Talk:Iowa gambling task

Old talk

In addition to dubious benefits of making available a cheat sheet for the test, the summary image of the contingencies attached to each deck isn't correct. In the first row the gain is not per deck it's per 1 card selected from each deck. The bottom row cannot be the number of rewards per 10 cards as the task rewards every card drawn and this value is therefore 10 for every deck. It looks to me that the numbers in this row represent the frequency of punishments. I'm going to remove the image from the article until someone wants to create a fixed version. I would do it myself, but I don't think i agree with giving the game away. 82.35.27.34 (talk) 11:35, 24 September 2008 (UTC)

Brilliant work on making this article and the picture is very good. I have one slight reservation however. I don't think we should give away the structure of the task. I originally linked to the IGT off of some edits I made on orbitofrontal cortex but didn't write it until I'd had clearance from Antoine Bechara to use a screengrab from the test (included). In it I asked for his comments on the OFC article which he said was fine, and he asked that we keep the structure of the task opaque. Similarly whilst I think the WCST is now much better I worry that by giving the game away we could impact upon the validity of others' research. I know you can find out the structure of both tasks other places, but this makes it VERY easy to find out. Bring on the flames... --PaulWicks 13:56, 7 April 2006 (UTC)

This page should have a warning as it basically destroys test integrity in anyone that may need to take it! I am a compulsive hoarder, and based on PubMed articles I came to look here, as they notice reduced scores on this. This is the test I did not have during a neuropsych exam, and now that I know how it works it would be of little use to me or my doctor. 209.226.121.18 07:39, 8 November 2006 (UTC)

I agree with the above and have revised the article accordingly.--Brad Patrick 03:20, 16 February 2007 (UTC)

Paul and others, you are spot on about this issue. I use the IGT in my research and it is pretty much essential to keep the task contingency opaque for naive participants, otherwise, what's the point? I guess providing the information for the masses is more important than the integrity of the science and psych assessments this task is used for. 6th April 2007.

Keep in mind that most people will not think of looking for this test, nor will they remember the contents of the test. Still I see that the opaqueness was promptly reverted. Maybe this should be sent to arbitration as was suggested by Nestify? -- KarlHallowell 18:03, 6 April 2007 (UTC)

Up to you really, karl. The contents of the test are quite easily remembered, I had a patient that we hoped wouldn't remember doing it just a few months previously, and he pretty much picked it up after several trials. There is also a learning effect with repeated task presentation in controls, and even in an amnesic patient [e.g. Turnbull & Evans, (2006) Preserved complex emotion-based learning in amnesia, Neuropsychologia, 44, 300-306] 21st August 2007 —The preceding unsigned comment was added by 86.132.231.13 (talk)

To rephrase a famous saying by the philosopher Santayana: those who don’t remember the experiments of the past are condemned to repeat them. It may be argued that the IGT is not new. Almost certainly Damasio and Bechara were unaware of the following very similar experiment, but the conclusions are markedly different.

Neal Miller’s Somatic Marker

In 1935, the psychologist and learning theorist Neal Miller conducted the following experiment.

To human subjects he presented in unpredictable order the symbols T (followed by electric shock) and 4 (not followed by shock). The shock was followed by a large galvanic response (GSR) that was soon conditioned not only by seeing the symbol T, but by anticipating it. From this and subsequent experiments Miller concluded that organisms should "behave 'foresightfully' because fear (i.e. anxiety), would be mediated by cues from a distinctive anticipatory goal response." Miller further concluded that the 'learned drive' of fear or anxiety, as marked by the GSR, obeys the same laws as do overt responses".

In other words anticipatory somatic changes mediate not choice, but avoidance, and may be described fully by learning principles.

Compare this experiment to Damasio’s IGT experiment, where an individual again is confronted with a succession of symbols (in this case, markings on a card), and with unpredictable aversive consequences and similar changes in an equivalent somatic measure, the SCR. In this case the unpredicted aversive consequences were large negative card values occurring from time to time. If it is assumed that unexpected 'bad information' is painful as well, then both experiments assume equivalence.

The difference between both experiments is not in their structure, which are more or less equivalent, but rather in the interpretation of the role of the galvanic skin response as a dependent measure. Specifically, the GSR for the Miller experiment correlated with a subjective response that was interpreted as anxiety or fear. For Damasio, the subjective response to arousal as marked by the SCR was subtler, or a mildly or non aversive ‘gut feeling’. In other words, if one assumes that the level of arousal was higher for Miller’s subjects than Damasio’s, the level of arousal could lead to distinctly different interpretations as to the role of arousal. Thus, it is easy to see how Miller assumed that tension based arousal (or anxiety) mediated avoidance, and why Damasio assumed that arousal mediated choice. In other words, if turning a bad card in the IGT experiment signified not a loss of play money but a loss of real money or a painful shock, then avoidance and not choice would have been a more likely interpretation.

So we are left with the original question: what is the role of autonomic arousal? If arousal is dependent upon learning, as both Miller and Damasio hold, what is its function: avoidance, choice, or some mixture of the two that is dependent upon the level of arousal?

One way to ascertain the role of avoidance is to simply examine whether elevated autonomic arousal occurs under response contingencies that either eliminate the ability to avoid or obviate the need to avoid. If results under a response contingency are all bad and unavoidable, then we have Seligman’s learned helplessness, and if all results are good and thus create no need to avoid, then we have Csikszentmihalyi’s ‘flow’ response. Both are marked by low autonomic arousal as marked by a reduction in SCR and a corresponding lack of reported tension, anxiety, or fear. Thus it may be construed that avoidance is an essential function of autonomic arousal. This of course does not directly challenge Damasio’s position, but raises the avoidance hypothesis front and center as an alternative explanation for his findings.

a. j. marr

Source:

Miller, N. E. (1971) Selected Papers, Atherton, Chicago

pp-123-171

Clarification needed in section "Task structure"

• They are told that each time they choose a card they will win some game money.
• Every so often, however, choosing a card causes them to lose some money.[clarification needed whether they know this or must find this out]
• The decks differ from each other in the number of trials over which the losses are distributed. [clarification needed whether they know this or must find this out]

Thus, some decks are "bad decks", and other decks are "good decks", because some will lead to losses over the long run, and others will lead to gains.[clarification needed whether they know this or must find this out]

Of course, extremely straightforwardly logically thinking, only the first phrase says "they are told". However we all know that wikipedia is written sloppily (and being read by not very logical people), and people may have an impression that "They are told that..." refers to the whole subsequent section. Therefore the description must directly say what exaclty the "lab rats" know and what is the hidden plot of the eggheads. Staszek Lem (talk) 00:21, 19 October 2016 (UTC)