Talk:Ambiguity effect

Untitled

Is the example given the best one of the ambiguity effect? Although the means are the same, many would prefer option X due to the shape of their utility of wealth curve. In X, there is a certain win, but Y (although it could have double the win) has the possibility of no black balls at all.

Fullel (talk) 20:58, 29 February 2012 (UTC) Elliot.

Exactly my thoughts. Both options are equivalent in terms of expected value, but X has less variance. As presented, since both options are "worth" the same amount of $, this isn't really the best example of a cognitive bias as no "wrong" decision is being made...Jamesa7171 (talk) 08:27, 30 April 2012 (UTC)

Is the one provided an example of the effect at all? In the analysis described the probability of all the events is known (that's how the outcomes having equivalent mean is known) -- I fail to see the ambiguity presented. — Preceding unsigned comment added by 196.27.25.141 (talk) 14:39, 19 February 2013 (UTC)

Probability

I have a problem with the wording "the probability is unknown". That is not the case here, at least not unknown in general. It might be unknown to the person in the experiment because the probabilistic modeling of that experiment is more complex in one case than the other. — Preceding unsigned comment added by 87.231.106.131 (talk) 07:41, 8 April 2014 (UTC)

I agree with you that, how this passage is phrased, it's not really clear if/whether probabilities are known or unknown, and by whom. The idea is that the probability of a black ball is somewhere between 0/30 and 20/30. Although the urn "experiment" was initially conceived as more of a thought experiment, we can consider "real world" analogue by suggesting that the experiment was set up as a double blind test, where some computer program randomly chooses a number between 0 and 20, puts that many black balls in an urn, then adds '20-number picked' white balls and 10 red balls. The probability of choosing a black ball is truly unknown, though it clearly exists. 2601:19B:B00:C7B2:B90C:D072:A6C7:1128 (talk) 16:29, 15 January 2018 (UTC)

Example

I question the use of this example for two reasons:

First, it fails to demonstrate ambiguity. The total number of balls is fixed, and thus the probability is fixed. What information is given to or withheld from the test subjects in order to create ambiguity?

Second, the details of the example are nonsensical in that they suggest a fixed total number of balls can contain two subsets of balls, i.e. red and black, that are equal in probability of being drawn but not equal in number. If there are 10 red balls in a pool of 30, then the number of black balls in the same pool must also number 10 or the probability is not equal. The example fails to mention a second pool of differing total size where black can exist in equal proportion to the 10 red balls but in a different quantity. Vikingnoise (talk) 19:34, 26 February 2015 (UTC)

Taking each point you raise one at a time-

First, what's withheld is the exact proportion on black and white balls. The subject knows these 20 balls are distributed between black and white, however the exact number of each color is not revealed. While the probability is fixed, given that it's not known ex ante, the subject must consider it an unknown variable.

Second, the trick in understanding how the two outcomes can be the same in probability is by considering the information set the subject has at the time of the decision. The probability of drawing a red ball is 10 out of 30- that's known with certainty. The probability of drawing a black ball depends on how many black balls there actually are. Using DeMorgan's laws, we should consider each possible "state of the world" (that is, each possible combination of white and black balls), consider the probability that state is the "true" one, and then consider, in that state, what is the probability of drawing a black ball. Since the number of black balls was randomly chosen in the interval [0,20], we should consider each case in turn. After considering what the probability of drawing a black ball is when there are 0,1,2,...,19,20 black balls in the urn, and weighting each one by the likelihood that there are, in fact, that many balls in the urn (which is a 5% chance any given number is the "true" one), we see that the unconditional expectation is also 10/30. 2601:19B:B00:C7B2:B90C:D072:A6C7:1128 (talk) 16:39, 15 January 2018 (UTC)

Dr. Villeval's comment on this article

Dr. Villeval has reviewed this Wikipedia page, and provided us with the following comments to improve its quality:

The article could mention references related to the experimental tests of this ambiguity effect in economics.

Suggested references: Becker, S., and F. Brownson (1964): What Price Ambiguity? Or the Role of Ambiguity in Decision Making,Journal of Political Economy, 72, 62-73. Camerer, C. F., and M. Weber (1992): Recent Developments in Modelling Preferences: Uncertainty and Ambiguity,Journal of Risk and Uncertainty, 5, 325-370. Colman, A., and B. Pulford (2007): Ambiguous Games: Evidence for Strategic Ambiguity Aversion,Quarterly Journal of Experimental Psychology, 60, 1083-1100.

Ivanov, A. (2011): Attitudes to Ambiguity in One-Shot Normal Form Games: An Experimental Study, Games and Economic Behavior, 71, 366-394.

We hope Wikipedians on this talk page can take advantage of these comments and improve the quality of the article accordingly.

Dr. Villeval has published scholarly research which seems to be relevant to this Wikipedia article:

• Reference : Li Hao & Daniel Houser & Lei Mao & Marie Claire Villeval, 2014. "A Field Study of Chinese Migrant Workers' Attitudes Toward Risks, Strategic Uncertainty, and Competitiveness," Working Papers 1047, George Mason University, Interdisciplinary Center for Economic Science.

ExpertIdeas (talk) 03:08, 25 August 2015 (UTC)

Dr. Alevy's comment on this article

Dr. Alevy has reviewed this Wikipedia page, and provided us with the following comments to improve its quality:

To some extent this review confounds risk and ambiguity aversion and it certainly jumps too rapidly from topic to topic. There is a broad range of theoretical and experimental literature that distinguishes risk from ambiguity, that places ambiguity in an axiomatic framework, and understands how ambiguity effects outcomes in both individual choice and market settings.

Further the notion that ambiguity aversion is 'irrational' (that is that it results from a cognitive bias') does not reflect where I think the modern literature is on this topic (see e.g. Gilboa, Postelwaite, and Schmeidler, 2009; Is it always rational to satisfy Savage's axioms?; Economics and Philosophy) and (Gilboa, 2015; Rationality and the Bayesian paradigm; Journal of Economic Methodology).

Trautmann and van de Kuilen survey experimental findings in The Wiley Blackwell Handbook of Judgement and Decision Making (2015); Chapter 3 'Ambiguity Attitudes')

We hope Wikipedians on this talk page can take advantage of these comments and improve the quality of the article accordingly.

We believe Dr. Alevy has expertise on the topic of this article, since he has published relevant scholarly research:

• Reference : Jonathan E. Alevy, 2011. "Ambiguity in Individual Choice and Market Environments: On the Importance of Comparative Ignorance," Working Papers 2011-04, University of Alaska Anchorage, Department of Economics.

ExpertIdeasBot (talk) 17:05, 27 July 2016 (UTC)

Another, more real possibility, perhaps, of why people choose options that seem safer

When there are more variables, there is more chance that someone is being cheated. This is obvious in the mortgage example - and the possibility of someone manipulation data to get the result they expect or want in the ball example, is a possibility - that people may have in the back of their minds (subconsciously?). There should be an experiment where the balls are, as far as possible, insured to be fairly distributed. And their should be an experiment where there is an advantage to choosing the option of having the unknown distribution . . . to have it pay off with an odds benefit. It would be interesting to see how much the odds would have to be increased to change the choice and how much the choices change with increasing odds. For example, if the possibility of non-equal distribution was in play, would an a 5% advantage cause people to choose that option? There are other things that could be done too. It's kind of sad that our societies' banking scandals have made the other example so painful. Benvhoff (talk) 10:17, 21 August 2016 (UTC)

An issue with the judgement of ambiguity aversion being a cognitive bias and heuristic

Regarding the choice between fixed and variable mortgage rates, it does not seem clear to me that this trick is against rationality (which a heuristic is defined can be).

Rationality in Peterson (Introduction to Decision Theory) is explained to be a concept on the normative optimality of a decision considering the information given at the point in time at which the decision was made (Peterson, 2009, p. 4).

Thus, it seems to me that the ambiguity effect (ambiguity aversion) as a decision rule (to delete outcomes for which probabilities are not known), a transformative decision rule if you will, is perfectly rational, since it does not make sense to consider the option of variable rates, since one does not know the probabilities of those variable rates. It is Knightian uncertainty which truly insufficient to be used with the (expected) utility maximisation calculus in classical economics. — Preceding unsigned comment added by SebastianGuy (talk • contribs) 17:55, 30 June 2021 (UTC)

A small error (or ambiguity) in probability, .. and own digression..

The probability of winning in the game Y is given at 1/3, the same for the game X, in fact this is only true if there is at least one black and at least one white ball in the urn. If we consider the possibility of zero or twenty black balls( 0 to 20 instead of 1 to 19) in the urn then the probability of picking a black is 0.35. It is counter intuitive, but without doubt. I checked with my old faithful Microsoft Mathematics and was surprised not to find 1/3 in the first place.. The author of the article wrote about equal probability between 0 to 20 black, where 1 to 19 would have been right to have a 0.33 win probability . It's even more uplifting at 0.35.

Another reflection : The riddle is very akin to 'would you play a game where you have 1/3 chance to win', ' would you play a game where you have 2/3 chance to lose'. Generally peoples tend to answer more positively at the first question prefering positive expectations. If you put the two propositions in the same phrase and ask for a choice... hopefuly they'll see the joke ! Not sure 100% will.. 2A01:CB14:4D6:A400:3C90:160:79B6:6E7A (talk) 14:38, 5 June 2022 (UTC)antSor