The Affect Heuristic – How we can use data to overcome our own bias in our decision-making
Oftentimes, our current situation of progress and success blind us to what is approaching on the horizon. It is very hard to avoid this, considering partly what makes us human is this ability to become comfortable in a present state, even if that present state will become harmful to us in the future. This is known as the Affect Heuristic, a way in which human beings show bias in making a decision, taking action that may be contrary to logic and objective thought.
Look at the financial crisis of 2008, where our lavish expenditures and comfort led to our own demise in many respects – a perfect example of how our society was blinded by the comfort we had come to inhabit.
We have seen the effects of what can happen when we let our guard down – when we start ignoring the signs that may be staring us in the face. Instead of letting the data or trends tell us what to expect and how to prepare for what may be approaching, we continue on our path of neglect and foolishly act surprised when the situation hits us hard.
Let’s take a look at a specific example from the recent past – the decline of sales in the U.S. auto market. The 1990s was a time of great economic success and excess. People had more money to spend, and they spent it on lavishly large vehicles such as SUVs, which the auto market in America was providing plenty of. One could argue that the events of the past decade were not foreseeable by the U.S. auto industry, and thus their inability to react was excusable and understandable – hence the bailouts. However, it was less than 30 years ago that the industry suffered much of the same declines as they did in the past decade.
A report released in 1980 by Natural Resources and Commerce Division of the Congressional Budget Office indicated that the auto industry in America was suffering unprecedented decline. The reasons cited for this decline may sound very familiar:
- Jump in gasoline prices
- Rise in interest rates and enactment of credit controls
- Economic recession
The impact that followed may also sound familiar – consumers switching to compact cars that met their needs which were more readily available by foreign automakers. The suggestions and predictions of the CBO stated that in order for the U.S. industry to become viable and competitive again, they would have to produce more compact cars. Perhaps it is just me, but I think there is not a clearer example of “history repeating itself” than this.
You may ask yourself how this affects you and your own situation. The U.S. automakers failed to listen to the data that was undoubtedly available to them. Are you in tune with what the data are telling you? Are you listening to it to make decisions for the future? Or are you blinded by perhaps recent success and letting that bias your decisions and direction for the future.
This isn’t a situation where ignorant people made foolish decisions. More so, it is simply a lack of understanding of how valuable data can be. Decisions can’t simply be made on gut instinct; and while we should all listen to our gut, using it as a sole means of direction can be misleading and dangerous to our own condition.
If history tells us anything, it says that it will visit us again – and the only way to overcome those reenactments is if you stay in tune with what has happened. A good consumer of research has the ability to take in is happening and be proactive in addressing it. Use the data to recycle what has succeeded and reevaluate what has failed. In short, don’t fall victim to your own comfort bias – but be objective and deliberate in your approach.
How to Really Reduce the Number of Smokers in KY
This past Tuesday (9-6-2011) the Center for Disease Control released a new report on smoking among adults in the US. The results weren’t surprising really. Smoking overall has declined but not as much as the CDC had hoped. Overall, about 19.3% of adults (roughly 45 million people) in the US smoke. That is down from about 20.9% who smoked in 2005. Furthermore, the people who smoke a lot (i.e. 1.5 packs of cigarettes per day) also declined, going from 13% in 2005 to 8% in 2010. So not only are fewer people smoking but they are smoking less when they do smoke. The Courier-Journal also wrote a pretty good story on this study.
While those trends provide some good news we also know that the smoking rate would decrease much faster if teenagers and young adults didn’t start smoking in the first place. After all, if the supply of new smokers is stopped then the smoking rate would only continue to decline as a result of other attrition elements. However, this opens up a much larger question about why do kids start smoking in the first place. There are myriad opinions and thoughts about this subject and unfortunately most of the conversations never progress beyond the opinion stage.
Based on prior research we have been involved with through the Drive Cancer Out program we know that school age children exhibit strong predictive patterns around their likelihood to try smoking. Those predictive patterns center around their beliefs that:
- Kids can smoke once in a while without getting addicted or suffering any harmful effects
- People who smoke are cool
The stronger a child’s association with these two statements, the more likely that child will be to try smoking as they become older…even when they tell you that they know smoking is harmful. This becomes powerful because when we can identify these children, intervention and deterrent programs can be provided. However, without the aid of statistics to isolate the key predictive drivers of smoking then all efforts to curtail the problem become subject to opinions and whims.
If the goal is to reduce the number of smokers then the only real path to success involves understanding why people start smoking and deter them prior to the habit taking shape. Data can help make this a reality whereas opinions often only succeed in expending needless energy and precious resources.
How We Think – The Return Trip Effect
A fundamental component of research is exploring how people think and what tools and processes we use when we form decisions. Over the last few years this field of exploration has increased considerably. This increase in knowledge has been a benefit for the research profession, but is also a benefit for anyone wanting to have a deeper understanding of what is really going on when a person makes a decision. Books such as Predictably Irrational and The Invisible Gorilla help to elucidate the hidden processes that take place for all of us.
This week I was again reminded of this field of study when I read an article on the Return Trip Effect. Most people have experienced this phenomena. You are taking a trip someplace, maybe you are leaving on vacation. The trip to get to your destination seems to take forever while the return trip seems to take less time…even though, both trips probably took about the same amount of time. Many people assume this feeling is because a person is more familiar with the landscape on the return trip and therefore can better anticipate the return trip time. As it turns out, this is not the real cause.
In reality, the perceived time difference is a result of people’s anticipation of the destination to which they are traveling. This anticipation influences their perceptions of the initial outbound trip time making it feel longer. When people are returning home, their anticipation is much more sedate and as a result they are better able to predict their actual travel time. As a result, the return trip doesn’t feel as long as the outbound trip.
To contrast this effect, consider your daily commute to and from work. If this is part of your normal routine then you likely have reasonable expectations about both the outgoing and returning trips. As a result, both legs of the journey seem to take about the same amount of time.
The article in USA Today does a pretty good job of summarizing the report. This finding is important not just to help us have better vacations but because of the underlying implications of these findings. We know that people are capable of accurately providing their opinions about a topic. However, those opinions are subject to change and those opinions are influenced by underlying factors that are not as obvious. By having a more full understanding of the components that impact a decision we are better able to understand behavior. And most importantly we know that the only way to change behavior is to fully understand what causes a behavior.
Understanding Report Statistics – An upcoming White Paper
Pardon the lack of relative dormancy here over the past couple weeks. Things have been rather busy around here.
Nevertheless, I want to make you aware of a recent white paper we will be releasing on being a smart consumer and reader of statistics. This white paper is sparked in part by a recent report we released on city services provided by Metro Government here in Louisville. The study showed that while Police, EMS, and Fire services are generally highly looked upon here, Waste and Transportation services are not to such a high degree.
The report was released and subsequent stories were published on it. It made headlines on the Courier Journal. Like usual, however, it sparked a debate over the validity of the results, namely in reference to how the data was gathered. We found that much of the community does not understand the way random sampling works – how we can collect a sample of 1,092 residents randomly throughout the city and that be representative of the entire community. In essence, if they weren’t personally asked, then how can we say, for instance, that 91% of Louisville is highly satisfied with Fire services.
Without starting a new discourse on random sampling, the point of the matter here is that such reports can only be useful if they are understood by those who read it. The audience must be educated to facts of statistics, and that is partly the task of the researchers.
Be on the lookout for our white paper, which we will undoubtedly post a link to on here once published.
Having a Successful First Year in College
Over the next few weeks thousands of students will begin a new school year. Many of those students will be going off to college for the first time. While this is an exciting time for
both students and parents alike, this is also time of fear for many students. While the excitement will make for an energetic first few weeks, many students will not return for the spring semester.
We have done a great deal of research on the attitudinal barriers to student success and there are specific fears that students have. Many of those fears center around family support, peer support, and academic ability. By applying some basic principles parents can help ensure that their kids will be successful their freshman year in college.
If you know someone going to college this fall take a look at our summary for a few key areas where you can help your student succeed.
STATS-DC – session – Best Practices and Public Data
This blog post is designed to pull some of the different themes from two different sessions. The first “Best Practices in Linking PK-12 ad Higher Education Data” was a more technology based discussion and the second was a session on NCES data that are available.
The first session focused on the best practice winners from the annual PESC best practices competition. Lately there has been a strong trend toward best practices that are submitted which focus on linking the work of K-12 with the needs and work of postsecondary researchers. 
This session primarily focused on the underlying data architecture that is required to “link” the data. By migrating away from EDI and replacing with XML but still retaining a common architecture the data are able to be used more broadly while ensuring greater quality.
Data interoperability is a significant issue. As a primary research company, IQS Research is fortunate to be able to gather the vast majority of the information we need for our studies. However, when working with education research there are huge opportunities when we can link our primary research findings, on the individual respondent level, to performance data within the education system. Furthermore, if those educational performance data are interlinked such that we can trace a student across state lines, across universities, etc, we are able to heighten the data we provide.
These data are currently being analyzed and compared at the aggregate and strata level, but further interoperability ensures that we can drill down to the student level in most cases. That will allow us the ability to identify attitudes and behaviors at the individual level and compare those to the self-reported information for purposes of building regression models and defining predictors for success.
Even though we cannot directly link our data to this information there is still a lot of great secondary research data available to the public at www.nces.ed.gov. Under their data tools tab and their surveys and programs tab there are a myriad of studies and data that can be accessed. Many of these are longitudinal in nature and track students from elementary into their careers (different studies).
Blog Contributors
-
KnowledgeDynamo
-
Joshua Holt
-
Ryan Gore
-
Shawn Herbig
Straw Polls – Should we listen to them?
As a researcher, I absolutely love election season. While I could say that the reason for this is that I am simply living up to my obligations as a citizen (partly true), the real reason I enjoy it so is because of all the polls that are released. And because so many polls are released, it can become difficult to decipher which ones are good and which ones are political nonsense. That is what makes it interesting for a researcher!
Let’s begin with the first question – what is a straw poll? There are two broad categories of polling: scientific and unscientific. Scientific polling uses random sampling controls so that the results from a sample that is drawn is statistically representative of the population. Previous posts have discussed this greater detail. Unscientific polling, on the other hand, has no systematic sampling controls in place that would allow for representation of a population. Historically, a lot of straw polls in the United States have been political in nature, and are usually fielded during election season by a particular political party. The very name “straw poll” alludes to their nature – it is thought that this idiom alludes to a piece of straw being held in the air to determine which direction the wind is blowing.
Most straw polls are very targeted, very narrow surveys of opinion. Their main purpose is to take a “snapshot” of a general opinion during a particular point in time. This seems valid enough, but the difference between scientific and straw polls exists within the methodology. Most straw polls use a form of convenience sampling that is a bit unorthodox, and the selection bias associated with can be extreme.
It is hard to assign a broad methodology to all straw polls (as each one is different in its own right), but many of them have candidates, such as in the Ames Straw Poll in Iowa, attract voters to cast their vote on who they believe should be the Republican candidate. If it sounds like political grandstanding, it’s because it is to some degree. It uses somewhat of an “honor system” whereby anyone can vote (within the parameters), which opens up a whole argument regarding the validity of the polls.
This brings us to our second question – should we pay any heed to the results of these polls? I previously stated many of the recent straw polls and their victors. There have been many polls, and there have been many different winners. But to answer this question, we only need to look at the candidates themselves. And they certainly place weight on these polls. Tim Pawlenty dropped out of the Republican primary because of the lack of support the Iowa poll showed for his campaign. Entire strategies are formulated based on results of straw polls. That is because these polls show the weaknesses of particular candidates. And for this reason, candidates are perhaps wise to take caution to what the polls are telling them.
However, are they good predictors of ultimate outcomes? In answering this question, we are reminded of the 1936 presidential election. The Literary Digest conducted its own straw poll, which showed Franklin Delano Roosevelt being defeated by a large majority. We all know this was not the case, and the reason for this catastrophic (as it led to the downfall of the Digest) miscalculation was in the methodology of the poll, which is the main criticism of any straw poll. The Digest used their mailing list to administer the poll, which consisted of motor vehicle registries and telephone books. The problem here? It was the Great Depression – many Americans were too poor to own a car or telephone, and thus a large sector of the population was neglected in this poll (selection bias at its finest), the very sector that was more likely to vote for FDR and his economic reforms.
The point of this post is this: take what you hear from these straw polls with a grain of salt. They do little to predict outcomes, but can be very valuable to the candidates themselves in adjusting and fine tuning their campaigns. Although there is a vast expanse of difference that exists between a lot of straw polls and scientific research, it can be surprisingly easy to muddle the reliability of each. However, knowing how to digest the results of research, both good and bad, will help you to avoid unsettling surprises.