Anchors Have Fear, Knowledge is Here: a report of how knowledge reduces the effects of anchors

We have known for a while now that different anchors, even if completely irrelevant, can influence people’s responses, especially for numeric based ideas. This is in part because once people have a number in their mind, when coming up with a response to a question or problem, they start at that anchor and then adjust from there. However, in situations where they are estimating the answer, there is often a large range of plausible answers. When adjusting, people tend to stop once they reach a plausible answer, meaning that they’ll stop at the lower end of the range if adjusting from below and a higher end if adjusting from above.

Since this anchoring and adjusting effect is so dependent on the range of plausible answers, it seems to make sense that if one had greater knowledge of the topic in question that they would be able to narrow down the possible answers; thus, they would be considering a smaller range, which should result in the anchor having a reduced effect on their response.

However, studies in the past don’t seem to confirm this idea; while there are some that provide evidence for this claim, there are others that refute it as well, and these are the ones that provide the idea that seems to be more universally accepted: knowledge level doesn’t moderate anchoring effects.

The logic behind why it seems like knowledge should influence the effects of anchoring is strong enough though that some people were unhappy with the current conclusions. That is why Andrew Smith, Paul Windschitl, and Kathryn Bruchmann decided to do some farther research into the matter. They conducted three causational or quasi-experimental studies along with one experimental study to explore the relationship between knowledge and anchoring affects. Their results led them to the conclusion they suspected, which contradicts the past-accepted idea that these two variables are unrelated. They found that knowledge level does in fact moderate the extent of the anchoring effects.

In the first three studies, the knowledge level of each participant was not modified; each participant came with their own level of knowledge on the topics being discussed. However, different measures were taken to account for this. For example, in their second study, Smith, Windschitl, and Bruchmann obtained participants from both the US and India, and then asked questions relating to either the US or India. This ensured that each participant answered questions about a topic that they were more knowledgeable about as well as for a topic that they were less familiar with. The results showed that there was a correlation between knowledge level and anchoring effects: the US participants were less affected by the anchors for the questions involving US topics, and the Indian participants were less affected by the anchors for the questions involving the Indian topics.

Since the past evidence contradicts these results, Smith, Windschitl, and Bruchmann also performed an experiment where they manipulated the independent variable (the knowledge level of the participants). This gives them a little bit more control over the situation so that they can try to confirm whether or not there is actually a causal relationship between these two variables. In this experiment, participants were given instructions and information before they took a survey asking them questions about prices for new midsized sedans (these questions contained either high or low anchors). The participants were randomly divided into three groups, and the only difference between each of the groups was the information and instructions provided at the beginning. The first group was given information that was not related to prices of midsized sedans, and simply instructed to look over this information. The second group received information about prices of midsized sedans, but again was only instructed to look over the information. The third group received the same information as the second group, but was asked to study the information, as it would be helpful later on. The important thing to note here is that neither group 1 or group 2 really gain any knowledge on the subject matter; even though group 2 is provided with important information; simply looking over information does not increase their knowledge about it. Therefore group three is the only group with advanced knowledge on the matter. This is important, as the results showed that they were the only group that was somewhat resistant to the anchors provided.

The series of studies performed by Smith, Windschitl, and Bruchmann puts some minds at ease because it provides evidence for what logically made sense: knowledge does in fact decrease the effect of anchors. The results in this series are also capable of explaining how past experiments came to the opposing conclusion, even though it no longer seems to be true.

Smith, A. R., Windschit, P. D., & Bruchmann, K. (2013). Knowledge matters: Anchoring effects are moderated by knowledge level. European Journal of Social Psychology, 43(1), 97-108

This entry was posted in PS253-2015. Bookmark the permalink.

Comments are closed.