On this episode of Skeptiko, Dr. Julie Beischel on how to conduct psychic medium research:
“I think people fail to understand that proper research design includes optimizing the possibility of achieving positive results. If you wanted to study plant growth, you don’t put a dry seed on the bench top in the lab and then say, ‘Plants can’t grow.’ You use soil, water, sunlight, and then you study the growth of the seed.” — Dr. Julie Beischel
Alex Tsakiris: Welcome to Skeptiko where we explore controversial science with leading researchers, thinkers, and their critics. I’m your host, Alex Tsakiris. Okay, now on today’s show we’re going to dig into the nuts and bolts of psychic medium research. We’re going to look back on some of what’s been done in the past and the criticisms of it, and we’re going to look forward to the kind of research we might do in the future in collaboration with open-minded skeptics like Ben Radford from the Skeptical Inquirer and Steve Novella from theSkeptics Guide to the Universe. And there’s no better person to join us and help us in this than Julie Beischel from the Windbridge Institute in Tucson. Julie, as many of you know from our previous interview with her, is researching medium communication as a way of answering or at least examining the bigger question of whether our consciousness survives death. So Julie, thanks for joining us again on Skeptiko.
Dr. Julie Beischel: Thanks for having me.
Alex Tsakiris: As you know, from the email dialogue that we’ve had back and forth we had Ben Radford on the last episode and we wound up kind of wading into this whole issue of medium communication. The work that you did with Gary Schwartz came up. I think what I’d like to do is start with a quote from Ben that kind of summarizes his feelings about that research, and then I want you to respond in general and I also want you to, and I think this will be interesting both to me and to our listeners, I want to work with you and go through the real timeline of what we’re talking about in terms of Gary’s publication of the afterlife experiment books, then the response by Ray Hyman, Gary’s response and then you joining the team at University of Arizona and what all of you went through, some of the changes you made and then your latest research. So, let’s kick that off by going back and listening again to what Ben had to say when he was on our show on the last episode.
Ben Radford: Let me give you another example, Gary Schwartz’s afterlife experiment. Gary Schwartz has published a couple of books and studies in which he’s claiming that there’s strong evidence for communication with the afterlife. Ray Hyman, whom I sure you know, the psychologist of the University of Oregon, an incredible statistician went through and looked at Schwartz’s claims. He found serious methodological flaws in the analysis and methodology.
Alex Tsakiris: Okay, so general response to that?
Dr. Julie Beischel: That book was published before I started performing mediumship research at the U of A, in fact, before I even knew what a medium was so I really can’t comment on his content because I wasn’t involved at all. Like you said there are several published critiques and responses if people are interested.
Alex Tsakiris: Here’s what I was trying to get at, Julie. It sounds to me like the chronology that we’re laying out here is this book comes out, Gary’s getting some criticism, and as you’ve related to us in the previous interview your meeting with Gary or the chance meeting that you had was really coincidental, and Gary saw in you the ability to maybe tighten up some protocols that he had and was developing and maybe wasn’t spending as much time on as he should have. Certainly your background would be something someone in Gary’s position would immediately see the value of in terms of what he’s doing. Tell me if that’s correct and how your background does fit with the kind of research that Gary was doing.
Dr. Julie Beischel: Well, I can’t really speak to what he was thinking when he met me, but what was conveyed to me because I had a strong background in methodological design in a “hard science,” Gary and the donor thought that would be helpful in the continuation of the research into yes, tighter protocols.
Alex Tsakiris: Because again, your background is pharmacology where you’re basically being trained to take new drugs and see if they’re safe and effective, right?
Dr. Julie Beischel: Yes, in essence that’s what pharmacology is.
Alex Tsakiris: So you and Gary started working together. When did you actually begin your research with Gary? When did you start developing a new protocol and then looking for participants and actually starting trials?
Dr. Julie Beischel: When I first came on board, I wasn’t very familiar with the field at all. He was sort of already in the process of asking questions like: Does the sitter need to be there for a reading to be successful or can a proxy sitter serve in place of the sitter? Does that work? Can you ask the medium specific questions about this carnate or does it need to be free form? So we did some of those sorts of studies at the beginning and then once I was learning more and more about the field, it became my goal to statistically establish the existence of the phenomenon of mediumship itself in a contemporary laboratory with modern mental mediums. I felt like that we had sort of skipped a couple of steps ahead, so we needed to back up to what we call the primary hypothesis which is can mediums do what they say they are doing? That’s when we performed, designed and recruited subjects and performed the study of that then went on to be published in the triple blind study. We nicknamed that the Primary Hypothesis Study, but we also call it the AIR which stands for Anomalous Information Reception which is what we call the phenomenon of process of mediumship.
Alex Tsakiris: Okay, so let’s deconstruct that a little bit. You felt and Gary felt obviously that maybe what needed to be done is to back up and look at the underlying hypothesis which is … how would you state the underlying hypothesis in kind of layman’s terms?
Dr. Julie Beischel: I could probably say this in my sleep. So the primary hypothesis is this: Skilled mediums can report accurate and specific information about the deceased loved one, termed discarnates, of living people termed sitters, without any prior knowledge about the discarnates or the sitters and in the complete absence of any sensory feedback.
Alex Tsakiris: So how do you think the works that had been done prior was not directly addressing that kind of underlying hypothesis? Where had we kind of ventured away from that a little bit?
Dr. Julie Beischel: Again, I wasn’t around for that. That book came out in 2002 so the current research is well beyond those methods. In our lab and other works that’s being done by, for instance, Archie Roy and Trisha Robertson in Scotland and Emily Kelly and Diane Archangel at the University of Virginia, so I don’t think it’s useful to sort of nitpick at that because we’re so far beyond it that it’s like saying, “Why did we think the earth was flat?” Well, it doesn’t matter because now we know that it’s not.
Alex Tsakiris: Maybe, but I think the part that I disagree with that is that the history of it I don’t think we have to run from and again, I think the historical context that I would add to it is that so if Gary Schwartz is the first one to, and he’s not the first one because there’s a hundred year history of this, but if he stumbles across this himself and is blown away by these readings and he goes in and he does what he does in afterlife experiments, I think it’s useful to look back and say, “Okay, objectively where was that maybe stretching the boundaries of what we really thought we know and where do we need to kind of pull in our ranks a little bit and go back to kind of the primary hypothesis that we’re looking at?” So, I’m not really asking you to comment on the research per se, but more why did you feel there was this need to, and we can move past this, but why was there a need to more tightly define the research hypothesis?
Dr. Julie Beischel: How about we can look at it like you said. It’s a hundred years of research so we can look at it in sort of a historical perspective: How did it used to be done and what we’re the problems and then how is the current research addressing those problems? Do you want to list those issues to me again and I’ll address those?
Alex Tsakiris: Sure, fair enough. I guess just to push that point a little bit further, I think the way that you described it and the way that you described your process of going back and saying, “Okay, what is the primary hypothesis?” I think it’s outstanding and if you’re recalling the interview that I had with Ben Radford last week, it’s a mistake that skeptics make as well. So you’re kind of saying, “Hey, I felt like maybe the lab was pushing the boundaries a little bit further than we needed to. We needed to take a step back.” Well, the same thing happens when you listen to skeptics and what they’ll do is when they get some data in that’s uncomfortable that pushes their boundaries like, “Hey, maybe I’m not in solid ground,” then they jump ahead and start asking hypothetical questions, “Well, then why isn’t the communication this way? Why can’t they answer this? Why can’t they answer this?” What I want to really make clear is how important it is to be crystal clear and focused in the research hypothesis and I think that’s something that you and Gary did to your credit on this second round or whatever round you want to call it, but you did. So again, let’s get back to the primary objection of that initial research from the University of Arizona was 1) a judging bias, 2) a controlled-group bias, and 3) the one that everyone really kind of grabs a hold of right off the bat is sensory leakage. Do you want to go through those?
Dr. Julie Beischel: Yeah, let’s go from the back and work backwards, so sensory leakage, yeah, if the medium and the sitter are in the same room obviously there’s sensory leakage even if you use a partition because the medium can as soon as the sitter says anything, the tone of voice, even if the sitter just says yes or no, like you can say yes or you can say, “Yes!” it’s very different and that gives a lot of information to the medium. So, you can’t have the medium and the sitter in the same room. There’s always going to be sensory leakage. Even on the telephone there’s sensory leakage for those same reasons where a lot of information comes through the sitter’s voice. So in the current research the sitter is not on the phone with the medium. It’s just a proxy sitter so I serve as a proxy sitter and it’s just the medium and I on the phone. I don’t know anything about the sitter or the deceased person. Then just the medium and I do everything.
Alex Tsakiris: That’s great and I want to make that crystal clear. On the readings that you are doing now, the medium never talks to the person they’re doing the reading for. Is that correct?
Dr. Julie Beischel: That’s correct.
Alex Tsakiris: So that part is totally out of the equation, and I want to just backtrack for a minute and point out that you’re talking about it in very tight scientifically-controlled terms, which is great. But the counter-claim to that has never been proven either, as far as I know, and that’s that sensory leakage, non-verbal cues, tone of voice can explain all the information that we know is passed between a reading. I’m not saying it is; I’m not saying it’s not. But sometimes when we hear that skeptical counter-claim, we have to keep in mind that that is an equally unproven scientific claim. In fact, cold reading demonstrations using the same kind of even lax controls that were used at the University of Arizona back in 2002 have never been done showing that you can get names of deceased relatives and be able to put all those together in a way that makes a meaningful reading. Now there’s no reason to go back and try to recreate kind of flawed protocol and see if it generates a good control to counter the work that was done, but I just feel a need to kind of point that out because sometimes when we see that we need to improve something it’s needed conceding that the counter-claim has really been established and I think in this case it hasn’t.
Dr. Julie Beischel: I would agree with that.
Alex Tsakiris: Let’s move on and talk about controlled-group bias.
Dr. Julie Beischel: Let’s go to the other one because this same idea controls for rater bias, doesn’t hear the reading as it takes place, then we can give the sitter two different readings to score without knowing which is which. So that controls for reader bias because the sitter doesn’t know which reading is theirs and they score one that’s theirs and one that isn’t with the same, I don’t know how you want to say it, amount of bias or level of bias and it sort of washes it out then.
Alex Tsakiris: Let’s break that down a little bit, because I think it slipped past me the first time I ran across your research. I’m trying to think what’s the best way to do this because we’re kind of talking about the end, the judging bias. Maybe at this point it would be more useful to go through and very quickly go over your protocol and then it just becomes crystal clear to anyone who’s objective that there cannot be any judging bias and there cannot be any controlled-group bias. So why don’t we start with going over how the protocol works?
Dr. Julie Beischel: Our current protocol uses a quintuple-blind methodology so there are five levels of blinding. How we do it is we start with a group of sitters and we screen them and they describe the one person they wish to hear from most that’s deceased. Then we take those descriptions and we find pairs of deceased people who are the most opposite and that creates a pair of sitters. Then each medium reads a pair of sitters and a pair of discarnates.
Alex Tsakiris: Now, hold on before you lose me and lose everybody else. So you start with all these potential people that would like to get a reading. One of the things that’s different from the prior research that was done at the University of Arizona is you narrow it down and say, “Okay, I want you just to select one person that you specifically want to talk to on the other side, if you will, one deceased person.” So this prevents the kind of fishing around that makes everyone uncomfortable both skeptics and believers where the medium goes, “It’s maybe aunt or maybe grandmother or a friend,” and all this because now the medium’s going to be tasked with just targeting in on just one specific deceased person. Is that correct?
Dr. Julie Beischel: That’s correct.
Alex Tsakiris: And the other thing that you do here that I think is really, really significant and it’s brought up or kind of, Ben Radford last week had a similar kind of thought without realizing that you had already done this in the research that you published more than a year and a half ago, and that’s you actually pre-select pairs of people based on how different the person that they’re trying to connect with is.
Dr. Julie Beishcel: Right, I wrote down a quote and I think this is close to what he said. He said, “Enough variation between subjects should exist so that meaningful distinctions between subjects can be made.” That’s exactly what we do.
Alex Tsakiris: So take us through an example of how that would work.
Dr. Julie Beischel: We pair the people to most opposite in age, physical description, personality description, hobbies, and cause of death.
Alex Tsakiris: Okay, so I come along and I’m a potential research participant and I say, “Gee Julie, I’d really like to connect with my grandmother who passed away. She passed away when she was 95 and she was a small woman in stature, five foot one, and she had olive skin and her hobbies were going to church and knitting.”
Dr. Julie Beischel: Okay, we would find another sitter who had a discarnate that was young and active and blonde and had died of an accident rather than like your grandmother who died of old age. We pair the person to be most different on a variety. It’s hair coloring and build and height and cause of death would be part of the body that was affected fast or slow, natural or unnatural. Then hobbies are inside or outside.
Alex Tsakiris: We had a little bit of technical difficulty and I think what you were going over was all the different factors that you consider in a reading anyway and those are the ones that you kind of match for the maximum disparity possible. The reason that you do that is because … tell us.
Dr. Julie Beischel: We pair up pairs of discarnates to be most opposite. They’re the same in gender but most opposite in everything else. Then the same medium reads both people in the pair. If you said you wanted a reading from your grandmother and I wanted a reading from my sister, then a medium would do a reading for your grandmother and from sister. Then I would score both readings and you would score both readings, but we wouldn’t know which one was which because we weren’t there when the readings took place. So that controls for reader bias and then we give a score. I give a score to each reading and you give a score to each reading in addition to item by item scoring. There’s a very complicated scoring process. So my score of your reading serves as a control and your score of my reading serves as a control. Then we statistically compare the scores given by the intended sitter to the intended reading and two of the scores given by the controlled sitter to the controlled reading.
Alex Tsakiris: Awesome. Now let’s back up because actually, and I know you’re trying to make it as simple as possible, but there’s a couple of other steps that you go through there that are also interesting and noteworthy. So we’ll put you in the place of the person who’s talking to the medium. The first thing you say to the medium is you give the first name of the person you’re trying to connect with, right?
Dr. Julie Beischel: Yes, so during the reading all the medium and I have is the first name of the discarnate. We start the reading and I give the first name.
Alex Tsakiris: Great, and that’s a good point, too. That’s all you have, so you don’t have any information because we’re substituting you here as we said as a pronoun but it’s not really you. You end up being a person who decided who the two participants or who the two deceased people that we’re trying to connect with. You would be blind to that.
Dr. Julie Beischel: Right. A different experimenter does the pairing and screens the sitters. It’s my research assistant, Michael. So Michael gets on the phone with all the sitters. He gathers all the information. He does all the pairing and then I say, “Okay, I’m ready to do two readings,” and he gives me a first name.
Alex Tsakiris: Great. So you sit down there and you say, “Medium, I don’t know anything else other than I have a person here who wants to connect with someone named Sarah. Ready, go.” They go, so you’re blind to it. They’re blind to it. The reading happens. Now tell us a little bit about what happens after the reading has occurred, the transcribing and the reporting on that.
Dr. Julie Beischel: Let’s back up even a little more. During the reading, I give the medium the first name and they’re allowed to just give some general information for about ten minutes and then I ask those same four specific questions: Describe the physical appearance of the discarnate, describe their personalities, what were their hobbies or how did they spend their time, and what was their cause of death?
Alex Tsakiris: Good point. So even in that part though, for ten minutes you go, “Okay, medium. Go ahead, what are you getting? Sarah.” They can just say anything and then you take them through a very targeted process, which is also a very important difference with the way the prior medium research at the University of Arizona was done. One big difference that we pointed out at the beginning is now we’ve targeted one specific deceased person. We’re really not interested in information that comes through other deceased people. And number two, we’re targeting in on certain specific bits of information that we want. What did they look like – build, height, hair? What was their personality – introvert, extrovert? Hobbies, cause of death. These are really specific things that you are now asking without knowing what the answer should be or could be. You’re just asking the medium for responses to those, right?
Dr. Julie Beischel: Correct.
Alex Tsakiris: Now the reading ends. What happens next?
Dr. Julie Beischel: Then I’m still blinded. I still don’t know anything, so I take the recording of the reading and I turn it into a list of single individual items of the information that the medium provided. So part of that formatting process is if the medium makes any reference to the name that I have then I pull out all of those references. I pull out all of the maybes and the could bes. So if the medium says, “I think maybe, I’m sort of getting that she might have had red hair,” my item is she had red hair because you can’t say whether she maybe had red hair if that’s true or not and we’re asking the sitter if this is true or not. I’m not totally versed in every single mediumship study that ever existed on the planet, but I think that’s relatively uncommon. I think that’s a newer protocol.
Alex Tsakiris: Great. So now you’re formatting a transcript. First you get it all transcribed and then you’re formatting it down to these single declarative kind of statements of facts. A couple of things that you mentioned, just to be clear, if I said you’re doing a reading for Sarah you would take obviously any references to the name Sarah because that would tip off that that’s who the reading is for.
Dr. Julie Beischel: That’s exactly right.
Alex Tsakiris: And then you have these number of other things like you said. You make the statements clear. You take any kind of medium speak of … give me an example of some of the medium speak.
Dr. Julie Beischel: I wrote a paper about all this methodology that came out in the Journal of Parapsychology and I think the example that I used in that paper is a medium might say, “I’m getting that Sarah is below you, below the sitter, pardon me,” then that means that they’re in a younger generation. If they said that they’re to the side of that means they’re in the same generation or above is a generation older. I put brackets and I sort of define what the medium speak means to the sitter.
Alex Tsakiris: Okay, and this isn’t something you came up with. This is just something, a shorthand way of talking that the medium has that you’re translating for the benefit of the sitter.
Dr. Julie Beischel: It is language that mediums use naturally and then I define it for the sitter.
Alex Tsakiris: Okay, great. So now let’s get down to the payoff. Now you’ve done a couple of readings with you being the intermediary. You’ve transcribed them, broken them down into single list of kind of declarative statements, “Her hair is red. She passed away this way. She was introverted. She liked this, blah, blah, blah.” Now what happens?
Dr. Julie Beischel: So now I send those two readings, and they’re just numbered Reading 1 and 2, I send those to a third experimenter, Mark, and then Mark and I say, “These two readings for these two discarnates.” We’ll name them in groups so Michael says, “Here are two names in Group A.” Then we do the readings and then I send the two blind readings to Mark and I say, “These are the two readings from Group A,” but he doesn’t know which one is for which name and he sends them to the sitters for scoring. So now none of us can accidentally convey anything to anybody because we don’t know anything to convey.
Alex Tsakiris: So now the next person in this chain who you said is Mark. Mark gets the two readings and he sits down with me and says, “Okay, Alex, you wanted to connect with your grandmother, Sarah. Here are two readings. I don’t know which one is for your grandmother Sarah, but one of them is and one is for another person.” Then I’m asked to do what?
Dr. Julie Beischel: He doesn’t sit down with you. He emails the readings to you and Michael has previously trained you on how to do the scoring. So Mark sends you two readings. You already know how to do the scoring and by yourself you score each of the readings and you email back your scores.
Alex Tsakiris: Okay, basically how am I scoring these?
Dr. Julie Beischel: You have the list of items, so you score each item for how we say, “Think to yourself, how well does this piece of information fit?” So you can give it a numerical score and there are not five isn’t one better than four. They’re just named zero through five. Five is obvious fit. It’s a concrete hit. Four is fit requiring minimal interpretation to fit, like they almost got it; that’s pretty close. Three is a fit requiring maximum interpretation like, “Ah, I guess if you really squinted your eyes that might be right.” Then two is it doesn’t fit the person whose name the reading was for, but it fits somebody else. One is no fit; that’s totally wrong. Zero is “I don’t know. I don’t have the information to know whether that’s right or wrong.” Then when we do the analysis, we use scores as four or five as hits and all the rest of them we consider misses. You also give the whole reading a score from zero to six, which is a different scale. That scale is based on the work of Russell Targ, the scoring they developed for remote viewing, like how you would score a remote viewer’s picture or response. The third way of “scoring” is we say, “Pick which reading you think is yours.” So you have this item by item percent of accurate items data, and then you have a whole reading zero to six score of the reading. Then you have a binary yes or no “Is this your reading?” so you can do statistics on each one of those pieces of data.
Alex Tsakiris: Right. The reason you can, again, is because the whole process we went through is just like me sitting down with these random facts that I have no idea where they came from and I have to pick whether they fit, whether they match, rate them on a scale on how well they fit and then come up with an overall evaluation. So at this point we’ve kind of addressed, I think, now let’s go back to the points we were talking about a few minutes ago, the judging bias now in the controlled-group bias are both addressed. What was the complaint or the criticism, and there was some validity to it, of the previous objection to the judging bias?
Dr. Julie Beischel: Well, if you know the reading is yours, rater biases, if you know the reading is yours you have a tendency to score it as either more or less accurate than it is in reality. But if you don’t which one is yours, then you’ll score them with the same amount of bias.
Alex Tsakiris: So if this is the reading that I got back, I might want to please either the medium or the experimenter or I might want to displease either the medium or the experimenter. Or, it might just be human nature that I want to cooperate. I want to get along so I start seeing things in certain ways. That’s kind of out of the equation now because I now have no knowledge which one of these readings pertain to me and there’s no way I could have any knowledge of them.
Dr. Julie Beischel: Right. If you score a bunch of things as right, you’ll score a bunch of things as right in both of the readings so statistically that will cancel out and if you score both things low, statistically that will cancel out. What the statistic looks for is the difference between the two not, “She got 77 percent right.” It doesn’t matter. It’s what the difference is between the scores given by the sitters to their own readings and the scores that they gave to someone else’s readings without knowing which is which.
Alex Tsakiris: Great, and what about the medium who is giving a lot of generalities to their answers and are just kind of fishing around? How does that not affect this process?
Dr. Julie Beischel: Well, we’re asking for specific information, so they can’t just give general because we’re asking for specific pieces of information. If they just give general, “He’s kind of short but kind of tall,” that’s two items now. He’s kind of short and he’s kind of tall. So a medium that just provided general information, one would have a low accuracy percentage and the sitter would have a lot of trouble discerning between the two readings and they probably would have a low record of the sitter’s choice.
Alex Tsakiris: And what you’d wind up with at the end of the day is chance, sitters picking readings and basically picking them at chance levels – 50/50 kind of scores. Or if you’re saying how “What score would you give it – one through six?” everyone’s going to be at the three kind of middle ground. That’s not in the published research that you’ve done published back in 2006, 2007, I’m sorry. But you did it in 2006, published it in January 2007, correct?
Dr. Julie Beischel: I think we probably did it in 2005, but go ahead.
Alex Tsakiris: I think the chronology is really important and I want to get back to this later, but this is work that has been out there for a while. So you published it in January 2006, it’s not like we’re springing something on people here.
Dr. Julie Beischel: January 2007, so it’s almost two years old now.
Alex Tsakiris: So in that published study, I think you had sixteen times where people had to choose “which reading is mine.” What percentage of time did they choose the correct reading?
Dr. Julie Beischel: 81 percent. So thirteen of those sixteen people.
Alex Tsakiris: And that’s a pretty impressive, I assume that was statistically significant if I remember correctly from the paper.
Dr. Julie Beischel: Yes, it’s quite significant like you’re looking for a P value of less than .05 and I think that P was .001.
Alex Tsakiris: Okay, with that in mind now we’ve gone back and I hope that isn’t too detailed and exhausting for people, but I just think that’s fascinating. I think the whole process is extremely tight in my opinion and I’m certainly open to, I’m not a scientist, and I’m open to other people who are critical of your work to come forward and say where the gaping flaws and holes are. But it sure seems to me like it’s addressed the main issues. With that in mind, let me go back and play a couple more clips from Ben’s interview on the last episode and then in light of what we now know and have talked about, let’s hear what you might have to say about a couple of these.
Ben Radford: I do know about the research and one of the problems, one of the issues it doesn’t address is that in many of these cases the verification of the information is provided by the sitter. That is this not information that is supposedly coming from the great beyond that is verified by a third party person. Much of this was information where the medium will say, “I’m getting information from your husband or grandfather or whoever else,” and the information is judged either accurate or inaccurate by the sitter and there is an inherent problem right now that has not been addressed.
Alex Tsakiris: I just want to remind people that I specifically said, “Julie Beischel is on this show and she’s done this new research,” and he responded, “Yeah, I’m familiar with that research.” So just as a grounding that’s where we’re all coming from on that. What are your thoughts on that?
Dr. Julie Beischel: I’m going to use that, our current peer review methods, to address that. He’s listing as a criticism the fact that the sitter is the person that judges the information as accurate or inaccurate. So it’s important to keep in mind the scope and the goal of the research. The goal isn’t to prove the existence of an afterlife are inaccurate. Anyone on my team would say we’re not trying to prove an afterlife. That is not the goal. What we’re doing is examining the processes of mediumship in its natural environment with the proper controls, so normal readings between a medium and a sitter. Again, the general hypothesis we’re testing is can mediums report accurate and specific information without any prior knowledge and in the absence of any sensory feedback. So with that being the goal, the sitter has to be the person that is the judge because the information was intended for the sitter. We don’t ask the hypothesized discarnate. I’m just going to put an asterisk there. Every time I say “discarnate” I mean hypothesized discarnate. I’m not implying we have established that the medium is talking to a dead person because we haven’t. So the hypothesized discarnate, we don’t ask the hypothesized discarnate to take an algebra exam and provide information a third party could determine is accurate. That’s not what a mediumship reading is. We’re asking them to communicate with their friends and family, the only people who can determine if the information is correct and applicable.
Alex Tsakiris: Okay, that’s a good, very, very good point and to just declare from where I think Ben was coming from one of the criticisms of the prior research done at the University of Arizona is that if a person comes through and says, “I think I’m connecting with an aunt figure and her name is Dolly,” then the person who is accepting the reading can say, “Yeah, yeah that’s right.” Then there isn’t the independent verification of whether that’s right or wrong but that was troubling to some skeptics.
Dr. Julie Beischel: Right, because how could you verify whether or not something was wrong about my entire life? I couldn’t tell you everything you would need to know to do that. I don’t know how that’s even possible.
Alex Tsakiris: You’ve gotten around that whole issue by how? How have you specifically kind of addressed this issue that there is someone there who’s making these subjective judgments about the accuracy of the data?
Dr. Julie Beischel: We’re controlling for reader bias by having the person score two blinded readings and we’re comparing their score of a reading intended for them to their score of a reading not intended for them.
Alex Tsakiris: So the big problem before is the skeptics will point out, and in some cases very correctly, is that the reading could pile up like positive points like, “That’s a hit. That’s another hit. That’s another hit. Oh, your percent is going up and up and up.” Now what you’ve done is kind of taken that out of the equation because it really doesn’t matter how high or how accurate any particular reading is, it’s more of a comparison. How does this reading compare with the other reading? That’s how you’ve really, I think, in very novel way controlled for this whole idea of rater bias. They can be biased one way or another way but it’s going to wash out with the controlled reading that they have.
Dr. Julie Beischel: Right. They’ll be the same bias for each reading because they won’t know which is theirs.
Alex Tsakiris: Here’s the next quote from Ben.
Ben Radford: Part of the problem here that those descriptions that you just gave, those can still be vague. Someone says, “You know the person who’s coming through is a tall man with gray hair.” Well, it turns out that when the person died, he was bald. This gets back to the problem of having a sitter verify the information because the sitter says, “Yes, he had gray hair,” then that’s a hit. That’s good information, but the medium could also say he was bald in which case the sitter would say, “Well, he had gray hair but in his last years he was bald.” So you can have a medium giving two contradictory pieces of information, both of which would be considered a hit by the sitter.
Alex Tsakiris: I think we’ve probably covered that, but go ahead.
Dr. Julie Beischel: I think this is important just to address and then I’ll get to that specifically. Ideally laboratory-based mediumship research has to include two things: 1) an environment that optimizes the process for everyone involved, the medium, the hypothesized discarnate, the sitter in order to increase the probability of capturing the phenomenon if it exists, and 2) methods that maximize the blinding to control for any conventional explanations for the information. So together those two factors optimize the possibility of achieving positive results while also controlling for experimental artifacts.
Alex Tsakiris: Hold on! Hold on, wait.
Dr. Julie Beischel: I have a real world example that will make that much more understandable.
Alex Tsakiris: Okay, go.
Dr. Julie Beischel: We use this metaphor: You can’t study football on a basketball court using baseball players and the rules for hockey because if you get negative results you can’t say, “I’ve disproven the phenomenon of football” in that case because you’re on a basketball court using baseball players and the rules for hockey. That’s not a proper experiment. Similarly, it’s not appropriate to claim that Jason Elim could kick a 95-yard field goal if you give him a Nerf football, an empty stadium and no defensive line. That’s not real football either. In order to study football appropriately only trained skilled participants and the regulation equipment, environment and regulations can be used. The same thing is true for mediumship. So, negative results from a study using methods that didn’t optimize the environment and positive maximized blinding are equally ineffective at establishing new knowledge.
Alex Tsakiris: And this goes back to what you’ve been calling a naturalistic or a natural setting that tries to put everyone in the position that they are when people report these fantastic readings and I think that’s so common sense and yet it get so mixed up in the minds of skeptics so many times when I hear it. It goes back to your initial research hypothesis or proposition. If we’re going to study if medium communication is real, shouldn’t we do everything possible to let the medium and the sitter have at it in the way that they feel most comfortable with the only caveat being that we want to put the proper experimental controls on it? That just seems like common sense to me.
Dr. Julie Beischel: Yeah, and some people call that, “Oh, you’re cheating then!” because you’re optimizing. But it’s the same thing as saying, “If you give the football player a football instead of a watermelon…” or you know. It’s not cheating. It’s optimizing so you can watch the process as it takes place take place. I think people fail to understand that proper research design includes optimizing the possibility of achieving positive results. If you wanted to study plant growth, you don’t put a dry seed on the bench top in the lab and then say, “Plants can’t grow.” You use soil and water and sunlight, and then you study the growth of the seed.
Alex Tsakiris: Right. Do everything possible. Once you put the proper controls in place, why not do everything possible to recreate the phenomena? I completely agree. Now, what about Ben’s point about the information being vague? Maybe I say, “The hair was black, but I was bald older.”
Dr. Julie Beischel: Okay, I think his main point was that two pieces of contradictory information can be judged as hits by the sitter. Again, we have to keep in mind the process of mediumship. So first, the medium is interpreting images, symbols, sounds, whatever so she actually may receive two items that appear contradictory but we instruct them to say what they see. That’s the process and we’re studying the process so we can’t mess with the process. Second, as is often the case, as anyone who has ever been in a relationship can attest communication between people is not black and white. Things can be contradictory and accurate at the same time.
Alex Tsakiris: So then let’s follow that piece of information or those seemingly contradictory pieces of information through your protocol and see what happens at the end and see if they do create this mix up in the final judging of the data. So the person says, “Black hair, bald.” Two things that are contradictory. Older in life they were bald. Earlier in life they had black hair. How do you score it? How do you report it? And then how does the sitter score it?
Dr. Julie Beischel: We’re not asking the medium to describe a photo of the discarnate. We’re asking them to describe aspects of the whole entire dynamic lifespan of the person. Just like that’s entirely accurate for the person to be bald at some points in their life and dark hair in some points of their life and the discarnate may present themselves to the medium in each of those ways. It’s accurate for the sitter to score each of those items as accurate because they are both accurate. During scoring, again we’re not ranking up the number of hits. We’re comparing the number of hits to the number of hits in a controlled reading. So if the medium always, if the one reading says, “Oh, he was bald at some point and he had dark hair at some point,” and the other reading said, “Well, he was bald at some point but he was blond at some point,” then those two times that she said bald cancel each other out because they’re counted as hits in the control group and in the intended group.
Alex Tsakiris: Great, which is really the important distinction that we want to make. Okay, next clip from Ben Radford’s critique of your work.
Ben Radford: I’ll tell you what, use subjects who are do not fit what most people would consider to be a normal profile, to have a subject who maybe lost his legs in an accident during the war.
Alex Tsakiris: I think we’ve hit this.
Dr. Julie Beischel: I did actually want to address that specifically. Again, we have to keep in mind the scope of the research. We’re not setting up a condition where the discarnate has to prove something. We’re interested in studying mediumship under conditions that exist normally so that is where the medium provides information along with the sitter to identify the discarnate.
Alex Tsakiris: I think where he was going was just to say, “Let’s get people that are different so that those differences are highlighted in the reading,” and that’s the first thing that you do. I just want to bring us back to that and make it clear.
Dr. Julie Beischel: Yeah, and then another issue with that is that even if we did that it would be very difficult because there aren’t a whole lot of people in the world who have lost their legs or have a birth defect so it’s a very small percentage of people. And, even if you did that experiment where all the people had something special about them, then all you could conclude at the end of day was mediums can report specific and accurate information that is special to these people. Well, we’re not interested in that study. We’re interested in how does mediumship work in its natural environment?
Alex Tsakiris: Let me go ahead and really go out on a limb here and speak for Ben Radford. I think where he was going was just taking the mundane out of it which is kind of the next quote that I wanted to play for you and I think they’re related. I think he’s saying, “Let’s look at people who are different,” because I think incorrectly on his part he had this vision that all these readings are being done for 80-year-old grandmothers who passed away. You’re doing just the opposite of that. So let me play this next quote because I think it relates back to exactly what we’re talking about.
Dr. Julie Beischel: Okay.
Ben Radford: Well, part of my problem with the whole notion of medium communication is that a lot of the stuff is so mundane.
Dr. Julie Beischel: I have one word – football. That’s like saying, “Oh, I don’t like watching football because the players don’t fly.” Well, that’s not how football works so we have to keep in mind how the process works. We’ve actually found mediums most often report three kinds of information: information that allows the sitter to identify the discarnate so I call that, “It’s me! It’s me!”; events that have occurred since the discarnate’s passing, “I’m here! I’m here! I’m still here. I saw you at that birthday party or I saw that you got married or whatever. I’m still in your life even though I’ve died”; and then three, messages of an emotional nature, “I love you.” So we also have to keep in mind the scope of the research. We’re not looking to prove the existence of an afterlife. We’re interested in the phenomenon of mediumship. Mundane or not, the information is meaningful to the sitter and that’s what we’re studying. I think it’s also important to recognize that the majority of all human communication isn’t mundane. I like to say we study human communication. One of the people just happens to be dead. You don’t study normal human communication by asking people to talk about quantum physics. You just ask them to talk to each other and that’s what we’re doing.
Alex Tsakiris: And of course, what’s mundane for one person or another or a third observer is incredibly relevant to an individual. I can say from my own personal experience with a reading using just kind of a very basic of controls so it’s not like a scientific reading, some information that I would tell that I could report to you about that reading would seem incredibly mundane and to me it was deeply, deeply meaningful, almost to the point of bringing me to tears in terms of how relevant that was to me on an emotional level.
Dr. Julie Beischel: I think that brings up another point. I think Mr. Radford said something about, like he was saying, “Oh well to me this means the discarnate says, ‘He loves you,’ and he literally said, ‘What value is that?’” Again, let’s remember what we’re studying. It’s a deceased person communicating with the people that they love who are grieving and suffering and missing him, not a graduate student defending their dissertation. So what would you say to your family if you had died and you could witness their mourning? I would say, “I love you.” I think the mistake here again is that we’re trying to prove an afterlife and we’re not. We want to look at a mediumship reading with normal people receiving normal messages from their normal deceased relatives. That’s what we study. And I think critics think mediumship readings contain life-altering or extraordinary pieces of information, but that’s like expecting those types of information to show up in phone calls or letters or emails from someone that you haven’t been able to talk to for awhile. To an outsider like you said, the information may seem mundane but it’s really meaningful to the person receiving it. And I actually want to correct something you said in that interview, Alex. You said we don’t score “I love you” as accurate, but after we have the medium answer those four specific questions about the discarnate’s physical life, we honor the sitter and the discarnate by asking the medium, “Does the discarnate have any messages for the sitter?” This provides motivation for the sitter to participate and the discarnate to participate, not just to have to jump through our hoops, “Hey, dead person, show up for our study and answer these questions and then go away.” This may be the only opportunity they have to convey important messages to their living relatives, so those messages items are scored in the same way that the other items are. But again, we compare the controls so if they always said, “I love you,” it would be right in the control reading and in the intended reading and it would wash out but I also want to make the point that “I love you” is not an automatic hit. If the medium said that my mom was saying, “I love you,” I would score it as wrong because we never said that in my family. I would say you must be talking to the wrong person because my mom would never say that. So it’s not like when someone says, “I love you,” that’s an automatic hit because not everybody would say that.
Alex Tsakiris: Right, good point. Thanks for that clarification. In wrapping up all of the comments that Ben had about your research, we have one more that we’ve touched on but let me play it one more time and have you address it directly.
Ben Radford: I have yet, and if you can point me to an example of this I’ll be happy to follow up and write it down and say that I was wrong, but I have yet to find a case either in psi research generally or in Gary Schwartz’s experiments specifically that have information that neither medium nor the sitter knew.
Alex Tsakiris: Okay, so the final clip. Thoughts on that?
Dr. Julie Beischel: I think I did address that. Again, we’re not trying to prove an afterlife. We want to study mediumship under its normal condition. There have been cases where the medium often, I’ve heard mediums say, “A reading doesn’t happen without the discarnate providing information that the sitter didn’t know because it’s the most evidential for the sitter,” but it’s not evidential in the lab because it doesn’t separate telepathy from survival as an explanation. So it’s not ideal. It’s not something we would look to collect because it doesn’t answer any question.
Alex Tsakiris: What would be the best way for Skeptiko to put together some examples for Ben to see where this has occurred? Taking out the survival issue and just saying, “Is another person able to connect with information that the receiving person didn’t know about?”
Dr. Julie Beischel: I think when you do your demonstration you can specifically … We did a small pilot study where we specifically used the question, “Please provide information that is unknown to the sitter, medium, or experimenter and can be verified later.” A lot of the mediums said they do that every time anyway, so you can specifically ask for that type of information and then specifically follow up with the sitter. I would suggest that because again, we haven’t done it officially in a controlled way so I don’t have data on that.
Alex Tsakiris: Great. We’ll do that.
Dr. Julie Beischel: I just can tell you it happens. I’ve seen it. I don’t have any specific examples because we haven’t looked at it specifically.
Alex Tsakiris: Okay Julie, well that takes care of the clips that I wanted to play from Ben’s interview but there’s one other point that came up and it’s really my mistake or my issue that needs clarifying and that had to do with the editorial process at the JSE, the Journal for Scientific Exploration. I’ve kind of danced around this in the past and then after our last broadcast I got an email from both Dean Raden which I want to read a little bit from in a minute and then I also wanted to ask you about the editorial process at the Journal for Scientific Exploration because you are very familiar with it.
Dr. Julie Beischel: I found it to be one of the most stringent peer interview processes of any journal I’ve submitted papers to. I think Dean said his recollection was only ten percent of the papers submitted to the JSE are accepted and published. That’s the same rate as for the Journal of Science so it’s not like they are just accepting and publishing everything they get. Ninety percent of the papers that are submitted to the JSE get rejected.
Alex Tsakiris: Great, I think that’s good information to get out there. I also wanted to include Dean Raden’s comment because he was nice enough to send me an email and try to clarify it as well. Here’s what he writes:
In your interview with Ben Radford, the JSE is mentioned with some uncertainty about its peer review process. I’m an associate editor for the JSE and I serve as a referee for a dozen other journals ranging from mainstream psychology and physics to complementary and alternative medicine. I can tell you that among the half dozen associate editors for the JSE collectively we act as referees for at least twenty mainstream journals in disciplines ranging from astrophysics to zoology. At JSE we use the same standards applied for articles submitted to any other journal, so JSE is completely mainstream as to its peer review. It’s just that the topics are not mainstream.
So I think that should clear up both for me and for anyone else this idea of the peer review process at the Journal for Scientific Exploration being somehow lacking. It’s clearly not and in this game, scientific peer review, we have to accept that the same people that were respecting enough to treat as editors on a wide variety on mainstream journals are doing the same job on the JSE that they’re doing on these other journals and therefore, there really isn’t any credible reason to discount the value of an article published in the JSE. But you had something else to add to that as well, Julie.
Dr. Julie Beischel: Speaking of journals and peer review and such, there’s a journal calledPsychological Science and it’s one of the top ten psychology journals in the world. We actually submitted that triple-blind paper to Psychological Science first and the editor rejected it without even sending it out for peer review. The reason he gave was, “There’s nothing new here.” Then we submitted it to Explore where it was peer review and then published. So when critics list the fact that these types of studies are not published in more mainstream journals as a drawback of the research, it’s not really a criticism of the work. It’s not that the material being produced by psi researchers is not of high enough quality to be published in those journals, it’s often the case that the journals refuse to publish them. I’m not saying, of course, all psi research studies are stellar but the reality is a bias exists that often prevents even the really good studies from being published.
Alex Tsakiris: Julie, we’ve really taken a lot of time. Thanks again for spending so much time with us and joining us and I do hope that I can close the loop with Ben Radford. I want to bring him in and get his response, bring him up to date on what you’re saying and hear what he has to say about it and hopefully get a dialogue going because I think, I get the sense in talking to Ben that it wasn’t deliberate on his part in terms of him not understanding the research. I have to admit that I didn’t fully understand it so let’s just try and bring everybody up to date and then see where we’re at that point.
Dr. Julie Beischel: Okay. I think it’s really great that you brought these historically separate and warring sides of scientific skeptics together because as a scientist I’m always thankful for intelligent criticism of the work. Educated and constructive analyses only make it stronger.
Alex Tsakiris: Great. Then we will be back in touch and thanks again, Julie.
Dr. Julie Beischel: Thank you.
Alex Tsakiris: Thanks again to Julie Beischel for joining me today on Skeptiko. As usual, if you’d like to check out any of the links associated with this show or check out our previous shows visit our website at skeptiko.com. You’ll also find links to our form and an email link where you can drop me a note if you like. Until next time, that’s going to do it. Bye for now.