posted 06-18-2009 02:52 PM            

I
SR
C
R
C
R
R
C

The scoring was explained as a plus in all spots and a plus 6 is NDI. A minus 3 in any spot is a DI (or a total of minus 6 overall)

I called DACA and was told that the only ZCT they teach is the standard Federal ZCT.

posted 06-18-2009 02:58 PM            

------------------
Hól vagytok székelyek, e földet biztam rátok.
Elvették töletek,másé lett hazátok.
Vesszen Trianon !

posted 06-18-2009 03:10 PM            

Those are the Utah scoring rules. Notice they don't require a +3 in every spot to conclude NDI (which makes a NDI harder than a DI that only requires a -3 in any spot).

posted 06-18-2009 09:25 PM            

C1 R1 R2 C2

You would score R1, for example, to the strongest reaction (channel by channel) of either C1 or C2. That is, you could "jump over" R2 to use C2.

Keep in mind that it's channel by channel, so you could score the EDA in R1 against C2 if C2's EDA was stronger than C1's. Then, if C1's cardio was stronger than C2's, you'd score that component (channel, whatever) to the cardio in R1. You'd do the same for R2, meaning you could jump R1 to score to C1 if that CQ had the stronger component of interest.

posted 06-20-2009 08:50 AM            

You are correct. The Utah 3 question format (what we sometimes call the Utah Zone) is scored left, to the preceding CQ. However, the Utah 4 question format (what we sometimes refer to as the Utah MGQT), is 4 RQs bracketed by 3 CQs, and is scored to stronger bracketed CQ per each component.

r

------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)

posted 06-20-2009 02:23 PM            

The Utah MGQT (the test Skip listed) does go against the rationale for the Utah ZCT format to which you refer. The N C R question order is used to "bring a little balance" to the test. The N serves to allow the a person to be at baseline when the CQ is presented. The RQ follows and assumes, that for the innocent, there's less "reaction room" on the RQ because the examinee likely hasn't fully recovered from the CQ.

You can score any CQT with the Utah scoring system, and yes, they score to the stronger adjacent CQ. That means if you're scoring a Canadian A-Series or a Federal ZCT, for example, you'd score R5 to the stronger of C4 or C6.

You can run the Utah ZCT with two RQs (Charles Honts at NPA 2009) if you'd like, but as he asked, "Why would you want to give up data?" You can run the Utah MGQT with three RQs instead of four (as Skip posted) because it doesn't make sense to add a fourth sham question just to make the questions "fit" a rigid format. (Sometimes four RQs is too many, which is why we do the same with the AF MGQT.)

posted 06-20-2009 02:57 PM            

The point of it all is not to comply with some rigid rules, but to adhere to some validated principles.

The N C R sequence seems to attempt to make use of the law of initial values, by using the N to help ensure the examinee's physiology has returned to stable homeostatic tonic levels, through a process called allostasis, thereby improving the maximum potential response and benefit to a truthful examinee from the CQ. The N C R sequence in the Utah ZCT is scored to the preceding adjactent CQ, unless the CQ is artifacted and unusable, and then you can simply score to the closest interpretable CQ.

The Int SR N C R R C R R C N sequence of the Utah MGQT seems to violate the inherent princple of N C R and the law of initial values. However, you can, if you like, think of the opportunity to score the RQs to the Stronger bracketing CQs as attempting to make up for the assumed loss of response the CQ, due to the change in question sequence.

Decision rules differ for Utah ZCT and Utah MGQT. Utah ZCT decision policies are total score only (+/- 6), with no consideration for individual spot. Utah ZCT is a single issue test with 3 primary RQs. No fancy-schmancy unproven hypothetical stuff involving weak RQs, or evidence connecting things. The guy did it (the bad thing) or didn't. Simple. So, three RQs, that describe the examinee's behavioral involvement in the bad thing, and that's it.

Utah MGQT is an adaptation of the Utah ZCT. It is intended for use when a multi-facet format or multi-issue (mixed issue) format is needed. Utah MGQT decision policies are total score (regardless of mulit-facet or mixed issue), as long as all values are + or all values are - (ignoring zeros). If the values are mixed (some +, some -) then the Utah MGQT, uses individual spot scores. This may not make intuitive sense to polygraph examiners, but to statisticians and people who play with inferential decision models it makes perfect sense (it's good science). To us polygraph examiners, who read the actual questions, the RQs are distinct logical issues. Once the data are gathered, the distinctness or similarity in the the logic of the question language no longer matters to a statistician or decision theorist. What matters is whether the measured mathematical values are similar or difference. So, even if the behaviorally descriptive topics of the RQs are distinct, if the measured numerical values are similar it is sound and valid science to work with them all together. The decision theoric advantage of this will be more powerful classification efficiency (and that means better accuracy and fewer INCs to you and me.)

BTW, OSS-3 uses a similar solution when using decision policies designed for mixed-issue screening exams - except the algorithms replaces some procedural instructions with a fancy piece of non-parametric math called a Kruskal-Wallis equation. At APA 2007 we showed a poster session that demonstrated that these policies are significantly better than what we would observe or expect from traditional spot scoring rules.

Innit great having techniques designed by scientists.

r

------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)