Psychological positive feedback effects.
Assigned:
- Charles G. Lord, Lee Ross, and Mark R. Lepper. 1979. “Biased assimilation and attitude polarization: The effects of prior theories on subsequently considered evidence.” Journal of Personality and Social Psychology. 37(11), Nov 1979, 2098-2109.
- Robert Rosenthal. 1980. “This Week’s Citation Classic: Pygmalion in the Classroom” (Commentary). Current Contents (Jan. 2): 156.
- Maria Elena Lara and Daniel N. Klein. 1999. “Psychosocial Processes Underlying the Maintenance and Persistence of Depression: Implications for Understanding Chronic Depression.” Clinical Psychology Review 19(August): 553-570.
- Christopher Peterson. 2000. “The Future of Optimism.” American Psychologist 55(1): 44-55.
- John T. Jost and Orsolya Hunyady. 2005. “Antecedents and consequences of system-justifying ideologies.” Current Directions in Psychological Science, 14, 260-265.
- Film:The Lathe of Heaven (screened May 2, ZDVD 1684)
Other (note: citation does not constitute endorsement of content):
Entries (RSS)
May 9th, 2007 at 12:11 pm
Making a critique of this week’s readings has proven itself to be difficult for me, as I take many of the concepts in the assigned readings for granted, probably because I have studied these and related theories in my social science courses; I look forward to hearing your opinions about problems with these ideas.
One idea I found to be a common idea throughout the readings is the presence of positive feedback loops based on how we as humans have come to process information. As we (as humans—or at least as white middle class Americans who were the initial focus of psychology research) strive toward consistency in our beliefs and actions, feelings of cognitive dissonance result when we come to find we are inconsistent, and thus we rationalize away our inconsistencies. The Lord, Ross & Lepper paper, for example, shows that our need to be consistent will cause us to interpret new or conflicting information in a way that affirms our preexisting belief system, while the Jost & Hunyady paper illustrates that our need to believe our current system/beliefs is/are right that we will avoid change even if it is in our best interest (and in fact, we will act counter to our best interest in order to remain consistent). Ross & Ward did another series of studies developing the notion of Naïve Realism, which illustrate that we will come to believe that people who are more polarized in one way or another than ourselves are extreme and irrational; and, interestingly, they found that most people come to share moderate beliefs but the way they process information has caused people to be unable to rationally discuss these beliefs. One common example of the Naïve Realist idea is that we see anyone who drives faster than ourselves as a maniac, and anyone who drives slower is too slow. I believe that this process of biased assimilation, as well as naïve realism and system-justifying ideologies, can be seen as positive feedback loops as they cause us to become increasingly extreme or closeminded in how we come to see the world. The Lara & Klein article incorporates this information processing idea, in that the more depressed individuals reach out for support, the less support they tend to receive, and the more depressed they become—and the depressed individual’s need for consistency would not allow her to recant her need for support or to go elsewhere for this need. The human need for support outweighs any rational processing that would allow either party to objectively recognize what is happening, furthering a feedback loop of reaching out and rejection.
I don’t think that the optimism paper necessarily applies here, but perhaps the interaction between student and teacher in the Rosenthal paper suggests, like the depression article, that other’s expectations of us influence our responses, in what also appears to be a continuous cycle.
May 9th, 2007 at 4:04 pm
I think that an interesting question that follows from many of the readings this week, but in particular from the Lord, Ross, and Lepper paper, is the question of what we ought to believe, or how we should go about believing things, in light of the fact that we process information in such a biased manner.
In the section “The Normative Issue”, Lord, Ross, and Lepper write that there is nothing inherently wrong with biased information processing — in fact, they write, “Willingness to interpret new evidence in light of past knowledge and experience is essential.” What they think is the “sin” of the subjects in the experiment is a kind of circular reasoning: using an accepted theory to process information in a biased way, and then using that information to support the theory. In other words, it’s OK for the subjects to process the new evidence about the death penalty in a biased manner, but when they then use that biased information to evaluate the bias in the first place, they go wrong. However, while doing so will certainly lead to positive feedback, divergence of opinion, and possibly unfalsifiability, it’s also not clear that we have any alternative to this procedure — how could we fairly test any hypotheses of which we have any opinion of if this opinion is going to color any tests we make? It seems like the answer is that we can never have really fair, objective evaluations, although if we introduce some negative feedbacks, such as consciously trying to be as evenhanded as possible, we can be fairer.
Another important issue is how are we to act on our convictions, and how are we to engage in dialogue with “nonbelievers”. The paper implies that rational discourse might not get us anywhere, in which case we might have to resort to nonrational methods of persuasion, such as the ones shown in _The Persuaders_. But this seems highly unethical — especially if it turns out that your convictions are the wrong ones. It seems that there is a choice between being effective and being reasonable/ethical, which is an extremely unfortunate dilemma. If we stop to ask reasonably, “What is the right thing to do?”, we most likely will be very ineffective and the right thing won’t get done. But if we don’t stop to reason, while we may manage to be effective, we won’t know whether our end is really good to begin with.
May 9th, 2007 at 5:05 pm
Looking back over the Lord, Ross and Lepper paper (since people seem to be attaching themselves to this one), I wonder how much of a bias in one direction people actually need to perceive information in such a biased manner. I could certainly imagine (and I would imagine the authors would agree) that the more polarized you already are, the more polarized you are likely to become, fitting in nicely with our course’s theme. However, they don’t address a much more subtle issue, which is what happens on the low end of the spectrum. If you are presented with a slight distributional bias in one direction, how quickly do you make that your own bias? This seems like a far more essential question from an evolutionary perspective (predators etc), and would inform the level at which we become biased processors of information.
I could imagine two experiments that would usefully examine this. The first experiment would be relatively simple: present people with a sequence of images, some largely blue, some largely red. Tell them they are asked to classify them into “red” and “blue”, flashing the image on the screen for a brief period and then removing them. They’d then be asked to class the image as “red” or “blue.” Now, the trick is that the images will be distributionally biased at the beginning to either red or blue, and then would become increasingly uniform, so that by the end of the experiment, the data would be at or near uniform on the whole. Then ask them to estimate the make up of the samples. If people are quick to create a bias, they would select the one that was more popular. Second, we could also look at reaction times. If they have created an implicit “motor-memory” bias, then they should react more quickly to the one at the beginning than to the other, since they will be expecting to press “blue” more often. If this becomes more exagerated even when the data is balanced, then we will confirm the Lord, Ross and Lepper hypothesis, even extending it to people that do not have a strongly held bias before the experiment starts.
Indeed, I would imagine the experiment listed here would have an even stronger motivation than theirs: quick object recognition ought to be a much more evolutionarily important trait, and forming quick biases (say, that thing is an enemy) is far more essential (and far more basic) than an evaluation of two conflicting arguments. Moreover, one could control for the amount of bias that people should construct, unlike their study, where they have varying biases (and varying evaluations of those biases…).
A second experiment would have the same set up, but would always have a balanced distribution of reds and blues. Instead, a confederate would confide in the subject that there is a certain bias for red or for blue.
This experiment would not suffer from any problems with the primacy effect (but might suffer other failures…). We would expect the exact same results, though.
May 9th, 2007 at 5:31 pm
Like Siobhan, I too find many of these ideas to be reasonable and familiar. Reading the blog posts thus far, I have little to add to their already treatment of the readings. What I will say something about it the “negative feedback” aspect of these social phenomena. As Jack and Siobhan already observed from the reading, the heuristics we apply in information processing are often necessary and good. But again it is only in some cases that their influences go awry: not all people are clinically depressed, or hold concretized beliefs. (One thing to ascertain would be if the relative number of people falling victim to these unchecked positive feedback thinkings is changes with respect to time, age, gender, location, anything.) So clearly for at least some people there are sufficient negative feedback influences in place to keep their thoughts lucid. It would be interesting, I think, to discover if these negative feedback mechanisms are natural, how they occur; and, furthermore, to investigate whether or not they can be reintroduced in cases of people who have lost them. If you’ve ever had a depressive friend, you’ll know that trying to convince them out of their self-defeating spiral of thoughts can be virtually impossible; but sometimes with serious consistent effort, or at the hand of a miraculous epiphany, their thinking will suddenly reverse. This is intriguing to say the least. I think understanding the negative feedback influences involved in information processing and opinion formation would be very fruitful. In closing, I think it is alright for me to plug education as an institution. Education systems serve to condition their subjects. I think utopian education would work to teach its pupils to deal with such human natures, directly or indirectly, so that they are in a position to transcend these heuristics, when necessary. So that they don’t get stuck in these loops. This, I think, is the importance of teaching towards critical thinking, skepticism, and self-realization. This is one of the most important means of empowerment.
May 9th, 2007 at 5:34 pm
I think that this week’s topic is my favorite, which makes me especially disappointed that I will be unable to make it to class tonight. I have a CS107 midterm from 7:00-10:00, so unfortunately, I will have to see all of you next week. In any case, though, I find the idea of fluctuations in mood being cyclical incredibly powerful. In fact, I would like to do my final presentation on positive feedback as it relates to optimism, or more specifically, psychological health in general (preferably on the 23rd). Alternatively, autoimmune diseases are interesting, since they are self-compounding and also dependent on stress levels and mood.
The reason why this idea fascinates me is outlook seems incredibly subject to individual control. Seeing the world through rose-colored glasses requires no resources and marginal genetic inclination. It is merely a conscious decision. The benefits that Peterson cites in his article about optimism are incredible: success, popularity, and even good health can all stem from positive self-esteem or confidence in other people. Knowing this, it’s difficult to justify the intellectual culture of belief in corruption and subtle insidious influences in society.
Delving deeper into the article, I was especially interested by the juxtaposition of pessimism and optimism.. As the article states, “The optimism and pessimism items in Scheier and Carver’s (1985) LOT prove somewhat independent of one another […] Some people consider both good and bad things to be plentiful.” I think in order to assess this claim, we have to go back to the idea of big versus little optimism. Big optimism takes overarching convictions of goodness and success and doesn’t really specify how these ends will be reached. Little optimism, on the other hand, takes a specific event and indiscriminately assigns a positive outcome to that event. Little optimism, to me, doesn’t seem like optimism since little pessimism and little optimism can co-exist. I can say “SymbSys 205 discussion will be really engaging tonight” and “I’m concerned about falling off my bike on the way there,” without any conflict. However, I can’t say “I think my night will go well,” and still think both of these things. Thus, although it’s more difficult to study, I think that big optimism is more pervasive in its positive effects and more interesting to consider (especially since its basis in reality is more difficult to determine).
Regarding the more popular article about belief bias, I have always been skeptical about belief bias claims. Just because mixed evidence strengthens one’s opinions does not mean that they are disregarding contrary evidence. Clearly, the reason why people develop viewpoints in the first place is because they find one side more compelling than the other. Even if they have carefully read and reviewed all the available evidence, it is possible that they could come to the conclusion that the literature available for one side of the issue has more merit, more credibility, or is simply more persuasive. Naturally, then, being exposed to this disparity again will reinforce the validity of this conclusion.
However, article managed to address this concern and the evidence presented in the associated study was striking. Since the experimental methods were not detailed on either of the index cards, I was shocked that people could conclude anything from these cards, much less designate one piece of anecdotal evidence as more compelling than the other. You could say that this article overcame my own belief bias with forceful enough claims to change my opinion.
Have a great discussion!
May 9th, 2007 at 6:17 pm
Like most of the respondents above, I was fascinated by the Lord, Ross, and Lepper paper. The results of this paper are very problematic for the philosophy and rationality of science. I guess biased assimilation is at the heart of the debate between Kuhn and Popper in the philosphy of science. Whereas Popper would say that science advances by theories and falsifications, Kuhn would reply (and Lord, Ross, and Lepper would verify) that once a theory is held it does not get falsified. A scientific community immersed in a particular paradigm (or strongly beliving a particular theory) will interpret all evidence as confirming. According to Kuhn, theories are not challenged, but rather the communities around them wither and die out as new practitioners are drawn more towards other theories.
I guess the most important part of understanding psychological positive-feedback loops how they will influence our actions regarding them. For example, what implications does the Pygmalion effect have for education? Should we pair students up only with teachers who believe in their success? What implications does positive-feedback in depression have for it’s treatment? How do we break the cycle of negative memories and ruminations? How have people who’ve recovered from depression managed to break the positive-feedback loop? These are all interesting research directions once the presence of a positive-feedback loop has been discovered.