|
Author
|
Topic: sacrifice relevant
|
detector
Administrator
|
posted 05-22-2009 05:26 PM
  
It is my understanding that the intended purpose of the sacrifice relevant is to allow the examinee to have an initial orienting response to the relevant issue so that the first relevant question sensitivity is not falsely heightened?Do we have research on whether this question fulfills its intended purpose? I guess a simpler way to ask this is...do we know if the sac Rel increases/decreases/stablizes sensitivity to the relevant issue? ------------------
Ralph Hilliard
PolygraphPlace Owner & Operator Be sure to visit our new store for all things Polygraph Related
http://store.polygraphplace.com IP: Logged |
pal_karcsi
Member
|
posted 05-22-2009 06:10 PM
In my practice I have found that if someone is reacting to this question, the result would be DI. Also I believe that the relevant sacrifice is still "a relevant question", no matter what the name is.------------------
Hól vagytok székelyek, e földet biztam rátok.
Elvették töletek,másé lett hazátok.
Vesszen Trianon !
IP: Logged |
detector
Administrator
|
posted 05-22-2009 06:21 PM
That would mean that the sacrifice relevant is not doing what it is supposed to do. If it IS a relevant question, we should be scoring it.What's happening to you in the real world application may be accurate, I guess I'm asking about its stated purpose and intent and any research done to test what it really does do or not do. ------------------
Ralph Hilliard
PolygraphPlace Owner & Operator Be sure to visit our new store for all things Polygraph Related
http://store.polygraphplace.com IP: Logged |
Barry C
Member
|
posted 05-22-2009 06:45 PM
I don't know that we've run a study in which we used a SR for some and not others to make the comparison. We do have data on techniques that use them and those that don't, and they all do about the same, which would seem to imply the SR doesn't do what it's intended to do. For example, the Reid Technique doesn't use an SR and it had an 88% accuracy. I suspect, but don't know without seeing and crunching the numbers, that it's not statistically different than the Federal ZCT or Utah ZCT (no difference in those when Don K. published his Validated Techniques article). The CPC test uses an "Introductory Stim" question in place of the SR, and it (the technique) is as good as any other CQT. Example: "Do you intend to lie to any question on this test?" There are a few thoughts on the SR though: One is it does what you describe. Another is that it helps the person habituate to the relevant issue. Another is that it does nothing, and yet another is that it make the truthful really nervous if they can count. That is, you ask him the same question four times in a test making them even more important (and salient) in his mind. With all that said, why are you trying to cause trouble? IP: Logged |
detector
Administrator
|
posted 05-22-2009 07:18 PM
LOLThanks Barry, I needed that chuckle. I'm asking for two reasons. 1. During the last GPA meeting it came up in discussion and I remembered something about the habituation part you mentioned...meaning that it could be doing more damage to the issue-innocent examinee. I wanted to understand the theory and any study behind that. 2. I had this happen in a test yesterday. Chart 1 clearly good responses to controls, chart 2 a little less, by chart 3 relevants and controls basically equal. Had I kept going I think it would have ended up DI, but I don't think that would have been accurate. Oh and examinee had major response to the sac rel each time. OSS included it in the rank scoring and it was #1. On that note, Ray, if you are lurking, is OSS3 supposed to be measuring the SR? I just picked up on this recently. Do you think I have a setting in my questions incorrect? I'm on Limestone and I was running a 2RQ AFMGQT. On what mechanism does OSS3 determine the question type?
------------------
Ralph Hilliard
PolygraphPlace Owner & Operator Be sure to visit our new store for all things Polygraph Related
http://store.polygraphplace.com IP: Logged |
rnelson
Member
|
posted 05-23-2009 07:51 AM
Ralph,OSS-3 should not score the SR - even if it is labelled SR. A niggle with some previous algorithms is that they would seem to use the question tags to parse the CQs and RQs, and any question with an R in the tag might be considered an RQ. Lafayette writes their own code for their OSS-3 tool. We simply provide whatever technical assistance we can. As far as I can tell, they use meta-data included in the question editor when you declare the question type for each question. As a results, the Lafayette tool seems to correctly understand that the SR is a sacrifice and not an RQ, and does not score it. The Limestone system regards the question tag itself as the meta-data. Current versions will correctly ignore the SR questions. The Rank algorithm, in the Limestone system is not (strictly speaking) OSS-3, but is a replication of the Relative Response Magnitude equation described in Raskin et al (1988). The tool will also calculate the rank-order algorithm described in Miritello (1999). That tool includes an option to score selected types of questions or all questions, and to score selected components. An interesting way to study the contribution of the SR question to scored results, would be to use a confirmed case sample that includes SR questions, and compare the results while including and excluding the SR in the analysis. It would be great to have a sample of 100 or so confirmed ZCT cases, for which we have published results for several different algorithms - like OSS-3, PA (replication), and even the HSS. Such a sample could even be scored with OSS-2,which continues to impress with its ability to provide robust results in its intended use. It just so happens that we may have such a sample. All it requires is some spare time to re-measure all the data because I didn't retain the SR measurements the first time I measured it. After that we could bootstrap a comparison of the results with and without the SR, for OSS-3 and the other algorithms. r ------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)
IP: Logged |
Barry C
Member
|
posted 06-03-2009 05:29 PM
I've been slow on getting back to this, but better late than never. Dr. Frank Horvath reminded me that he did a study on the SR and found it didn't seem to do what it was intended to do. It was a lab study, so it had some limitations, but he didn't see what was predicted, and there was some evidence the SR was less than helpful, shall I say.It's in an old Polygraph journal, and I'll try to get the cite for those who want it. It's good to go back and look at those things now and then. I'd read the study some time ago, but I had forgotten it until he wrote me. IP: Logged | |
Polygraph Place Bulletin Board
