|
Author
|
Topic: MGQT
|
Poly761
Member
|
posted 10-29-2005 08:46 AM
  
What is now considered the better or best MGQT format? What is the question sequence (1-10)?IP: Logged |
Barry C
Member
|
posted 10-29-2005 01:58 PM
I think it's a matter of opinion. There is plenty of evidence to show a format in which the RQs precede the CQs (e.g., the Army MGQT) the scores are more negative. Since it appears many examiners score with a bias against the truthful, adding that variable is probably not wise.Backster has a multi-facet test, the CPC has one (their "B Series"), the Utah test is an option, and my preference, the Air Force MGQT is out there as well. There's also the Navy CIS MGQT, the Secret Service MGQT, and more I'm forgetting, I'm sure. I prefer the AF MGQT because it's so flexible. You can use it with two, three, or four RQs, and you can even run it as a single-issue test. Off the top of my head, I think the order is as follows: 1) Neutral
2) Sacrifice Relevant
3) CQ
4) RQ
5) CQ
6) RQ
7) CQ
8) RQ
9) CQ
10)RQ You can chop off the last zone or two to do a short version. Compare, channel by channel, to the strongest adjacent CQ. A +3 in each spot is necessary for an NDI. A -3 in any spot is DI. Anything else is INC. [This message has been edited by Barry C (edited 10-30-2005).] IP: Logged |
armoredshoe
Member
|
posted 10-29-2005 10:49 PM
I have had great success with the AF MGQT.I much prefer it over the ArmyIP: Logged |
Poly761
Member
|
posted 10-30-2005 09:37 AM
Using the AF MGQT, is a 6 or 8 question exam permissible?In the "old" MGQT, I use + or - 2 or more in at least 2 of 3 tests to make a call. In order to make a call with the AF format I'm guessing we still need at least two tests? Is this accurate? END..... IP: Logged |
Barry C
Member
|
posted 10-30-2005 02:15 PM
I would always run at least three charts. Then total the spots and plug the numbers into the scoring cutoffs I explained above.You would never want to run more than four issues (or "facets" of the same issue to be a little technical) in an AF MGQT. If you do, you're making up some other test. You can run up to six in the Secret Service MGQT, but I think that's pushing it, and your accuracy is going to go down. If you want to run more issues - though I can't imagine why - just run two different tests (over six charts). I don't understand how you scored the "old" MGQT. Also, what are you calling the "old" one? IP: Logged |
Poly761
Member
|
posted 10-30-2005 03:20 PM
The MGQT I referred to is REID (Modified), relevants at 3,5,8 & 9. Primary relevant is at #5, secondary relevants at #3 and #8; and knowledge is #9. Controls at #6 & #10. An additional relevant was permitted as #11 but I seldom used this. I Don't recall running less than 3 tests for one (specific) exam. Two are required for an opinion. In the event there is a problem with 1 of the 3 I can ignore the problem chart for scoring or continue with a 4th test. What I do after the 3d test is determined by what occurred leading me to want to ignore/discard one of the charts for analysis. It appears we're stepping on our terminology a little. I define one EXAM as a minimum of two TESTS, same questions. As to scoring. I determine a "sub-total" by assigning a value for the pneumo, GSR & cardio of each question/spot during each test. The total for each question/spot is determined by the value of the sub-totals for all tests. For NDI/DI to a specific question/spot, the total would have to be +-2 or more in 2 or more charts. Anything else is INC. END.....
IP: Logged |
Barry C
Member
|
posted 10-30-2005 06:00 PM
That's the same as the Army MGQT, or the "DI test" as they called it in polygraph school. Reid uses inclusive CQs and the Army MGQT uses exclusive, earlier in time, CQs. Are you saying if you got a -2 in chart 1, a 0 in chart two, and a -1 in chart three for each spot, you'd call it INC? (Let's just pretend every spot worked out the same. I know it probably wouldn't.) I - and most here - would total the spot scores for a - 3 in each spot making the call DI. Actually, it the scores looked like this, we'd call the guy DI: RQ3 - +10; RQ5 +12; RQ8 -3; RQ9 +5. So, I'm still confused - or maybe I'm not - how you're scoring. IP: Logged |
Poly761
Member
|
posted 10-30-2005 07:33 PM
The scoring for a DI call to each relevant question (3-5-8 or 9) requires, at minimum, a score of -2 in 2 out of 3 charts. It is not a total of the 3 charts that results in an opinion. There must be -2 or more for each relevant in at least 2of3 tests/charts. END.....
IP: Logged |
skipwebb
Member
|
posted 11-02-2005 07:36 AM
I'm confused! What validated testing technique allows only two test charts (and therefore two askings of the relevant questions) as sufficient to make a determination of deception or non-deception?
Every validated testing technique that I am aware of requires three test charts.IP: Logged |
Barry C
Member
|
posted 11-02-2005 02:49 PM
Thanks Skipp. I thought it was me.IP: Logged |
Barry C
Member
|
posted 11-02-2005 02:55 PM
I should add, Matte and Backster have different cut-offs depending on the number of charts run, and some, I believe (Tedd, correct me if I'm wrong) will stop at two. I can't remember what Reid does. It's been too long, and I don't use it. I've never heard of the decision criteria here though.IP: Logged |
Ted Todd
Member
|
posted 11-02-2005 04:36 PM
Barry,You are correct. Backster says a DI or NDI call can be made on two charts.
2 charts > -9 is DI with two relevant questions. Two charts with three relevants is > -13 for a DI call. I do remember Tom Gray telling the class that three charts were preferred but never less than two for any call. Ted IP: Logged |
Poly761
Member
|
posted 11-03-2005 01:06 AM
I've been away from exams for three years and see the confusion I've created by my statements. The MGQT req'd 3 tests, 2 were straight through and 1 was mixed sequence. Barry - In response to your Q, 10-30 @ 6:00 PM, I would call spot totals of -2,"0" and -1 as "DI." I preferred to see -2 in at least two tests, especially if the GSR recorded the highest score due to ratio. I am lost with regards to the scores you list, para 2, line 5 of your post. END..... IP: Logged |
Barry C
Member
|
posted 11-03-2005 07:41 AM
In regard to those scores, what I was trying to say was if all spot totals except one were very positive scores and the other one was a -3 or less, the call would be DI. To illustrate, let me break the scores down in a hypthetical series of three charts:RQ3:
Chart 1 spot total: +3
Chart 2 spot total: +4
Chart 3 spot total: +3
Total score for that spot: +10 RQ5
Chart 1 spot total: +3
Chart 2 spot total: +5
Chart 3 spot total: +4
Total score for that spot: +12 RQ8
Chart 1 spot total: -1
Chart 2 spot total: 0
Chart 3 spot total: -2
Total score for that spot: -3 RQ9
Chart 1 spot total: +1
Chart 2 spot total: +2
Chart 3 spot total: +2
Total score for that spot: +5 The call would be DI based on the -3 score in RQ8. It's just as DI (unless it's a Utah or CPC test, but that's another discussion) as if every spot score total was -3 or less. Now you can argue all kinds of hypotheticals and say something's wrong with RQ8, but let's pretend they were all fine for the purposes of scoring and making a call based on the recognized rules. I hope that helps clear up my point.
IP: Logged |
J.B. McCloughan
Member
|
posted 11-03-2005 09:47 AM
Poly761,I am glad you are actively seeking the best methods to conduct your polygraph examinations and your participation here speaks of that. I have used the MGQT and the ZCT. I personally like the KISS principal when it comes to an examination. Because of this, I have come to conduct a BI-ZCT or You Phase whenever possible. This should be easily amendable to every criminal specific issue test. Although the BI-ZCT might not serve you initially in a screening examination, it can be used later to focus on a single area of concern. As you already know, deceptive examinations take care of themselves, for the most part. It is the truthful examinees that we are doing all of the work in the pretest for. You will have to work a whole lot harder using a MGQT than a ZCT to achieving this. Again keeping to the KISS principal, I would suggest using a ZCT or a BI-ZCT to ease an already intricate task. Whichever examination you choose to use, I would suggest that you research the current protocol for that examination, as you already know they change over time and with new research studies. Barry is giving you some of that information here. Barry, You have given a great textbook example of a basic polygraph school test on proper scoring. As you have already alluded to, I most likely would re-evaluate the examination with other acceptable analytical methods if those were the results, just to make sure. However, that is best left for another discussion, as I don’t want to add confusion here.
[This message has been edited by J.B. McCloughan (edited 11-03-2005).] IP: Logged |
Poly761
Member
|
posted 11-03-2005 09:52 AM
Barry -Thanks for the detailed explanation and I understand save for one statement: " - if all spot totals except one were very positive scores and the other one was a -3 or less, the call would be DI." I understand -3 in any of the 4 spot totals will be the basis for a DI. What I'm not sure about in your explanation is "-3 or less." What if -2 is a spot total and all others are solid positive scores? Thanks again..... IP: Logged |
Barry C
Member
|
posted 11-03-2005 10:03 AM
A -2 is greater than a -3. Numbers less than -3 would be -4, -5, -6, etc. You're confusing yourself by thinking like the negative scores are positive numbers.The above is good advice, and if you can, stick to a good single-issue test. Do a couple if you have to, just to keep up accuracy. If you have to use an MGQT I suggest you use a format that has a CQ for every RQ (e.g., the AF MGQT) in order to give the truthful a fighting chance. IP: Logged |
Poly761
Member
|
posted 11-03-2005 08:54 PM
Thanks to all of you for the clarification.I accept your comments about the examinee being administered this exam, but you believe the AF MGQT offers a better opportunity for the truthful? As I haven't tried this yet a -3 at any one spot for an overall DI call is very different. Is the same scoring criteria followed if a 6 or 8 question test is permitted? I reviewed the 61-page DODPI Test Data Analysis and I don't recall reading of a 6 or 8 question test/analysis. Thanks again..... IP: Logged |
Barry C
Member
|
posted 11-04-2005 08:31 AM
When you say 6 or 8 question test, do you mean 6 or 8 relevant questions? If so, there is no recognized 8 RQ test format. Using such an exam will probably give you a validity equal to that of CVSA, which is worthless, so I'd avoid it. I'd avoid the 6 RQ format for the same reason. When would you ever have 6 factest of the same issue anyhow? I have a hard time coming up with four. Arsons are about the only time I go searching.If you really want to do a six question version, go with the Secret Service MGQT because it is at least a format recognized by the polygraph community. I've used it as a screening test, but I got away from it because it asks too many RQs, and more than 5 RQs violates ASTM standards. Here's the format if you don't have it: 1) Neutral
2) Sac Rel
3) C
4) R
5) R
6) C
7) R
8) R
9) C
10)R
11)R
12)C Score it like you would the AF MGQT with a +3 or more as NDI, and a -3 or less as DI. Remember, with the DoDPI family of tests, a -3 or less in any spot means the whole test is DI. You don't call three spots NDI and one DI. (There are some who do, but that's a discussion we already had somewhere back a ways if you're interested.) IP: Logged |
Barry C
Member
|
posted 11-04-2005 08:35 AM
By the way, you can shorten the above test like the AF MGQT and run it with four RQs. Again, I like the AF because it has more CQs, which theoretically should favor the innocent.The rules for all of these tests are spelled out in the AAPP's examiners handbook, which is due for revision here shortly. IP: Logged |
Poly761
Member
|
posted 11-04-2005 10:29 AM
I'd be surprised if there was a 6 or 8 relevant questions format. In asking about the 6 or 8 questions I was referring to your statement on 10-29 @ 1:58, "You can chop off the last zone or two to do a short version."To shorten the AF MGQT by eliminating one zone (Q9 & Q10) I see this as an 8Q test (1-8). To eliminate two zones (Q7-Q8 & Q9-Q10) would it not be a 6Q test (1-6)? Are you considering the SR as 1 of the 4 RQ's when deciding to shorten the SS MGQT? Thanks again for your help & the add'l MGQT format? END..... IP: Logged |
Barry C
Member
|
posted 11-04-2005 10:54 AM
Okay, I get it now. So, yes, you score the same whether you have 4 or 3 RQs.The SR is not considered an RQ, so don't count it when determining how many RQs a test has. With the SS MGQT, you just chop off the last RQ RQ CQ and are left with a 4RQ (+the SR) test. When you do that though, you end up with what is - or even could be considered - a Utah multi-facet test, which is scored differently. (As I mentioned, we talked about that here somewhere, so if you have time and want to torture yourself, search around for the discussion.) IP: Logged |
Poly761
Member
|
posted 11-04-2005 11:43 AM
As "JB" indicated, I'll keep it simple. I like the format of the AF MGQT & I'll probably stick with this test. I like the ability to shorten this test as you explained & not having to worry about another (Utah) scoring system.Thanks again. IP: Logged | |
Polygraph Place Bulletin Board
