Stopping Unwanted Behaviors

By the end of January every year, many people who made New Year’s resolutions discover that it is hard to keep them.  By mid-February, most people who resolved to change their behavior will find that they have failed and are back to their old habits.

One difficulty with changing behavior is that you have to stop doing an undesired behavior, and ultimately replace it with one or more desired behaviors.  As I discuss in Smart Change, your actions are governed by two brain networks that can be called the Go System and the Stop System.  The Go System engages goals and promotes action.  The Stop System inhibits behaviors that the Go System has begun to execute.  Your ability to stop a behavior successfully, then, depends on two factors: how strongly the Go System is engaged and how effectively the Stop System can inhibit an action.

Psychologists have developed a number of laboratory tasks that can be used to study the Go and Stop systems in action.  One interesting task was used in a paper in the December, 2013 issue of Psychological Science by Julie Bugg and Michael Scullin.  This task is called the Prospective Memory task.  Participants are told that they will see a number of strings of letters (like DUCK or DCUK) and they will have to respond whether the letters form a word by pressing buttons on a computer keyboard.  This task is called a Lexical Decision task.  However, participants are also told that there are two special target words (say, CORN and DANCER).  If they see one of the target words, they should press a third button to indicate they saw the target.  

After doing that task for 80 trials, participants are then told that they will continue to do a lexical decision task, but the target task is now over, so they should only respond whether the letters they see form a word.  Even if one of the target words from the first part of the study appears again, they should not press the special key.  They key measure in this study is whether participants mistakenly press the special button when they see one of the words that had been a target in the first part of the study.

Previous research with this task suggests that the more times people see the target words in the first part of the task, the more likely that they will be to mistakenly continue to press the button in the second part of the task.  The idea is that when you see the words many times in the first part of the task, it activates the Go System more strongly, and that makes it harder for the Stop System to overcome the temptation to press the key in the second part of the task.

In this latest paper, though, the researchers explored an interesting case.  Some participants saw target words 4 times in the first part of the study.  Other participants never saw target words in the first part of the study.  In the second part of the study, all participants saw words that had been targets 10 times over 260 trials.  Surprisingly, 56% of the participants who never saw targets in the first part of the study mistakenly pressed the target button at least once, but none of the participants who saw the target words 4 times in the first part of the study mistakenly pressed the target button.

What is going on here? 

The first part of the study creates the goal to respond to the target words.  When people see those target words 4 times in the first part of the study the Go System recognizes that the goal has been achieved, and so its activation is dampened in the second part of the study.  When people never see the target words in the first part of the study, the Go System remains active in the second part of the study, which makes it harder for the Stop System to completely overcome the desire to act.  

As a demonstration of this point, another study in this paper repeated this study.  This time, participants were once again focused on two target words in the first part of the study.  This time, 4 targets appeared in the first part of the study, but participants saw only one of the two target words (so half of the participants saw only CORN and the other half saw only DANCER).  In the second part of the study, participants were much more likely to mistakenly press the special button when they saw the target that had not appeared in the first part of the study than when they saw the target that had appeared.

Why does this matter?

When you try to change your behavior, you probably focus most on stopping the undesired behaviors.  It may be even more important to find ways to dampen the activity of the Go System to make the undesired behaviors easier to overcome.  One way to do that can be to perform the undesired behavior a little in order to make the Go System believe that it has achieved its goal rather than avoiding the undesired behavior altogether.

Obviously, there is a gap between these simple laboratory studies and the real world.  So, there needs to be more research that helps to translate these intriguing findings into specific recommendations. 

Testosterone, Trust, and Social Status

Over the past several years, much has been made of the psychological influences of the hormone Oxytocin on behavior.  It has been called the love hormone and the trust hormone.  And there are some findings out there to support that increases in oxytocin levels can increase positive thoughts toward a partner and trust in some cases.  

In that same time period, there has been much less discussion of the influence of testosterone on social relationships.  This topic is explored in a paper by Maarten Boksem, Pranjal Mehta, Bram Van den Bergh, Veerle van Son, Stefan Trautann, Karin Roelofs, Ale Smidts, and Alan Sanfey in the November, 2013 issue of Psychological Science. 

In this paper, participants were women who were either given a dose of testosterone or a placebo.  Women were used, because men’s level of testosterone fluctuates significantly throughout the day.  The testing was done about 4 hours after the administration of the testosterone, because previous studies suggest that the strongest influence of testosterone on behavior occurs about 4 hours after the testosterone is released into the bloodstream.

Participants played two economic games.  In the first, they were given 20 Euros, and were told that they were playing with a partner.  They could give as much of that 20€ to the other participant.  The experimenter would then triple the donation, and the partner could return as much of that amount as they chose to the original donor.  For example, if the participant gave 10€ to the partner, the partner would get 30€, which they could split ay way they like with the participant.  This game is a measure of trust, because the participant has to trust that they will get back from the other player at least as much as they gave up. 

After playing one round of this trust game participants were then told that they were playing the other side of the game.  Now, they were given 60€, with the explanation that the other player had given their full 20€, and were asked how much they wanted to return to the other player.  This game is a measure of how much people are willing to reciprocate in social situations.

The trust game results suggest that testosterone decreases trust.  Participants gave about 54% of their money to their partner when they received the placebo, but gave only about 38% of their money when they received testosterone. 

The reciprocation game suggests that testosterone increases reciprocation.  Participants returned only about 43% of the money when given the placebo, but returned about 53% of the money when given testosterone.

Other measures taken during this study suggest that testosterone did not make participants more tolerant of risky options or more tolerant of ambiguous situations.

What does all of this mean?

The finding that testosterone decreases trust is consistent with a lot of work that suggests that testosterone makes people vigilant for violations that may disrupt a person’s social standing.

The reciprocation effects are interesting, though.  When a person reciprocates in a game like this, they are sending a signal that they can be relied on in social situations.  When people can be relied upon, that generally increases their social standing.  So, testosterone may be helping people to act in a way that improves their social position.

Of course, the doses of testosterone used in this study are much higher than what women will naturally experience.  Further research needs to explore whether similar effects can be obtained with testosterone levels that are similar to those that occur naturally.

The Attractiveness of Average and Familiar Faces

Over the years, many studies have examined what people find attractive in faces.  One important factor is symmetry.  If you draw a line down the middle of someone’s face, the more similar the right and left sides of the face, the more attractive it is seen to be.  Evolutionary psychologists have argued that we like symmetry in faces, because it is a sign of health.  

Symmetry is not the only factor that affects attractiveness, of course.  For example, Marilyn Monroe had a prominent mole on only one side of her face.  This mole may have helped to draw people’s attention to her face, which then increased her perceived beauty.

One demonstration that symmetry is important to judgments of facial attractiveness comes from morphing studies.  In these experiments, people rate the attractiveness of a series of faces along with one face that is the morphed average of the set of faces.  This morphed average is generally seen as more attractive than the mean rating of all of the faces from which it was generated.  A big reason for this advantage for the average face is that it is much more symmetric than the individual faces from which it was generated.  The morphing process eliminates the asymmetries in the individual faces.  

An interesting paper by Jamin Halberstadt, Diane Pecher, Rene Zeelenberg, Laurent Ip Wai, and Piotr Winkielman in the November, 2013 issue of Psychological Science demonstrates that the familiarity of a face may play an even bigger role in the attractiveness of the faces than symmetry.  

The authors started with 28 pictures of celebrities from the Netherlands and 28 pictures of celebrities from New Zealand.  They created morphs of pairs of faces to create 14 morphs.  Then, participants from the Netherlands and New Zealand rated the attractiveness of the morphs as well as the attractiveness of the individual faces.  All of the morphs were rated first.   

Participants were also asked whether they recognized the morphs and the individual faces. 

When participants rated faces of celebrities from another country, the typical averaging effect was observed.  The New Zealand participants found the morphs of two Dutch celebrities to be more attractive than the individual photos.  The Dutch participants found the morphs of two celebrities from New Zealand to be more attractive than the individual photos.

When people judged celebrities from their own country, though, familiarity took over.  On average, the ratings of the individual faces of celebrities from that culture were seen as more attractive than the morphs.  The more strongly that a person recognized the celebrity, the more strongly they preferred the individual faces to the morphs.

This study is a nice demonstration of the influence of familiarity on judgments.  We are wired to prefer familiar things to unfamiliar ones. For example, the mere exposure effect demonstrates that seeing something even once makes it more desirable than something that has never been seen before.  So, even though there may be general factors that can make something attractive or desirable, there is no substitute for making it familiar.