Wednesday, April 28, 2010

Or we can overinterpret scientific findings

And apparently, we can do it in the other direction as well. As the folks over at Freakonomics illustrate, when the science seems to agree with our instincts or politics, we can take a very complex situation and oversimplify it to make a point that we would like to be true, whether it is or not. In their example, the relationship between environmental causes and the disappearance of honeybees is tenuous at best. But there are many environmental groups that claim it is clear as day - because this advances what their hearts believe to be true.

It is not just environmentalists that make this error. Many domains are very complex. Economies, ecologies, political systems, religions . . . . In most cases, there is not a direct link between any one input and one output. But if we WANT it to be true, we can find some scientific finding that we can take out of context and use to make our preferred point.

The general public just doesn't get science

Another example of how valid and reliable scientific evidence is often discounted or ignored by the general public.

A new study by researchers at Tulane University found that spanking leads to increased levels of aggression in children. They controlled for a "host of" factors such as depression, alcohol and drug use in the mother, natural levels of aggression, and more. They followed the childrens' behavior over many years, several ages, and 20 US cities.

And then in an interview, some TV talking head says the study is BS because he was spanked and he grew up just fine. This is a sample size of just 1, has no controlling factors, may not even be true (maybe he would have been even less aggressive if he hadn't been spanked), etc. etc. But I am sure that half the audience just nodded their heads and agreed with him. The insights provided by the science are wasted.

I guess we get what we deserve, don't we?

Tuesday, April 27, 2010

Power and hyporcrisy

Dan Ariely points to another great piece of research (full paper PDF linked on his blog). He describes a study that provides more evidence that our ethics are much more malleable than we think. Basically, this research finds that when you are in a position of power, your ethical expectations of yourself get easier and your ethical expectations of others get stricter. It further found that when your power came illegitimately, the opposite happens.

This illustrates what is commonly referred to as a licensing effect. It is the same as when you order a healthy meal at a restaurant and then reward yourself with a big dessert or side dish. Our logical/rational brains may want to do what is right, but our emotional brains feel entitled to maintain a balance. So if your power is legitimate, you are entitled to more unethical deviations than others are as a "balance." But if your power is illegitimate, you are entitled to even less unethical behavior than others - also to "balance" the power.

What I find very interesting is that the researchers were able to induce the effect just by getting their participants to imagine stories and situations where they had legitimate or illegitimate power. It doesn't take much to get these effects going. This reminds me of the study I blogged about recently where even imagining that you order a healthy meal gives you license to order dessert. I guess it is pretty easy to fool ourselves into doing what our emotional brains wanted all along. Face it - we are suckers.

Tuesday, April 13, 2010

imaginary licensing effect

The licensing effect happens when individuals engage in something that boosts their self-image (giving to charity, dieting), they are more likely to engage in something self-indulgent subsequently. And vice versa (being self-indulgent leads to subsequent self-image boosting behaviors).

An article in the latest Sloan Management Review reports on some research that this also happens when people just imagine the self-image boosting behavior. If you plan to go on a diet next week or imagine yourself giving to charity, you are more likely to be self-indulgent. Yet another example of us fooling ourselves so that we can do whatever we want. In this case literally having our cake and eating it too.

They suggest some good marketing ideas. If a store puts
donation bags or charity collection messages at the entrance of a store/web site, it would allow visitors to imagine giving or planning to give later and then they would buy more or higher margin stuff. It's a little hypocritical, but it would work.

Social Media Advertising

A recent study by Psychster had some interesting findings on web ads. What I want to focus on today was the one that surprised me. The ad types that were most perceived as ads (banner ads and signing up for company newsletters from an ad link) were in the mid-range in "I would buy products from this brand". And banner ads were also among the highest in "I would recommend this brand to a friend." So users in the study were receptive to the ads, despite all that has been said about them.

Of course, it was a very artificial task, so the results have to be taken with some huge grains of salt. The users did not interact with real pages or ads. They watched a video of someone interacting with the ad with a narrative explaining what was happening. So there was no possibility of banner blindness and all users had to interact with the ad because they had to watch the video to the end.

Tuesday, April 06, 2010

Divine politics

Another great article from Dan Ariely. You have just GOT to read this.

The obvious finding: We have similar views about divisive issues like gay marriage or the death penalty as we assume G-d does.

But is this because we create a G-d in our own image or because we gravitate towards a religion where the G-d already agrees with us? That is the fascinating part of this study.

They use a common manipulation to shift participants views about these subjects (by having them give a speech on camera arguing for the other side) and then measure how their view of G-d's opinion changed. When their opinion moved, their assumptions about what G-d believes moved too. Not because G-d was convinced by their speech, but because G-d always thought this way.

Sunday, April 04, 2010

Disclosing Conflicts of Interest

As usual, Dan Ariely (one of my favorite behavioral economists) has a great take on an important topic. When companies or individuals disclose their conflicts of interest, it actually makes the situation worse!!! For some very interesting reasons too.

The most interesting to me are:
1. When people disclose their conflicts, their behavior gets more conflicted because they feel it is more acceptable because of the disclosure.
2. The public has less suspicion of the conflict because the person must be honest if they are disclosing.

So in sum, the situation gets worse!!!!

Friday, April 02, 2010

Effective Elementary School Teaching

I read a research study today that investigated how to improve elementary school teaching. The study looked specifically at 5th grade math, but there are some important generalizations that can be made. The major findings:
  • Experience teaching that grade (5th) for that subject (math) led to higher student achievement. Other experience did not lead to better student achievement. Apparently, experience in one area or grade doesn't translate well to other areas or grades.
  • Connections from one team (i.e. all the 5th grade math teachers) to another team (e.g. 8th grade math or 5th grade English) didn't improve student achievement either. Again, the practices of one don't translate well to others.
  • When one teacher (among the 5th grade math team) had an advanced degree, all of the other teachers in that team (other 5th grade math teachers) were able to learn from and use what he/she had learned. Any additional investments in formal higher education for the team did not lead to better student achievement.
  • Strong ties from a teacher to the principal led to better student achievement of his/her students. But the benefits of these ties did not spread to students of other teachers on the team like the advanced degree did.
  • Strong connections developed between effective teachers so that they shared best practices and both improved. But the less effective teachers required the whole team to have a strong social network among all of the teachers in the team. Apparently, the effective teachers don't share as much with the less effective teachers unless the entire social network among the team is strong.
Implications:
  • Schools should invest in formal higher education for one member of each team, but not more. Additional funds should be used to provide incentives and tools to help teams develop strong ties and a dense social network so that all of the teachers can benefit from the insights and abilities of each.
  • Once a team is formed, stability is important. The school should try to identify where a teacher is best suited and keep them in that position for a long time.
The researchers note that the study is relevant to elementary school teachers, but not for other domains. The boundary conditions are likely to be very important. For example, University teaching best practices are probably more generalizable between years and subjects.