“First the worst, Second the best, Third the golden eagle.”

The above is a rhyme from my youth1, the implication being that you really don’t want to be first in anything. First is the worst after all. Obviously this is not true in the world beyond the playground, firsts in every field are celebrated, sometimes in great disproportion to their inherent worth2. How then is the position of first regarded by the human psyche? Do we give it more weight in other areas of our life? Can the order that we see things influence how we think about them, can we be manipulated into seeing things a certain way just by controlling what we see first?

Perhaps unsurprisingly the answer is yes. I should qualify that, it should be surprising that the order in which we observe something affects our decisions about it. We are after all rational beings aren’t we? We take in information, process it and come to conclusions based on the merits of what we have considered. Well, sort of. We are in actuality a mess of conflicting thoughts and desires, we all come with our own preconceptions and biases that can distort how we view the world without our really being aware of them. Is it really so surprising that we’ve found another one? Not to me, though this one is weird.

So what am I talking about really?3 Work performed by Dana Carney4 and Mahzarin Banaji looked at how the order we are exposed to things can affect the way we think about them and lead us to conclusions that otherwise we might not be justified in making. In a series of six experiments the question of how much influence the order we are exposed to stimuli affects our actions is asked. The experiments progressed from relatively cautious beginnings showing an implicit bias (using the Implicit Association Test – IAT as discussed in my previous post here) without explicit bias (self reporting) or real world consequences to actual choice biases in the real world.

In the first two experiments pairs of images (horses or people) were presented and then were rated for preference by the participants using either the IAT or an explicit seven point scale (eg “I strongly prefer x to y”). The implicit test showed a bias towards the first picture seen while the explicit measure did not.

Progressing from this the second two experiments attempted to determine if this implicit bias could be revealed in actions in the real world. Both used a pair of small consumables, either chewing gum or a lollypop, that would be placed on a table sequentially. The participant would then choose which item they preferred and would get to keep the item. In the first experiment subjects had to choose as quickly as they could (“within one second or so.”), this mimicked the IAT. In this condition 75% of participants chose the item placed on the table first.

The second experiment in this pair took place using commuters in a train station. Once again subjects were asked to choose one of two items which they were shown sequentially, this time however they were asked to either choose quickly or to take their time (mimicking the explicit test). In this set up 62% of the subject chose the first item in the time-pressured condition while no preference was seen if they could take their time.

Finally the study authors decided to test whether the preference for items seen first was due to an impulse to regard these items more positively than later seen items or if it was because the condition enhanced previously held beliefs about the object. In other words perhaps the first gum you see isn’t taken because your are made to think it’s better than the second gum but because you already have a positive view of gum and have this positive view reinforced by seeing one type first and so chose it for that reason.

To try and tease out these factors the authors used pictures of convicts in the final two experiments in order to use a stimulus that would naturally be seen negatively. In this set up participants were asked to indicate which criminal was a better candidate for parole. If first seem items have a more positive aspect imparted to them by virtue of being first then the first convict would be chosen. If on the other hand a baseline feeling about objects is enhanced then the second convict would be chosen. Once again however the IAT showed a preference for the first convict seen, labelling them better suited for parole. Also again, no preference was seen in the explicit measures.

What does this tell us about the general human capacity for decision making. Well superficially it’s good news. Given the time we will deliberate and make decisions based on the evidence available. If however we are working under time pressures then our implicit biases may come to the fore. There are however certain decisions that the authors of this study point out are known to be subject to implicit bias, such as consumer brand preference or even treatment preferences of doctors for specific patients.

We should be wary then of this new bias that has been thrown into the mix. Our heads are already full of biases that we may or may not be aware of, only by attempting to identify them do we stand much hope in correcting or tempering their influence on us.

So next time you need to make a snap decision, try to slow down and think it through. You may make a better choice5.

1. There seem to be versions of this around the world but the one above is most common in New Zealand as evidenced by this report.

2. Looking at you gold medal in Olympic BMX.

3. And how many rhetorical questions can I fit into one article? (does this one count? [or that one? {ok I’ll stop now}])

4. Of the “Is Your Boss A Better Liar Than You? Probably, Yes” post.

5. On the other hand you may not want to think complex decisions through too much.

Enhanced by Zemanta