Simplifying Decisions
By Richard Erdmann & Christine Drew
All decisions ultimately come down to a Yes or No choice; yes we will do this or no we won’t, but too often we allow the process of deciding to also be binary. By binary, we mean a ‘whether or not’ process that examines only one course of action or an ‘either-or’ decision in which two options are considered but only in an all or nothing context. Paul Nutt, author of Why Decisions Fail, released a study in 1993 of 183 decisions. Only 29% of those decisions involved more than one alternative. Think about that for a moment. In 81% of decisions, only one alternative is even considered. As a decision-making process, considering only one option doesn’t seem like much of a process. In addition, Nutt concluded from his overall work on decision making that whether or not decisions failed 52% of the time versus only 32% of decisions involving two or more alternative choices. Let’s look at the risk with two alternative choices first.
The advantage of two alternatives is that they can often be combined for a solution superior to either alternative. As we begin the deliberation or the combination, there is still a risk that the process becomes an ‘either-or’ process and combining is either not considered or viewed as a compromise and the argument goes underground for a while. We have a classic example – the reading wars.
The reading wars of the 1990s, and re-emerging again today, positioned phonics against a literature-based approach. Framed as an ‘either-or’ process, the fight became mean spirited and political. The science was and is clear that phonics should be included in the reading equation. The real question is whether the mastery of phonics requires the learner to only be exposed to consistent sound-symbol relationships for some period of time before being exposed to a broader range of literature. If the answer is that sound-symbol relationships need not be consistent or need to be consistent for only a very short period of time, balanced literacy should have emerged much earlier and ended the debate. Instead, balanced literacy emerged late in the debate, battle lines were drawn as though it was trench warfare, and balanced literacy won a battle but not the war. Over time a literature-based approach pushed phonics off the table, teachers came into the classroom never having learned to read using phonics nor trained in phonics, and the ‘either-or’ decision never really went away. New research in the cognitive sciences indicates that a phonetic approach should never have disappeared, which is very different than saying phonics only.
But, that leaves the ‘whether or not’ decision in which only one alternative is considered. There is a study of 83 decisions made over eighteen months in a German firm. The decisions were well documented and archived. Some years later they were opened to researchers. The researchers took the decisions to the board and asked the board to rank the success of the decisions having the benefit of hindsight.
The firm had a rule for decision making; there could never be more than three alternatives considered. Even with this rule 40% of the decisions involved only one alternative (whether or not), 55% were two-alternative decisions, leaving only 5% as having three alternatives. The board ranked 40% of the multi-option decisions as very good versus only 6% of the ‘whether or not” decisions. That is a staggering difference.
Why don’t single option, ’whether or not’ decisions work? The answer reflects two very different kinds of problems. The first is outright bias. No alternatives are ever considered and the deck is often stacked in favor of yes. The decision becomes easy and involves very little rigor. The second problem becomes one of focus. Since the decision is often weighted from the beginning, the focus is not on the decision itself but on implementation. The focus is on how to garner support and how to make it work. Our example comes from the teacher accountability movement.
As the teacher accountability movement gained political traction, two decisions were made. End-of-year tests would be the measure of student learning and be a part of teacher evaluation, and there would be an observation protocol through which teachers would be evaluated and coached. Although most states picked one of two popular protocols, several states did not. We worked in a state that preferred a home grown version. We also worked in a district that was one of the pilot sites.
In reality, the state only considered developing its own solution. The two popular protocols served more as models than options. With these two protocols as options, alternative approaches to teacher accountability were never considered. While it was a ‘whether or not’ decision, it was stacked heavily in favor of a particular solution. Federal funds were attached to their decision. Like the nationally adopted protocols, this state developed a very complex, complicated, time consuming protocol that required a considerable amount of training. With this in mind, let’s return to Paul Nutt’s conclusions.
- Were alternative protocols considered that might have been considerably less intrusive or threatening? An alternative, recommended by research on decision making, is multitracking. Comparing this state’s approach to the two prominent protocols, Danielson’s and Marzano’s, is a tweaking comparison. Were there any real alternatives piloted? No. Do they exist? Yes.
- Was the pilot focused on whether or not the protocol worked (did test scores improve) or did it focus on how to make it work? To our knowledge, the question ‘would it work’ was not really asked but assumed. The question was how to make it work.
- Being a state program, it was mandated, but there was considerable energy spent on how to get everyone on board and behind it.
Both of these processes are extensions of mental shortcuts. The ‘either-or’ process bends to the human tendency to frame in the extreme. One is a liberal or a conservative. One is pro-testing or anti-testing. Hans Rosling, author of Factfulness, calls this the gap instinct. The ‘whether or not’ process is an extension of confirmation bias. This is the tendency to bend new evidence to fit existing beliefs.
To combat both of these tendencies, thinking needs to encompass broad swatches of diverse input. Multiple alternatives need to be considered.