statistical significance. The non-significant experiments were not in the wrong direction, and one of them came fairly close. Of course, being close doesn’t count. But it certainly was not the case that the non-significant results opposed and cancelled out the significant results!

    Radin took the liberty of combining the experiments statistically and coming up with astronomical significance. But I guess he knows, and probably assumes that his readers know, that is does not mean all 4 experiments were successful. Only 2 were.

    This article describes extreme diligence on the part of the researchers. They know their work will be examined under a microscope by skeptics, and maybe even some pseudo-skeptics will glance at it.

    The presentiment effect is small, and it is expected to be small. That is why we use statistics. The Ns were large, as they should be when trying to detect small and possibly variable effects. But more power is probably needed anyway to get everything up to significance.

    I see nothing devious or deceptive or careless about this research. And I am a skeptic.

  16. SDRon 16 Apr 2009 at 8:35 pm

    pec,

    You have consistently shown that you are nothing near a skeptic. No one is buying your claim to be one. Give it up already.

  17. artfulDon 16 Apr 2009 at 8:50 pm

    pec writes: “- subconscious awareness is not limited to the past and present, but may sometimes extend a bit into the future. ”

    The function of the so-called subconscious is to make predictions about the future in terms of expected consequences of past and present actions. The only futures one can be aware of are our non-conscious and conscious predictions of their nature.

    Predictions can be accurate, but that accuracy is not confirmation after the fact that we were or are clairvoyant. How can you not be skeptical of any assertions to the contrary?

  18. rc_mooreon 17 Apr 2009 at 3:57 am

    I find the opinions of the medical practitioners in the linked press report (on the homeopathic study) to be quite disturbing. I don’t think people with cancer go to oncology clinics with the expectation they will be given placebos (even if they ask for them). And the continuing observations that the studies are not good enough to show homeopathic remedies are neither effective or non-effective are becoming quite old. Since homeopathy lacks any proven theoretical basis, at this point I doubt that the “right” studies will ever come along.

    The off-topic Radin study was interesting, mostly in how bizarrely overly careful it was in its description of methodology (pec was right about this I think, the researchers expect scrutiny) to then throw all cautiousness to the wind in the section on anticipatory biases. One again in this type of research, when the population averages do not meet expectations, isolate a few “gifted” individuals from the population and promote them to high statistical improbability. When I read statements like “a subset of participants specifically selected for exhibiting strong differential results suggestive of a genuine presentiment effect” in experiments that up that point were very careful about control and double blinds, my alarm bells immediately go off. Who did the selecting, and what were the objective criteria? Is it that will they prove the point you want to make, when the general population will not?

  19. eiskrystalon 17 Apr 2009 at 6:07 am

    -If a study with an herbal product that is not diluted to non-existence is called “homeopathy”, then that is scientific misconduct on the part of the authors of the study-

    No, the study conclusion should be changed to reflect that the medicine had an effect and that homeopathy didn’t. I think that would be much more damning. Accidental inclusion of medicine in homeopathy test shown to be much more effective. LOL.

    -And I am a skeptic-
    Only about reality Pec…only about reality.

  20. Steven Novellaon 17 Apr 2009 at 7:19 am

    Radin’s methods do seem pretty tight – which is probably why the results were essentially negative. Radin’s “gift” is in what he does with the data. He is the master of pulling a signal out of noise with fancy statistics.

    That effect sizes in ESP experiments are so tiny is very important. I don’t know why anyone would think a-priori that effect sizes should be small. That sounds like special pleading.

    But dubious phenomena tend to be right at the threshold of detection. Radin’s data is akin to a distant fuzzy picture of a “blobsquatch”.

  21. SteveAon 17 Apr 2009 at 7:21 am

    Pec,

    With regards to your comment:

    “The point is to investigate the hypothesis that subconscious awareness is not limited to the past and present, but may sometimes extend a bit into the future. This idea may seem strange, but is not at odds with what is currently understood in theoretical physics.”

    Could you expand on this? How does theoretical physics explain subconscious precognition?

  22. Neuroskepticon 17 Apr 2009 at 8:02 am

    I write about this study here and I am planning to contact the authors of the review to see why they included the paper about the herbal ointment which doesn’t seem to be homeopathic.

    Hopefully all will be revealed soon…

  23. pecon 17 Apr 2009 at 9:32 am

    “why the results were essentially negative. ”

    NO THEY WERE NOT!! Why are you making statements that anyone can see are not true?

  24. CKavaon 17 Apr 2009 at 10:08 am

    SteveA> “Could you expand on this? How does theoretical physics explain subconscious precognition?”

    You just had to ask didn’t you ;) and pec- you are a skeptic, come on! You’re about one magic bean away from worshipping the tooth fairy.

  25. Steven Novellaon 17 Apr 2009 at 10:32 am

    there were four trials. Two were dead negative. The other two had a statistically significant but tiny effect. To me – that is “essentially negative”.

    The meta-analysis he did to conclude that all were positive when taken together is worthless in my opinion.

    Bottom line – tiny effect blips only visible with dubious statistical analysis and inconsistency across trials.

  26. pecon 17 Apr 2009 at 11:14 am

    “there were four trials. Two were dead negative. The other two had a statistically significant but tiny effect.”

    That is WRONG. 2 were statistically significant with very small p values! The size of the effect is entirely irrelevant. You don’t know what you’re talking about!

    And NONE were dead negative. They all showed the effect in the predicted direction. Did you even look at the data?

    And he did not say all were positive. He analyzed all of the together just to show that a larger N would probably settle the question.

    You are just trying desperately to discredit research that has nothing wrong with it. The same effect was even found by other researchers.

    You are not even trying to be scientific. And I’m sure some of your readers can see that.

  27. rc_mooreon 17 Apr 2009 at 11:42 am

    pec –

    In regards to the Radin study:

    Are you saying the results are significant, even though the study itself indicates that the problem of anticipatory bias is not well addressed?

    If I agree that a statistically significant delta exists in pre-stimulus reaction between the emotional and calm picture in some to the trials, but conclude this is due to some other effect, such as anticipatory bias, then do believe the study is still positive? (In other words, to you ignore alternative explanations to statistical significance)

    Do you also agree that with such very small effects seen, that in order to remove such things as noise, investigator bias, etc, more independent trials are necessary before any real judgment can be made as to the validity of Radin’s research?

    Do you also agree that subsequent studies have not been able to validate Radin’s research? Why is that, do you think?

    And I must agree with Dr. Novella that your claim that such small effects are to be expected seems to be without any foundation whatever.

    It would be a fallacy to assume that things that do do not show up in measurements exist but are merely very small.

    If you argue that skin conductance is a poor measure (in terms of resolution) of pre-stimulus reactions, and that is why we expect a small delta, that would have some foundation. But I don’t think that is what your are claiming. And it would lead me to wonder why an experimenter would design an experiment using such an admittedly poor methodology. It is as though they are hoping to exploit the ambiguity of the data.

    I would also like to know “…how does theoretical physics explain subconscious precognition?”

    Please explain how you get over the theoretical hurdle that the mass and velocity of the atomic particles in my body precludes any direct involvement with sub-atomic (quantum) effects at all (via Planck), especially in terms of information transfer. Or maybe you are saying subconscious precognition is not quantum in nature, but electromagnetic. But is so, then please explain how it has escaped our measurement, as all energies at that level are will known and quantified.

    I await your response in good faith.

  28. pecon 17 Apr 2009 at 12:04 pm

    “more independent trials are necessary before any real judgment can be made as to the validity of Radin’s research?”

    More independent trials are always helpful when a phenomenon has not been generally accepted. I’m not sure how many it would take to convince pseudo-skeptics — something near infinity I guess.

    [Do you also agree that subsequent studies have not been able to validate Radin’s research? Why is that, do you think?]

    No, I did not know of any subsequent studies that failed to validate Radin’s research. Would you let me know what they are?

    “how does theoretical physics explain subconscious precognition?”

    It doesn’t explain it. There is just nothing currently known in physics that would prevent it. We are not talking about well understood phenomena — but if they were well understood we would not be talking about them.

  29. Steven Novellaon 17 Apr 2009 at 12:04 pm

    not statistically significant = negative.

    I never said the other two studies were not statistically significant, I said the effect sizes were tiny. This absolutely does matter as there is potential bias and noise in the data. Even a tiny bias can give a tiny but statistically significant effect.

    And it matters that the alleged presentiment effect sizes were much much smaller than the reaction effect sizes. This means that they are likely not physiologically meaningful, which calls into question using skin response to infer presentiment.

    pec – you continue with ad hominem attacks, straw men, and non sequiturs. You almost never address points I actually make. You merely assume whatever negative stereotype suits your purpose at the moment.

    I have confidence that my regular readers will discern whose arguments are scientific and logical.

  30. pecon 17 Apr 2009 at 12:14 pm

    “And it matters that the alleged presentiment effect sizes were much much smaller than the reaction effect sizes.”

    I does not matter! OF COURSE the reaction effects were much larger! We have brains that are tuned to this physical and temporal reality! We are animals living in this world! That does NOT mean other things can’t be going on in addition. Most parapsychology studies small effects, which are over-shadowed by ordinary sensory experience. If you discount it on the grounds that the effects are small, than no amount of high quality statistically significant research can ever convince you. You simply don’t want to believe it.

    What I like most about this presentiment research is that the subjects are not making any conscious effort. I think that most forms of ESP operate outside of conscious awareness. I think Radin is doing some great work and I wish more scientists had the resources to confirm, or disconfirm, his results.

  31. superdaveon 17 Apr 2009 at 12:18 pm

    Pec, the effect size is significant because it means that there is a possibility that what the experiment is measuring is not what Dean thinks he is measuring.
    This is especially true when taking such sensitive measurements. I just read Dean’s paper in which he suggests that human thought can effect laser interferometry. This technique is so sensitive that there are hundreds of possible sources of error that can account for the low levels of effect size that he records. You cannot say offhand that the results are not due to ESP but I think given the plausibility of ESP and the plausibilty of other sources of bias, the likelyhood of the results being caused by ESP are very low.

  32. Steven Novellaon 17 Apr 2009 at 12:25 pm

    Pec – again, you miss my actual point and simply launch into irrelevant attacks at my motivations and psychology.

    My point was that a small effect size calls into question the ability to infer presentiment from skin reactions.

    I do not categorically reject small effect sizes in science. It is absurd to infer that from anything I wrote. But, small effect sizes are harder to confirm. This means that research must evolve tighter and tighter controls until even subtle forms of bias are worked and. It means we need higher standards of replicable data before we accept the small effects.

    This does not exist is psi. As studies are replicated and improved, effects tend to disappear. In fact Radin and other researchers invented a concept to “explain” this – the “decline” effect. ie – cannot be replicated.

    It is no coincidence in my opinion that questionable and implausible phenomena tend to have effect sizes on the threshold of detection. Saying we expect this is special pleading. That does not mean it is not true – but never-the-less it is a post-hoc excuse for small effects. Occam’s razor favors a simpler answer – the effects are not real.

  33. TsuDhoNimhon 17 Apr 2009 at 12:28 pm

    In fact, the calendula cream found to be effective in one study is not diluted at all and thus it cannot, to all intents and purposes, be considered to be a typical homeopathic remedy.

    Calendula cream is a common herbal remedy for skin irritation. It has a mild anti-inflammatory effect, and works better than the same ointment base without the calendula. WOW! They rediscovered a skin cream that helps with dermatitis.

    That’s sooooo research-worthy.

  34. tmac57on 17 Apr 2009 at 1:06 pm

    Pec-”I think that most forms of ESP operate outside of conscious awareness.”
    Am I to take it that from this statement that you are convinced that ESP is real? Because in your previous statements you seem to be saying that there ‘might’ be something there, leaving the door open for ambiguity.
    Of course, I felt all along that you were a believer, but maybe you could tell us unequivocally .

  35. pecon 17 Apr 2009 at 1:38 pm

    “As studies are replicated and improved, effects tend to disappear.”

    That isn’t true. There are vast quantities of evidence for psi that did not ever disappear. And you had said Radin’s presentiment experiments had essentially NO effect. That is very different from saying the effects were small. And when an effect is shown repeatedly over thousands of trials, always in the same direction, and overall reaches a high degree of statistical significance — it is NOT reasonable to say it’s all a result of error!

  36. pecon 17 Apr 2009 at 1:40 pm

    “Am I to take it that from this statement that you are convinced that ESP is real?”

    The evidence strongly suggests that it is real. I would change my mind if I saw evidence that strongly suggested otherwise.

  37. rc_mooreon 17 Apr 2009 at 2:12 pm

    pec —

    Thanks for your response. It actually does not take an infinite number of independent experiments to convince me, but I do admit it takes more than one.

    I admit that I am not aware actually of any experiments that repeated Radin’s work (some were listed I think, but I am unclear on that point), so I guess I retract any implied claim that verifications have failed. If you hold that Radin’s work has validity, however, I think that your claim should be backed by verifications that have succeeded, just to be fair, under you own criteria.

    And I would still like your opinion on the subject of anticipatory bias, which Radin clearly acknowledges, but does not account for. He suggests that other studies have shown it may be effectively removed with very large sample sizes, but he gives not indication that his studies meet this criteria. And I am still confounded as to the significance of the “thirteen” subjects he remarks on.

    I must take objection to this response however:

    “There is just nothing currently known in physics that would prevent it. ”

    I specifically listed my objections, arising from theoretical physics (Planck’s limits). Did you miss that, or are you ignoring it? Perhaps we could both claim we do not know enough about theoretical physics to actually give a knowledgeable opinion either way.

    Again, thanks for you responses.

  38. artfulDon 17 Apr 2009 at 4:10 pm

    pec excerpted:
    ‘ “how does theoretical physics explain subconscious precognition?”
    It doesn’t explain it. There is just nothing currently known in physics that would prevent it.’

    Actually what we currently “know” in physics would prevent us from seeing the future without at the same time altering it by bringing it into the present.

    And it’s then pretty much anybody’s guess as to what we might learn in that heretofore inscrutable future that actually could unprevent us from exercising such visionary perspicacity.

  39. artfulDon 17 Apr 2009 at 4:23 pm

    The point being that pec has simply found another way of phrasing the “anything is possible” alibi.

  40. pecon 17 Apr 2009 at 5:19 pm

    [The point being that pec has simply found another way of phrasing the “anything is possible” alibi.]

    No. If it doesn’t defy the laws of physics, and if there is a lot of evidence for it, then I think it’s possible. I don’t believe in things just for no reason.

  41. straightgodlesson 17 Apr 2009 at 5:38 pm

    pec-

    Richard Bierman has demonstrated in a meta-analysis that over time effect sizes in every area of parapsychological research fall to zero. So your claims of strong replication are false.

    Here is Ray Hyman in the July/August 2008 issue of the Skeptical Inquirer:

    “Dick Bierman, the Dutch parapsychologist, for example, carefully re-analyzed major meta-analyses of parapsychological research on mentally influencing the fall of dice, the Ganzfeld psi experiments, precognition with ESP cards, psychokinetic influence on RNGs, and mind over matter in biological systems (Bierman 2000). He looked especially at the relationship between effect size and the date when the studies in each of these research areas was conducted. ”

    “Bierman fitted a regression line to the data in each area. In all cases, the regression line revealed a consistent trend for the effect sizes to decrease with time and to eventually reach zero.1 In addition to these linear trends from the meta-analyses, Bierman and other parapsychologists point to dramatic failures of direct attempts to replicate major parapsychological findings. These particular failed replications cannot be dismissed as being due to low power, which is the excuse commonly offered by Utts, Radin, and a few others.”

    So the claim that psi research has produced compelling evidence is DEMONSTRABLY false. So how can people like Dean Radin justify their claims? They can’t. Dean Radin responded to this article with a pathetic little essay, but he didn’t even bother to address this damning evidence.

    Will the presentiment experiments produce demonstrably reliable and repeatable effects? Looking at the history of this field I consider the probability of this occuring to be very very low. However, as good skeptics we have to watch for new data and be objective in our examination.

    We should also keep our eyes peeled on the “dogs that know when their owners are coming home” phenomenon. Take a look at this (pilot) experiment conducted by OpenSourceSciences. Dr. Novella I would appreciate it if you would watch these two short videos and comment. I don’t know what to make of it. The controls seem adequate and the data is highly significant.

    They are on the Skeptiko channel here:

    http://www.youtube.com/skeptiko

  42. rc_mooreon 17 Apr 2009 at 6:06 pm

    straightgodless –

    I know you were directing your comment to Dr. Novella, but I have had an interest in this effect before — so thanks for the link. I watched two of the videos, and I have a few comments:

    1. The study is not scientific. I am not sure what hypothesis is being tested, or what criteria are being established for acceptance or rejection.

    2. I am unclear on how significant the results are, because I don’t see any methodology for repeated testing. Are we looking only at successes? How far from the owner returning home (time and distance) does the effect occur. What time boundary do we place for the dog laying in its “home owner returning spot” do use to call the trial a success.

    3. What is the time distribution the owner is away — the video claims random. Does this mean the owner leaves the house for completely indeterminate times, between 0 and 72 hours say at completely determinate intervals?

    I would expect that if you let me collect data on the average homeowner leaving over a 5 year period, I could make some excellent statistical predictions on when they are returning home.

    4. If the researchers truly want to demonstrate the effect is due to telepathy, why not do that, by creating a methodology that removes all other variables. Place the dog in a full sensory deprivation environment, let a computer choose completely random times and intervals and locations for owner leaving, and see what the dog does.

    I have suggested this to Rupert Sheldrake supporters, only to be told, once again, that highly constrained experimental conditions tend to mask paranormal effects. How Sheldrake determines this is true, I have no idea. He does not test it as a hypothesis as far as I know. I refuse to buy his books to find out the results of his experiments, since I have access to published journals.

    5. I felt the result of 0% of time spent by dog when owner not on the way and 87% of time waiting when owner was on the way was very misleading. These numbers, I believe, apply to this one trial only, not a statical average over all trials. I could be wrong on this, the video is not clear.

    I am curious though, why did you feel “The controls seem adequate and the data is highly significant.” ? Is there something I am missing? or maybe you have more information on the research.

  43. straightgodlesson 17 Apr 2009 at 6:47 pm

    He (Alex Tsakiris) has a podcast called Skeptiko and he has talked about the details of the protocol on there-random return times and so on.

  44. artfulDon 17 Apr 2009 at 7:08 pm

    In the words of the pec herself, “Why do you waste time on that? Science is a practical method that does not concern itself with absolute certainty.”

    Plus I haven’t seen any attempts to get the dog to predict when the owner will leave home on one of these trips, or to signal a sense of what the owner is doing at any particlar time while away and before the decision to start back home is made.

  45. rc_mooreon 17 Apr 2009 at 7:12 pm

    Straightgodless –

    Thanks for the info.

    I checked the podcast, and found a little more information — his protocol has even more holes than the youtube video suggested.

    That is some shtick Mr Tsakiris has going on his site. I had never been there before. By balancing paranormal research and skeptical interviews he creates some credibility, but I was not impressed when I read through some of the transcripts.

    You asked Dr. Novella what he thought, and I noticed on one of the older podcasts he had already commented, so if you haven’t listened, you may find some answers there.

    I am afraid for myself however, I must continue to keep this one in my “if they want to really know, why don’t they do it right?” file.

    I can only assume that they prefer to confirm their pre-existing beliefs in the paranormal.

  46. rc_mooreon 17 Apr 2009 at 7:20 pm

    artfulD said:

    “Plus I haven’t seen any attempts to get the dog to predict when the owner will leave home on one of these trips, or to signal a sense of what the owner is doing at any particlar time while away and before the decision to start back home is made.”

    Ha, I would love to design some other tests. Like, what happens if the owner on their way home, receives a phone call and goes somewhere else. Or gets a flat tire. Does the dog leave “home” base and go somewhere else? What if the owner parachutes from a plane into backyard. What happens if a friend or taxi brings the owner home? What if the owner dies on the way home (not in the protocol, or course) — does the dog never go lay down on the home base spot again?

    Inquiring minds want to know!

  47. straightgodlesson 17 Apr 2009 at 9:45 pm

    rc_moore-

    I don’t think Dr. Novella has commented on this particular experiment. I think your confusing his comments with the ones he made on the Kane experiment conducted by Sheldrake.

  48. rc_mooreon 18 Apr 2009 at 1:19 am

    straightgodless –

    This is the podcast I was talking about: http://www.skeptiko.com/index.php?id=45, and I think you are right.

    I will give it a listen anyway when I get a chance.

    Thanks for the correction.

  49. Neuroskepticon 18 Apr 2009 at 9:33 am

    While I don’t want to interrupt the Great Telepathy Debate, the authors of the Cochrane review of homeopathy have responded to an email I sent them. I write about it here.

  50. tmac57on 18 Apr 2009 at 9:55 am

    The video may be somewhat interesting, and probably resonates with pet owners that feel that they have special connections with their pets, but it doesn’t appear to be a very good scientific experiment ,as pointed out by others. My guess is that they are cherry picking the desired results, which by the way is how I view the homeopathic studies. IMO only.

  51. HHCon 18 Apr 2009 at 11:47 am

    Neuroskeptic, Thanks for the information. Its kind of presumptious on the part of homeopathic medicine distributors to pretend to know what treatments cancer patients need and sell their wares over the counter. I am wondering whether the highly diluted homeopathic remedies are being used by persons with allergies to the normally prescribed products. I am suggesting that there may be a desensitization process ongoing in the patient.

  52. rc_mooreon 18 Apr 2009 at 1:35 pm

    Straightgodless –

    Sorry to continue off topic, but I think this hits at the heart of all psuedo-science, as it relates to homeopathy, the paranormal, especially in regards to experimental results that are promoted as being scientific, but in fact, are not.

    Neuroskeptic has uncovered the fact that the calendula cream cited as a positive homeopathic result in the study cannot in fact be claimed to be scientific result, because the definition of homeopathy applied in the study is not scientific.

    Garbage in, garbage out.

    In this spirit, I listened to the relevant (dog telepathy) skeptico broadcasts, and have come to the following conclusions:

    1. Mr. Tsakiris has no idea what the word random, in regards to scientific research, means. Or control. Or double-blind, or ….

    2. Mr. Tsakiris has never taken any courses in statistics, or has ignored the. In particular, the concepts of p-value and signficance are not a part of his critical thinking process.

    3. Mr. Tsakiris feels the real problem is not a lack of scientific evidence for the paranormal, but that skeptics refuse to empty their brains of knowledge, lower their standards, or accept anecdotes as objective evidence. If all skeptics could be as uninformed and gullible as Mr. Tsakiris, the paranormal could be accepted as fact.

    I draw these conclusions from Mr. Tsakiris’ own podcasts.

    Again, garbage in, garbage out.

  53. pecon 18 Apr 2009 at 4:29 pm

    [Bierman fitted a regression line to the data in each area. In all cases, the regression line revealed a consistent trend for the effect sizes to decrease with time and to eventually reach zero.]

    I would be surprised if Bierman actually said all the effect sizes reached zero. Hyman is a CSICOP fanatic so I don’t necessarily believe what he says. In fact he found nothing wrong with the SRI remote viewing results, over 20 years — no one ever said their effect sizes ever reached zero. Hyman recommended ending RV at SRI, even with its positive results, because he just could not believe it was because of psi.

    However I have heard from elsewhere that psi research can be tricky. You would never know that from reading Dean Radin’s books (yes he’s probably biased the other way). I think one major problem is you can’t screen out experimenter effects.

  54. pecon 18 Apr 2009 at 4:32 pm

    Oh wait — I guess Hyman meant the effect sizes were going down, so if the trend continued they would reach zero! Quite a different story. We don’t know from that statement if the effect sizes decreased by just a little, and Hyman therefore assumed (hoped) they would disappear.

    Our old friend bias at work.

  55. HHCon 18 Apr 2009 at 6:46 pm

    Briefly looked at Radin work. In only Experiment 3 & 4 was his equipment working properly as planned. He looking for emotional or calm responses to anticipated pictures. I predict you could obtain similar results asking subjects to get excited over flipping a penny, heads you win, tails you lose. Those studies are real dogs! :-D

  56. pecon 18 Apr 2009 at 6:56 pm

    HHC,

    Your comment makes no sense.

  57. [...] depth skeptical discussions of the review and it’s problems and limitations Steven Novella at Neurologica and the Neuroskeptic provide excellent [...]

  58. rc_mooreon 18 Apr 2009 at 8:17 pm

    pec –

    I thought you had left the thread (no one keep coming back forever).

    Any thoughts on my previous questions? Still awaiting your reply on Radin’s acknowledgement without redress per the anticipatory bias issue, and the basis in theoretical physics, in answer to my objections.

    Thanks, I will check back.

  59. artfulDon 18 Apr 2009 at 8:51 pm

    pec is only obeying the dictates of her future persona.

  60. DevilsAdvocateon 18 Apr 2009 at 9:11 pm

    Pec: “And I am a skeptic.”

    SDR: “You have consistently shown that you are nothing near a skeptic. No one is buying your claim to be one. Give it up already.”

    SDR, you have to understand the nomenclature and special definitions. Pec isn’t saying she’s like us. We are defined as pseudo-skeptics, while Pec is the ‘real’ skeptic, you see.

    It’s sort of like with the UFO folks, where a UFO is just an unidentified flying object, but a real UFO is an alien space ship.

  61. sonicon 19 Apr 2009 at 3:25 am

    straightgodless- rc moore- pec

    You have over simplified Dick Bierman’s position. Here is a link to a recent article.

    http://m0134.fmg.uva.nl/publications/2001/Benjaminschapter.pdf

    Regarding time and physics-
    Try
    http://www.npl.washington.edu/TI/
    Or
    http://www.npl.washington.edu/npl/int_rep/ti_over/node2.html#SECTION00020000000000000000
    This is “Transactional Interpretation” to physics by John Cramer. It proposes ‘backwards in time’ elements to the universe. It is considered a fairly reasonable interpretation in physics circles today. (It could be correct because it fits the data and is one of the efforts to get the ‘observer’ or consciousness out of physics. Apparently having backwards in time signaling to particles is less problematic to some than the notion that consciousness is a valid aspect of this universe.)

    If that is too much, look-up Brian Josephson or Freeman Dyson as prominent physicists who believe in paranormal phenomena. (Brian won a Nobel Prize for physics in 1973).
    I grant that the fact they believe it is meaningless as to the actual existence of the phenomena, but the notion that it is bad science or ‘impossible’ or unscientific always comes from people who know less about science than these gentlemen. And that is meaningful.

  62. CKavaon 19 Apr 2009 at 5:58 am

    Sonic, I suppose it must be equally as meaningful that critics of vitamin C megadosing for treating illness tend to have less nobel prizes than it’s main advocate Linus Pauling?

    Or it could be a sign that being the recipient of a nobel prize is no guarantee that your positions reflect the scientific evidence on all issues you care to comment on. Which seems meaningful to me.

  63. rc_mooreon 19 Apr 2009 at 11:56 am

    Sonic –

    thanks for the links, I will peruse them. My questions to pec were not related to any of Bierman’s conclusion, but it looks like interesting stuff.

  64. artfulDon 19 Apr 2009 at 1:43 pm

    pec and sonic
    When you look into the future, are you then required to do now what the future shows you will have done when you get there?
    Or has the future given you a glimpse of itself as a function of telling you what to do in the present?

  65. pecon 19 Apr 2009 at 2:51 pm

    ” the fact they believe it is meaningless as to the actual existence of the phenomena, but the notion that it is bad science or ‘impossible’ or unscientific always comes from people who know less about science than these gentlemen.”

    That’s right. I do not have a preconceived bias regarding the paranormal; I just consider the theories and evidence and keep an open mind. I started reading the Bierman article — thanks for the link — and will read the whole thing later. My initial impression is that he might be saying something similar to what I have said for years, in discussions of parapsychology — that conscious awareness interferes with psi and should be eliminated from parapsychology experiments.

    I had even written to Sheldrake and to Radin about this, because they never seem to mention it. I think paranormal research should use animals or young children as subjects, or adults who do not know the purpose of the experiment. Ganzfield, pk with RNGS, and remote viewing all depend on conscious efforts of subjects.

    The presentiment research seems promising to me because at least the subjects are not expected to make a conscious effort to see the future. Of course they probably know what the experiment is about, and that might be a problem.

    Presentiment research could be done with animals, and I think the effects might be stronger and more reliable, and easier to replicate.

    PK (mind interacting with RNG) experiments have been done with animals and were successful. I could not understand why the PEAR researchers always used human subjects who were trying consciously to influence the machine.

    So hopefully parapsychologists are starting to wake up to this problem. I will read the Bierman article and see if that’s what he was talking about.

  66. rc_mooreon 19 Apr 2009 at 3:06 pm

    artfulD asked:

    “When you look into the future, are you then required to do now what the future shows you will have done when you get there?
    Or has the future given you a glimpse of itself as a function of telling you what to do in the present?”

    Interesting question. I have a question that is much more basic, drawn from the experiments on presentiment described: How much delay in time, between the measurement of change in skin conductivity, and the presenting of the target, can occur and still show correlation.

    It seems contradictory that a presentiment (knowing something before it occurs), is affected by the passage of time (how long before the presentiment and the effect). I would expect that the statistical significance begins to below expected thresholds as the delay increase.

    This is why the problem of anticipatory bias must be addressed. This is a real effect, which also diminishes with increasing delay.

    Until this is resolved, I will use an Occam’s Razor approach, and attribute Radin’s result to this well known data confounder, not the paranormal (As I have noted, Radin himself brings this up, without resolution). I am open to possible explanations, (pec?), if offered.

  67. artfulDon 19 Apr 2009 at 4:20 pm

    People that give consideration to the possibilities of observing the future should also give consideration to redefining the concept of future, because as it stands now, “future” refers to something that will happen at a later time, yet is unclear about whether this is by way of a predetermined plan coupled with omnipotent power or by way of a non-causal yet inevitable sequential eventuality – or is perhaps something viewable by the rare talent of the perhaps even rarer omnisciently prescient animal. Just for starters.

  68. pecon 19 Apr 2009 at 5:16 pm

    Any kind of scientific research into precognition or presentiment will have to be full of problems and philosophical confusions about causality, etc. That doesn’t mean it shouldn’t be tried.

  69. artfulDon 19 Apr 2009 at 5:29 pm

    It does mean that you should at least be able to understand what it is that you are testing for, and thus be able to structure your experiments and analyze the significance of your results accordingly.

  70. pecon 19 Apr 2009 at 7:17 pm

    They are trying artfulD. Science is not simple.

  71. pecon 19 Apr 2009 at 7:25 pm

    straightgodless: “Richard Bierman has demonstrated in a meta-analysis that over time effect sizes in every area of parapsychological research fall to zero. ”

    That is just not true.

    Hyman: “Bierman fitted a regression line to the data in each area. In all cases, the regression line revealed a consistent trend for the effect sizes to decrease with time and to eventually reach zero.”

    But I just read the Bierman article and he definitely did not say that. He said that in some research areas there has been a decline effect. He did not say the declines reached zero, and he did not say this was true in all areas.

    Hyman framed it in that way so readers, such as straightgodless, would assume the effect size reached zerio. But he was only talking about the trend.

    And if you read the Bierman article, you will see that he says there is very strong evidence for psi, strong enough to be beyond doubt. But in some areas there has been a decline, which Bierman does not think means the effects aren’t real.

    Some research programs, such as Stargate, did not experience any effect size decline. Others, such as PEAR, had a decline and then a rebound.

  72. artfulDon 19 Apr 2009 at 9:06 pm

    pec writes: “Presentiment research could be done with animals, and I think the effects might be stronger and more reliable, and easier to replicate.”
    definition: presentiment |priˈzentəmənt| noun
    an intuitive feeling about the future, esp. one of foreboding : a presentiment of disaster.

    That’s of course how animals or any life forms survive. By suffciently accurate predictions about how its choices will affect its future in its particular environment.

    You can brush aside philosophical aspects of scientific hypotheses, but I can conceive of nothing an animal can choose to do, consciously or otherwise, as evidence that it has seen a future event and whose actions can then be proven not to have been, in and of themselves, determinant of that consequent future.

  73. artfulDon 20 Apr 2009 at 2:22 am

    To put it more succinctly, it can’t be demonstrated that the sequence of events allegedly observed wouldn’t have been altered by the act of observing them, and therefor not to some extent inaccurate and/or indistinguishable from a prediction.

  74. eiskrystalon 20 Apr 2009 at 3:58 am

    and Pec derails another thread.

    This was a thread about medicine…the results were so close to negative i doubt it would be worth the cost to take it and one of the successful trials used a known antiseptic cream, not homeopathy. Way to sqew the results.

    Pec can gab on all day about quantum effects, but it doesn’t change the fact that medicine has to work noticeably better than placebo and be cost effective.

  75. pecon 20 Apr 2009 at 8:06 am

    Someone else brought up the presentiment research. The blog author said the effects were essentially negative, which was just not true. So I felt I should say something because I was sure all the materialist readers here would believe it, even though it wasn’t even close to being true. And they would not even check.

    Doesn’t it bother you if someone makes statements that are ideologically motivated and unrelated to scientific evidence?

    It would bother me just as much if someone made claims for paranormal research that were untrue.

    In general, I don’t like deception and the substitution of ideological bias for scientific curiosity.

  76. pecon 20 Apr 2009 at 8:28 am

    And by the way, if something shows a declining trend that does NOT mean it’s inevitably going to zero! I’m sure you can think of your own examples.

  77. DevilsAdvocateon 20 Apr 2009 at 9:51 am

    I despair of the day when these paranormalists might understand that skeptics apply skepticism to skeptical writers’ offerings, in fact, even more rigorously, and chuckle at their attempts to step in and show us the errors of our sycophantic ways.

  78. HHCon 20 Apr 2009 at 7:20 pm

    OK, I see pec wants to do paranormal research only with animals, young children, and adults who do not know the purpose of the experiment. Does this include the senile?

  79. eiskrystalon 21 Apr 2009 at 3:53 am

    -The blog author said the effects were essentially negative, which was just not true-

    Depends what you mean by “essentially negative”. I’d say 3 out of 4 homeopathy medicines showing no significant difference against placebo is pretty close to negative. Also the one study that showed a positive for homeopathy only had 32 people in it.

    So…one small interesting result, but essentially negative for homeopathy in general…ie…their “medicines” don’t do what they say they do 3 times out of 4.

  80. pecon 21 Apr 2009 at 2:11 pm

    That comment was about presentiment research, not homeopathy. Specifically, 4 experiments where 2 had significant positive effects with very low p values, and two were non-significant but in the predicted direction. One of those had only 10% odds of being by chance. When all 4 were analyzed together, the p value was miniscule.

    So no one would consider that essentially negative. It was not definitive, since only 2 out of 4 reached significance, but very very far from negative. Experiment results are not absolutely positive or absolutely negative, and to call results essentially negative you would need an overall flat or negative effect.

    The blog author also complained about the effect size, saying it was too small to separate the signal from the noise. However, separating signal from noise is exactly the purpose of inferential statistics.

    And he thought the effect was small because the post-stimulus effect was much larger than the pre-stimulus effect. But that is what you would expect in this kind of research.

Trackback URI | Comments RSS

Leave a Reply

You must be logged in to post a comment.