Availability Bias

Introduction

A bias is a shortcut rule, which helps us calculate what we are supposed to do with low computational cost. Availability bias basically chooses whats important based on an events’ availability. The true value system could be hiding behind the biases. If we remove this availability bias, would it mean that we can see our true value system?

This could be helpful in answering the following: “In such a hypothetical event that there are 20 million people dying out of hunger, it being required that UN generate 4.4 billion $’s and only 10% of that available, should I be working towards this cause in full flow or should I try to improve my social skills so that I get laid? Rather, ‘what is important to my value system’.

What do you expect

In the essay on “the way ahead” I wrote about what I expect from writing, I put it here for the sake of completeness.

I expect to know what I should be focusing on in life. I don’t need to know the full solution, i.e., something that can guide me to eternity. I just need right now, something that can inform me what is more important to my true value system and why it is important. Something in the range of “Yo, people are the only thing that is important to you and you need to go to the extent of being called of being an extremeist because …” or, “ Yo, people are not important at all, you should focus on buying ferraris, because …”. The reason cannot be just any reason, it should be backed by science. For example, “Yo, people are the only thing that is important to you because, when we look at “X” bias, we see that although the bias sucks, it appears to be intended so that you save as many people as you can, hence pointing to your real value system.”

Availability bias

Availability is defined as the ease with which a particular idea is available to us. When we use this availability to judge our reality as to how often an event can occur, we are said to use the availability bias. My friend recently confronted me, regarding my hypothesis that I made a 3 years back. The hypothesis was related to getting a girlfriend once one moved to foreign country. This I believe I made when I was hearing a couple of stories around that people got girlfriends when they moved to a foreign country and hence have a higher availability. Needless to say I was excited to go to a foreign country for pretty much the same reason I think.

People apparently overestimate the likelihood of dying in a dramatic event such as terrorism as they are highly publicized and therefore have a higher availability. This heuristic is one of the reasons why people are more easily swayed by a single, vivid story than by a large body of statistical evidence. -Wikipedia article on Biases in general

Today, I was told two stories regarding people losing their deposit on the house. Its a lot of money and frankly a matter of concern. I’ve been stressing about it. Availability at it’s max. Whereas it should only be worthwhile to look at statistical evidence. I think I know where I can run such a survey on (on housing websites maybe!). Inherently I assumed that if I only heard 2 stories then there should be more. Seems like the availability bias is taking over. But maybe if I am really concerned, I should get some data on that. For me I think I would better be served to know more about my current landlord and if she took away the deposit of my previous tenants. She seems nice for now. Will hear from the tenants soon.

When people are asked whether there are more English words with K in the first position or with K in the third position, they use the same process. It is easy to think of words that begin with K, such as kangaroo, kitchen, or kept. It is harder to think of words with K as the third letter, such as lake, or acknowledge, although objectively these are three times more common.
-Wikipedia article on Availability Bias in general

In another experiment, subjects heard the names of many celebrities, roughly equal numbers of whom were male and female. The subjects were then asked whether the list of names included more men or more women. When the men in the list were more famous, a great majority of subjects incorrectly thought there were more of them, and vice versa for women.
-Wikipedia article on Availability Bias in general

In one experiment that occurred before the 1976 U.S. Presidential election, some participants were asked to imagine Gerald Ford winning, while others did the same for a Jimmy Carter victory. Each group subsequently viewed their allocated candidate as significantly more likely to win. The researchers found a similar effect when students imagined a good or a bad season for a college football team. The effect of imagination on subjective likelihood has been replicated by several other researchers.
-Wikipedia article on Availability Bias in general

When you get to hear about some people who are unemployed, you think, “It is very hard to find a job in Holland” 3 days back when you met your friends who found a job, “Almost everyone seems to have found a job in our batch (as we were thinking of people who didn’t have jobs, none came to mind). Only God Statistics can save you.

In summary, availability is the ease with which a particular idea/event is available to us. We use this availability/ease of retrieval of examples to judge the frequency of a given event. In other words, the availability bias provides us our Map. In the wikipedia article it is also said that,

Explanations

There seem to be two explanations for the availability bias informing us on what variable gives us the estimation of the probability of an event.

Tversky and Kahneman argue that the number of examples recalled from memory is used to infer the frequency with which such instances occur.

Participants were asked to listen to the 19 famous-men-names and 20 non-famous-women-names. Subsequently, it was observed that the students could recall more famous-names than the non-famous counterparts. Majority of the subjects estimated that they heard more number of men names than female names (19 men names and 20 women names). Thus Tversky and Kahneman tried to prove that the number of men-famous-names recalled from memory was used to infer the frequency at which men-names occured, according to the bias.
Source:WIKI

Schwarz and his collegues, on the other hand, proposed ease of retrieval explanation, in which, ease with which examples come to mind, not the number of examples, is used to infer frequency of a given class.

Participants were asked to make either 6 or 12 examples of their assertive behavior. Pretesting revealed that almost all the students were capable of generating 12 examples, but it was a difficult task. Participants who were asked to make 6 examples (an easier task) rated themselves as more assertive than participants who were asked to generate 12 examples. This seems to inform that the ease of retrieval of examples influences the outcome, i.e., the inference of the frequency with which instances might occur in the future [Source:WIKI]. This appears to disprove the hypothesis of Tversky and Kahneman as number of examples recalled clearly didn’t influence in this case.

It appears that Schwarz redefined the availability bias, and for all purposes, we are going to stick to only Schwarz’s explanation.

Importance

It (AB) is a mental shortcut that relies on immediate examples that come to given person’s mind when evaluating a specific topic, concept, method or decision. If something can be recalled it must be important… The availability heuristic operates on the notion that if something can be recalled, it must be important, or at least more important than alternative solutions which are not as readily recalled. -Wikipedia article on Availability Bias in general

Let’s take the experiment by Schwarz, and the above discussion on importance and connect them. According to his experiment, we predict that we are more assertive, when we are asked to make 6 examples of our assertive behavior(easy task) as compared to when we are asked to make 12 examples (difficult task).

Based on the ease of recall for an event (e.g., when one has been assertive), we can predict that there will be either more or less such events (of one being assertive). This means that __ is important than __.

It appears that when one was made to recall 6 examples, he did it with ease and hence the event of ‘being assertive’ seems important as compared to the other events such as ‘being a good son’, which might not so easily be recalled. And if the same person was made to recall 12 examples, he would have found it hard to do it. Hence the event of ‘being assertive’ becomes less important as compared to the other events which are easily recalled. I am not able to wrap my head around an explanation on whats important than what for Schwarz experiment. But,

Based on this ease of recall of an event, AB is able to decide for us, which event is important. Maybe the following is an example: When looking back at my last month, it was easy to recall examples of my feeling stressed last month because of money-matters, rather than recall examples of me feeling sad as a result of not having girlfriends. Hence I had a stressful month according to my bias, rather than a sad month because of my lack of girlfriends. Hence money-matters seems to be more important, as a result of ease-of-recalling, in comparison to not-having-girlfriends.

Availability only deals with predicting future probability of an event and then this can also be related to the importance of a task.

Availability bias and social proof

The AB (availability bias) asks us to use ease of recall of an event, to decide how important a task is. Whereas, social proof is the tendency of assuming the reactions of others, in an attempt to reflect correct behavior for a given situation. Can be seen as similar to bias.

I assume for now that social proof and AB (Availability Bias) are different, and try to differentiate them, so that I do not confuse one for the other. I see that the use of the word availability allows me to confuse with social proof, as social proof is based on in some sense the availability of people doing some actions. But, when we talk about AB, we are not really talking about availability of information, rather, ease of recall of examples. So we will taboo the word availability when talking about AB.

When you see examples of Indian people having girlfriends on facebook, you can’t but help feel bad for yourself or try to think of ways to get one. This information is “available” to you, and based on it you understand what is the right way to react. This appears to be a good example of social proof rather than AB, where we deal with ease of recall of examples.

In the case of AB, if I can easily recollect many examples of my friends have gf’s, then I guess it only allows me to predict that there are a lot of people who have gf’s. If I can recollect easily, then it is more important than other events that are less easily recallable.

Summary until now

The AB (Availability Bias) informs us through the Ease of recall of examples, the influence on ones decision in determining the frequency at which an event might happen in the future. More ease of recalling an event implies, more important the event is in comparison to other events which have less ease of recalling.

More ease of recalling examples implies more important something is. Of course this is biased. By that I mean to say that its just a Map, its relation to reality is still unknown.

Ease of Recall

People overestimate the likelihood of astrologers getting things right. For my cousin sister, its clear that astrology works. When asked she promptly gives the story of how the astrologer killed it with one of his predictions of someones marriage happening in a particular year. It seems like this information is very easy to recall for her. It appears that more than one bias could be at play here. Social proof might also be adding to the success of astrologers. Moving on…

Ease of Recall seems to be a function of some variables. Looking at Schwarz’s experiment it appears that the ease of recall is determined by the number of examples and the event. In this case, the number of examples asked to recall were either 6 or 12 and for the current event of ‘being assertive’, 6 was easy and 12 was hard for the current event. For now, Ease of recall is just a variable by itself. I am not going to get further into the number of examples and the event [Open issue 1].

For example, after seeing news stories about child abductions, people may judge the likelihood of this event is greater. People tend to rate “newsworthy” events as more likely because they can more readily recall an example from memory.

Recent information seems to have a huge effect on ease of recall and hence predicting the probability of future events. This also leads to ‘how important an event is to us’. In this case, ‘how important child abductions are as opposed to being in a plane crash’, for example.

So if we were bombarded with news that aleppo is fucked up, then the ease of recall at that moment increases, and we predict a higher probability for future events being fucked up in allepo. and This becomes important to us than other countries being fucked up.

Exception 1

So if I was bombarded with news of people dying, people fucking and people buying “equally”, then the ease of recall at that moment increases for all the events. All the events become more important. Maybe there is a superclause/superbias over this, to help us identify what the true value system want’s, once we make all events equally available, there by bypassing AB?

The problem with answering such a question is there are so many things unknown to me about AB and how exactly it works. For example, when I say “being bombarded with news “equally”, what can equally even mean? How will I know reliably what it means? Will there be a superbias which takes over in case of having reached an impasse? What is that super bias?

re-inventing the wheel

I feel like I am trying to reinvent the wheel. Maybe the greats have already figured this out. Maybe I am looking in the wrong direction. Maybe it’s a good idea to revert back to an STM, to get some guidance. I have spent quite some time on this, and will for now move on. It feels like a good exercise to rack ones brain.

Way forward

STM to the resue?

Open issues

In the Schwarz example, one gets asked to think of certain amount of examples. In reality though, you are on your own. How does the AB, impact life on a normal day, when you are not really asked to recall x number of examples.

Based on the number of examples asked to recall, one can believe he is either more assertive or less assertive. Such a ridiclous bias, which is not valid in this day and age atleast! How do we start to bypass it

Is AB the one that is responsible for the thing that I worry the most?

Is STM’s Idea based on this example!

In day to day life, we often think about many things, things such as work, play, women, writing, being social, people dying. Can I say that this is based on the availability at that moment, that it appears that certain things come to my mind and that hence they are available?

how do you compare one ease of recall with the other? if things like ‘number of examples’ exists then it is easy and tang