Problem-solving can be difficult, but you don’t need a master’s degree in statistics or sophisticated mathematical models to get started. What you must do, though, is to eliminate biases.
There are more than 100 frequent cognitive errors that any of us can make, according to researchers. Consider confirmation bias, which is the tendency to pay attention to data that supports our previous ideas while ignoring information that contradicts them. Then there’s the sunk-cost fallacy, in which we double back on losses because we don’t want to admit we’ve made a mistake. The list could go on and on.
So, what’s the best strategy to stay away from these snares? In a nutshell, teamwork. Individual biases can be overcome via egalitarian work methods. Take it from Superforecasting author Philip Tetlock. The book concentrates on the art of foreseeing the future and discusses teamwork. When it comes to projecting future developments, Tetlock’s data reveals that well-organized teams usually beat the most skilled individuals. They can even outperform computers capable of processing massive amounts of raw data in some circumstances.
But, in this context, what does “highly organized” imply? According to Tetlock, the greatest teams optimize their problem-solving methods to foster an egalitarian environment in which everyone’s ideas are heard equally. The consulting firm McKinsey has a principle called the commitment to dissent that is heavily rooted in this idea.
This strategy means that junior team members are not only encouraged but also required to speak up when they disagree with senior employees. Superiors, on the other hand, pledge to pay attention to these points of view. What is the significance of this? Well, according to McKinsey, poor problem-solving is frequently the result of one type of bias: assessing ideas based on the position of the person proposing them rather than their quality. When everyone has a say, on the other hand, the team is much more likely to act on the greatest ideas.
A practical technique to create this kind of transparency and avoid senior team members from dominating talks is to assign team members ten votes represented by sticky notes. Add each proposal on a whiteboard, and have everyone in your team place sticky notes next to the idea they like most. As an added benefit, you may ensure that senior members vote last and do not influence the votes of others.
It’s one thing to collect data; it’s quite another to use it to develop useful solutions. That’s simply how data works. Data, as important as it is in addressing problems, cannot tell you anything on its own; you must make it speak.
There are good and bad methods to do this, as evidenced by the old statisticians’ joke about incompetent analysts torturing data until it says what they want to hear. That strategy will almost certainly lead you astray, but what is the alternative? It’s time to discuss heuristics. If you treat your data well, it will reward you with useful information.
The term “heuristic” derives from the Greek word heuriskein, which means “to find.” A heuristic’s objective, as the word’s origin suggests, is to assist you in finding something, specifically, a solution that fits the data in front of you. Let’s look at a handful of examples in more detail.
First and foremost, Occam’s razor. William of Ockham, an English philosopher, perfected this logical technique in the fourteenth century. The simplest option is usually the correct one, according to this rule. Your best bet, whatever your situation, is to go with the hypothesis that requires the fewest assumptions.
Let’s have a look at a simple math problem. Let’s pretend you have four assumptions, each with an 80% probability of being right. When you do the math, you’ll find that the chances of all four being true are just over 40%. By contrast, if you simply make two assumptions, it’s 64%. To put it another way, the less assumptions you make, the more likely you are to be correct.
Another useful heuristic is the 80:20 Rule. It is also known as the Pareto analysis because it was created by the twentieth-century Italian economist Vilfredo Pareto. According to the study, 20% of causes frequently decide 80% of outcomes. It’s fairly uncommon, for example, to learn that 20% of a product’s consumers account for 80% of sales.
To perform a Pareto analysis, make a list of your issues, which could include things like client complaints, missed orders, or broken products. Next, assign a score to each problem based on how significant a difference it will make if it is solved. After you’ve enumerated your issues, figure out what’s causing them, such as a lack of training, faulty equipment, or confusing protocols. Finally, sort the issues into groups based on their root causes and total up the scores. The bigger the impact of resolving this issue, or cause, the higher the total score.
Organizations frequently seek to comprehend the consequences of their policies. Take, for example, governments. Is it true that lowering taxes stimulates the economy? To discover out, you should do an experiment. How? So, within a specific income category, you could designate a control group, keep their tax rates the same while slashing everyone else’s, and see what happens. However, this type of real-world research is ethically questionable and, in many cases, illegal.
This is only one example of how a company could be unable to collect data. Budgetary constraints have a similar effect in other situations. However, there is a way past these roadblocks. If you take the time to look, you can find a wealth of important information in the actual world.
Take it from two political scientists, Evan Soltas and David Broockman, who sought to see if American voters discriminate against minority candidates in elections. They couldn’t think of a way to answer this question without conducting their own experiment, so they relied on a natural experiment.
Natural experiments are tests that happen to be done by the world and produce the data you’re looking for. It was a voting technique adopted by the Republican Party during presidential primaries in the state of Illinois in the case of Soltas and Broockman.
Instead of voting for candidates such as Trump or Romney, voters choose delegates to represent them. This isn’t rare in the United States, but Illinois has two peculiarities. First, these delegates’ names appear on ballots, despite the fact that they are politically unknown and frequently unsearchable. Second, voters are not required to vote for their favored candidate’s full slate, they can mark two of Trump’s delegates or two of Romney’s while ignoring a third.
As a result, voters have a decent notion of the ethnicity of the delegates; José, for example, is more likely to be Latino, whilst Tom and Dick are more likely to be white. This also means that voters will be able to choose amongst delegates who are running on the same platform. If voters truly discriminate, it stands to reason that minority candidates with names like José or Miguel would earn fewer votes than delegates with names like Tom and Dick when compared to the overall number of votes cast for Trump or Romney.
This is an excellent natural experiment since it provides the researchers with the data they require to begin answering their issue. All they have to do now is sift through the data, a procedure that takes far fewer resources than conducting their own trial. What is the moral of the story? If you look hard enough, you’ll probably find that data from someone else can answer your query!
One of the most crucial talents in the modern workplace is problem-solving. So, what’s the best way to go about it? The most crucial aspect of the procedure is correctly defining the problem. You may then start breaking it down into smaller portions and prioritizing solutions once you’ve done that. Look for high-impact, high-influence results in this case. Creating egalitarian work methods will help you eliminate prejudices, which will make problem-solving even more successful.
Check out my related post: How could you be better at solving problems?