Testing Assumptions Can Result in Better Decisions

There is an excellent Jewish expression called “bubbie-meise”, which refers to old wives’ tales. Some examples are: It’s bad luck to open an umbrella in the house! You can’t go swimming for one hour after you’ve eaten or you’ll drown! Eat all your food, there are starving children in Europe!  And my favorite, Chicken soup can cure anything!

One does not have to belong to any religion to believe in such things.  We generally tend to accept assumptions, beliefs, or superstitions as valid. Such anecdotes have been passed down for generations. But in the world of business, clinging to faulty data and being too stubborn to accept change cannot be cured buy a bowl of chicken soup, however tasty it may be.  We need to test assumptions to ensure better decisions.

Perceptions and inherent patterns cause us to rely on invalid assumptions.  People tend to be risk adverse, avoid change, and accept the most comfortable alternatives. However, just as we can develop routines to help us through our day, we can also develop routines to reduce the chance our assumptions are wrong.  Risk can never be fully eliminated, but understanding how it can be reduced can help us immensely.  It’s why we check the weather forecast before going outside or use the crosswalk to cross the street.

Analytics alone cannot resolve a conflict; it needs to be supplemented with passion, effort, commitment, and focus. Analytics can be less reliable when the data is wrong, when relationships are invalid, when sampling is inappropriate, and when risk is not considered. For example, the more creativity and uncertainty involved in any given situation, the more intuition and a little luck will be required.

Many economic proposals ignore that the economy is getting even more diverse. In 2023 nearly all the stock market gains are in seven stocks. The comparable returns of stocks and bonds seem to gyrate every day. Inflation, bank results, foreign activities and other factors seem to affect the economy every day. 

"When I let go of what I am, I become what I might be."  - Lao Tzu

Some suggestions to better test assumptions are as follows:

Review and evaluate processes and decisions. For example, it is unreal to me that objective testing regularly outperforms personal evaluations in employees’ personnel decisions. The reason is mostly poor training and bias.

Data needs to constantly updated. The latest census shows some dramatic changes in the makeup of our country, and that diversity needs to be considered when we gather and analyze data. Different regions have significantly varied ethnic as well as economic characteristics. A great amount of data also needs to be adjusted for the impact of the pandemic. For example, comparisons to last year or 2019 can show quite different results because of social changes brought on by the pandemic; there are more stay-at-home workers than ever before.

Facts are frequently more independent that we think. If you flip a coin a certain number of times, the odds will always work their way back to 50-50 regardless of any streak of heads or tails.   Cause and effect are frequently assumed rather than analyzed. Differing and multiple goals (such as short-term and long-term goals) can impact the understanding of cause and effect. Medical symptoms are frequently attributed to certain issues, while other factors may be the real cause.  That’s why it’s important to get a second opinion if something does not sound right to us.

Bias is one of the greatest complications when it comes to accuracy in the analysis of decisions. This includes statistical problems like sampling, measurement, and development of information. I also believe that social bias can be more impactful than statistical bias. This includes our preconceived perceptions and assumptions about factors affecting decisions. Cultural and environmental factors also affect bias. Dress, demographics, weather, location, and culture all affect perceptions in the decision-making process.  It is important to never assume anything based on the past.  I would refer to the Odd Couple episode “My Strife in Court” to illustrate what the word “assume” is made of.

"You know what happens when you assume..."

Risk assumptions and tolerance are critical to effective actions. Predicting results where there are significant and consistent historical data can be fairly simple; however, predicting results for new programs or with little or inconsistent data requires developing educated estimates. Assumptions regarding risk tolerance also need to be considered. For example, you generally need to be more cautious with regards to safety than low investment high reward opportunities like the lottery.  It’s crucial to know where every dollar is going, and where it can reap the greatest benefits.

Have you considered unconscious bias training?

Organizations need to be open to measurement and feedback and understand cultural parameters. Observing, understanding, and sharing financials, operations reports, and sales reports are the first step. A management style such as “walking around” and checking in with employees can be priceless. Balancing short-term and long-term goals, understanding challenges, and tolerance for failure are examples of understanding the cultural environment.

Analytics, tradition and experience are all valuable tools to improve decision making. However, you need to ensure that the assumptions behind those tools are accurate and reliable. In particular, our rapidly changing environment involving issues like Covid requires regular testing and validation. Similarly, creativity and intuition that defy some analyses and are becoming increasingly required. We can help you objectively search alternative causes and solutions. Understanding that there are various solutions to the problems we face out there only helps us improve our business practices.  It does not mean that chicken soup still does not hit the spot.

Related: Balancing Decisions for Effectiveness