Author
|
Topic: Test Format
|
Capstun Member
|
posted 05-02-2005 04:45 PM
Trying to do too much at one time here! The case name is Gilliard, not Gillian. That's what get for staying up and watching reruns of the X-Files. IP: Logged |
Barry C Member
|
posted 05-02-2005 05:16 PM
This case resulted in a he said / she said type of argument between Honts and Abrams that didn't do polygraph a whole lot of good. Honts and Abrams had a go round over it in the APA Journal some time back if anybody is interested in self torture. I got the impression they're not particularly fond of each other.Abrams added one directed lie to the final chart of his test, and he used that info to draw his conclusions, which Honts jumped all over. Honts used a mix of PLCQs and DLCQs when he started playing with the concept of DLCQs, but he tossed the PLCQs because he found the DLCQs to be more effective. Keep in mind the Utah guys stim to CQs in any test. Everybody says intuitively that it should result in more false negatives, but that's not supported by research. What it does do (apparently) is result in symmetrical cutting scores, something that isn't true of other methods. (Research shows that truthful people respond to CQs at about 75% of the intensity that liars respond to RQs, which is why many studies show a bias against the truthful. It also shows Backster - and now Matte - was right in creating asymmetrical cut-off scores.) It seems this post turned back to the Utah topic we had elsewhere, and I think it caused another record. I've never seen two pages of posts!) IP: Logged |
CHSBOY Member
|
posted 05-02-2005 06:38 PM
Based on my understanding of Abram's concerns as stated in his writings in Polygraph years ago, I agree with his concerns about over-emphasis on the DLCs and lack of balance by many using the technique. That can cause problems.I may get around to checking on the 'hybrid' issue when I return (from a road trip). For some reason I thought that Honts' hybrid was the mix of DLC and PLC but it really doesn't matter. I haven't read that stuff in years and was always concerned about over-emphasis. And by the way, I think you're smart not to use this technique on specific issue/criminal testing. It may work well there but that's not what is was designed for and we don't have good research on its use in specific issue/criminal testing where a subject has been questioned or interrogated extensively. To 'balance' such a test, one would probably feel an urge to make the DLC strong and that's a recipe for disaster. Good discussion. IP: Logged |
CHSBOY Member
|
posted 05-02-2005 08:11 PM
Boy, You're right (Barry). Two pages! I didn't notice earlier when I posted a comment following what I thought was the 'last in line', a response from CAPSTUN. Now checking in, I see that you had already commented on the Honts - Abrams affair and the DLC-PLC mix. Good discussion thread. Good night, folks.IP: Logged |
J L Ogilvie Moderator
|
posted 05-03-2005 06:11 PM
If my memory serves me right, I think Dr. Honts also did some research in which he concluded there is no such thing as "to strong of a control question".I am not sure I agree 100% but I wonder how difficult it would be to move the psych set of a quilty person to a comparison question? That doesn't seem likely to me under most circumstances. I do believe it may be possible during a child molest case if you use sex control questions that could include other unknown deeds and/or victims that may be more threatening to the subject. Also, remember, in the Utah test they do stim the controls between charts but they also mention the relevants so as not to focus soley on the controls. Jack ------------------
IP: Logged |
Lieguy Member
|
posted 05-08-2005 11:42 PM
Hey Barry;I have a lot of confirmed data....it's all in axciton format. Would you like me to send it to you or do you have enough already? Chip Morgan Boise Police Department 208-373-5488 chipm@cityofboise.org
IP: Logged |
Barry C Member
|
posted 05-09-2005 07:17 AM
I don't have enough yet, so I'd appreciate any confirmed tests with one two RQs (single issue) in them.Thanks! IP: Logged |
CHSBOY Member
|
posted 06-23-2005 07:47 PM
Does anyone know of any school teaching the use of DLCs in any test format, issue or screening (other than DoDPI)? Just curious.IP: Logged |
Capstun Member
|
posted 06-23-2005 10:51 PM
The Arizona School of Polygraph Science teaches it in Phoenix. Tom Ezell is the director. Here is their link: http://www.azpolygraphschool.com/ IP: Logged |
Poly761 Member
|
posted 06-27-2005 05:44 PM
I've chosen to use the MGQT for most criminal issues: 3. Were you involved in the robbery in any way? 5. Did you rob Johnson's Jewelers? 8. Do you know who robbed Johnson's Jewelers? 9. Right now, can you take me to any item taken during the robbery of Johnson's Jewelers? ALL Q'S BASED ON KNOWN EVIDENCE & PRE-TEST. Examinee may not have been present at robbery but assisted the person(s) committing the robbery. If not involved in the robbery, may have item(s) taken during the robbery (Q9). As designed, the ZOC deals w/a single issue and in my opinion is too restrictive: 5. Did you rob Johnson's Jewelers? 7. Did you rob Johnson's Jewelers this past Saturday? 10. (Evidence connecting) I've received the best results in a criminal exam with the MGQT. On occasion, I have used a ZOC as a follow-up to the MGQT. ..... IP: Logged |
detector Administrator
|
posted 07-03-2005 02:32 PM
Hey Everyone,Although I really love the discussion going on in this thread, it has gotten a bit spread out. Rather than starting new questions on the tail end of another discussion, please click the button to start a 'new topic' when it is logical to do so rather than appending a new discussion to an old topic. I'm going to close this topic for now. If you still want to respond to it, no problem, just start it as a new topic. ------------------ Ralph Hilliard PolygraphPlace Owner & Operator http://www.polygraphplace.com
IP: Logged |