Daniel Salvagni — Yet another software engineer

Managerial Decision Making

Early this year I decided to focus my studies on the leadership & management sides of organisations, and I enrolled for Strategic Leadership and Management Specialization, a specialization from the University of Illinois on Coursera. I found the courses really complete in terms of content and recommended material, and, so far, these courses gave me a great background on topics such as the foundations and applications of leadership, group decisions and fostering decision, conflict management and so on.

When I took Designing the Organization: From Strategy to Organizational Structure, I thought it would be far from what I’d like to learn at this moment, however, it helped me to understand why organisations are designed in the way they are. I’m currently in the 4th course of 7, called Managing the Organization: From Organizational Design to Execution where I’m currently learning about effective organisational decision-making. This is such a great topic that worth to leave my notes with kind of a course summary here.

This is going to be a long article and most of these topics would require an article for each, however, this article aims to share an overview of them. I’m going to share information about the different types of decision-making in organisations, prescriptive and descriptive models, rational choice and cognition and cognition process, availability biases, representativeness biases among other topics. The most important is: how can we do better in decision-making?

Different types of decision-making in organisations

There are basically two different critical dimensions of decision-making, known as agreement on the cause/effect relationship and agreement on preference.

The first is related to how much the parts agree that a particular action could result in a particular outcome. Let’s assume that one department of our organisation said that we need to increase our profits (effect/outcome) and we are given two different solutions (cause/action): 1) invest in marketing and sales; 2) reduce the costs of production; Both seem reasonable actions for the expected outcome, how can we agree on which of those are the most effective? That’s when it comes to the second decision-making type, the agreement on preference. How can we agree that invest in marketing and sales is the better option?

Agreement on the cause/effect relationship

The degree of an agreement the members of an organisation have about a particular action (cause) would produce an expected outcome (effect) in a given decision situation.

Agreement on preference

The degree of an agreement the members of an organisation have on the desired outcome of a given decision situation.

When we all agree on the cause/effect relationship and we also agree on preferences, we call it Computational Decisions, or Rational Model. When we have a high agreement on cause/effect, but a low agreement on preferences, we call it Compromise Decisions or Political Model. A low agreement on cause/effect and a high agreement on preference is called Judgmental Decisions, or Incremental Model. And, finally, low agreement in both axis is called Inspirational Decisions, or Mixed Model. We can consider a low agreement as disagreement as well.

If we understand the principles and process of decision-making, why do we fail in making successful decisions?

Decision-makers fail in identifying the cause/effect relationship and this is the main reason why managers fail to make successful decisions. It means, basically, that we are dealing with the problem of uncertainty.

There are two models that can give us detailed answers, Prescriptive and Descriptive Models.

Prescriptive models describe how people would behave if they follow certain requirements of rational decision-making. Descriptive models say we need to investigate and learn how people behave and their strategies of making decisions.

Prescriptive models of decision making

This model tries to understand where they made a mistake and develop methods to show the right way of taking the decision.

In other words, it prescribes the right way of making complex optimal decisions. For instance, developing models to calculate the Return of Investment it’s a prescription whether you should take the decision of investing or not.

Prescriptive models are also known as Expected Unity Theory and they are based on fundamental assumptions, as the following:

  • Ordering of alternatives Decision-makers should be able to compare any two alternatives.
  • Dominance Decision-makers would never select an alternative that is dominated by a better one.
  • Cancellation A choice between two alternatives should depend only on those outcomes that differ, not on outcomes that are similar.
  • Transitivity If A > B, B > C, then A > C
  • Invariance Decision-makers shouldn’t be affected by the way the alternative is presented.

The paradox of Rationality and Bounded Rationality

Even though we aim to make rational decisions, we often face the paradox of rationality. The paradox is that we use different methods to decide on different cases. In a rational scenario, we would choose the same method for different situations. It means that we are bounded to rational instead of being completely rational.

In the view of the Bounded Rationality, decisions-makers acts as Satisficers seeking satisfactory rather than optimal solution. This aspect is called Satisficing.

For instance, while buying a new laptop, we don’t look at all models available or we don’t go to all stores to compare models. Usually, we go to a couple of dealers and choose one that satisfies our decision.

Some people argue that we have bounded rationality due to limited willpower. We save less money for retirement but we are willing to spend this money on something ephemeral such as a dinner or tickets for a concert. That’s our Bounded Willpower.

Besides that, it says that people are not only interested in their outcomes, but also other’s people outcome. That’s the reason, for example, why there are such competitions in organisations about salaries, raises, and positions. We care, or we act like we’d care, about others, and that might make us to not take 100% rational decision. This is called Bounded self-interest.


People choose a path that satisfies their most important needs even though the choice may bot be ideal or optimal.

Bounded willpower

People give more weight to present concerns rather than future concerns or needs.

Bounded self-interest

People care about the outcome of others.

Descriptive models of decision making

Descriptive models are also known as Prospect theory and it says we need to investigate and learn how people behave and their strategies of making decisions. One reason is to understand why people deviate from rational decision-making systems is to develop methods to avoid predictable errors in order to make better decisions.

There are some cognitive limitations that we need to pay attention to understand where we as individuals make mistakes and look for solutions to avoid it.

Decision-making is tricky as expected and bellow you are going to find some of the traps that we might fall into.

When we are facing potential losses we become more aggressive and willing to take risks. However, we are conservative about gains.

  • Gain and losses are defined in terms of deviation from a reference point. We have different points of view about what’s good or bad.
  • Decision-makers are usually loss averse.
  • Value of something increases when it becomes part of one’s endowment. People think that what they own values more than the market says.
  • Preferences depend on the way the problem is framed.
  • Certainty effect: we pay more for certainty even though the chances of going wrong are low.
  • Psychological accounting, we think in percentages instead of actual numbers.

What’s the difference between rational choice and cognition in decision-making?

It’s easier to compare side-by-side the list of principles of each one.

Principals of Rational Choice

  • Goals are clear and agreed.
  • Decision-makers can calculate all alternatives and their outcomes;
  • and can evaluate all alternatives simultaneously, not sequentially.
  • and can use absolute standards rather than relative standards.
  • Decision-makers uses factual information to choose among alternatives and chooses the alternative with the highest payoff.

Principles of Cognition and the Cognitive Process

  • Our goals are ambiguous.
  • There’s not always an agreement on the goals and the way we frame the decision situation, our preferences and goals may change.
  • Decision-makers always have limited information and limited information processing capability.
  • We can’t compare hundreds of alternatives all the time and we usually evaluate alternatives sequentially.
  • We compare A e B and chose one and then go to another one;
  • Decision-makers evaluate alternatives against an implicit favorite (usually the status quo). For instance, doing nothing. So we compare if we don’t do anything versus if we do something.
  • Decision-makers process information that is usually distorted.
  • Decision-makers usually choose alternatives that satisfy most of the criteria you have developed.

How do these factors actually influence the way managers makes decisions?

When faced with a situation that requires a decision, the goal of a decision-maker is to make the most effective decision. Although, there are cognitive elements, also known as Managers’ Cognitive Structure, that will impact whether we make effective decisions or not.

That’s why we need to develop organizational structures and processes in order to reduce these elements’ impacts in making the decision.

“the tendency to believe that what people do reflects who they are”.

Fundamental Attribution Error is a psychological concept that explains people behavior which tends to emphasizes internal factors rather than external situations. It’s our tendency to attribute another’s actions to their character while attributing our own behavior to external situational factors.

Since we can not always see the whole picture, what’s in evidence most of the time is the behavior. For instance, if you notice that your work colleague is getting late at work in the morning or to some meetings, your first judgments tend to be about your colleague’s characteristics and motives rather than the situational context. It could be delayed due to car traffic, by taking the kids to school, by any other reasons that aren’t related to that person’s personality.

Information is one of the most valuable assets. We can only take rational decision by having enough information at our hand. Nevertheless, decision-makers can be misled by information when they face distorted data or not enough random samples of subjects in their collection. Below you are going to find three sources of biases faced by managers in decision-making, as seen in professor Huseyin Leblebici’s class.

The Illusion of Correlation and Causality

Does employees satisfaction lead to high performance or company’s success have a stronger impact on employees happiness?

The Halo Effect

The tendency to look at a company overall performance and make attributions about its culture, leadership, or practices;

The Illusion of Connecting the Winning Dots

If we pick a number of successful companies and search for what they have in common, we’ll never isolate the reasons for their success, because we have no way of comparing them with less successful companies.

Escalation of Commitment

Escalation of commitment is our tendency to increase our commitment to our initial decision in spite of getting negative feedback.

For instance, let’s say that the “Innovation Lab” of your organisation came up with a project of software that compresses movie files to 75% of its original size by losing 10% of quality and your organisation decides to invest in this project. Later in the same year, the Innovation Lab has developed already 80% of the software and you read in the news that your competitor is about to launch an even more efficient compression software that compresses a file to 90% of its original size without losing any quality. At this point, your organisation has invested EUR 90 million out of EUR 100 million of the approved budget. What should you do? Would you keep investing the remaining EUR 10 million in the project and finish it or would you throw it away and save this money?

Most people would keep investing in the project. They might think that the project might succeed in the market, but in reality, it’s not.

We pay more attention to confirm information rather than to disconfirm it.

Why does this occur? It’s hard to admit to ourselves that we made a mistake in our decision. It has to do with management impressions as well. We want to be seen as consistent managers and we don’t like to admit failures. Besides, this is an example of judgmental biases. We are risk-averse to positively framed problems and risk-seeking to negatively framed problems.

Decisions must be made about the future rather than about the past.

The investment done in the example above are sunk-costs. They are already in the past and you can’t count on it for the future. It’s better to use it as a lesson an take the remaining money and invest it in something new. Obviously, this is not an easy decision, although there are some rules and process that can help us to avoid unproductive investments. We are going to see it later.

Availability Heuristics

If I’d ask you what is the most cause of death in Brazil, Stroke or Interpersonal Violence, what would you choose? You probably are going to choose Interpersonal Violence as it’s more common and usual in the news than Strokes. However, according to the IHME, Stroke has a higher position than Interpersonal Violence in the death causes report.

That’s Availability Heuristics, we tend to assign a probability to an event by how quickly and easily we can remember similar events.

These associations are made because it’s easy to recall, easy to retrieve this information from memory and then we presume these associations. We see that several violent interpersonal attacks happened and it was reported in the news during one week, we presumed to assume that it might happen the next week as well.

How can we avoid Availability Heuristics Biases?

It’s important to maintain accurate records about particular events that occurred in the past, as it makes easy to minimize primacy and recency effects and other distortions.

It also helps to get expert knowledge to avoid wishful thinking. Sometimes the probability of desired outcomes is overestimated.

Finally, we should break compound events in smaller events to be analyzed independently, estimating the probability of each simple event first.

Representativeness Heuristics

We have presumed associations in our mind that makes we judge and categorize a decision situation based on a pattern of our previous experiences or expectations. People are insensitive to base rates and to sample size. Sometimes we don’t consider the context and that we don’t have the right amount of samples to have a conclusion.

There’s also the Misconception of Chance and Regression to the Mean. An event that occurred today might not happen tomorrow. Same as an extreme event measured in the first time might be an average one when measuring the second time.

How can we avoid Representativeness Heuristics Biases?

Sometimes we can be misled by a highly detailed scenario because in such cases, it seems to be a representative scenario, although, it might be only an exception. Moreover, consider base rates as it’s especially important when an event is very rare or very common.

A hint from this class is also that Chance is not self-correcting. It doesn’t mean that a negative event that happened today might be a positive event happening tomorrow.

And last but not least, not misinterpret the regressions to the mean. Regression to the mean is typical when an outcome depends in part upon chance factors.

Confirmation Bias

As mentioned before, we tend to pay more attention to confirm a piece of information rather than disconfirming it. We usually search, interpret or recall information in a way that confirms our hypothesis and give less attention to alternative possibilities.

Critical traps in decision-making

We should be aware of some well-know traps in decision-making and one of the lessons give these two examples. The Sunk-Cost trap is one that we already talked about previously, let’s take a look also in the Anchoring Trap.

The Anchoring means that we might be anchored by initial information that we give disproportionate weight. For instance, investing in a particular stock: you have the information that a particular stock value increased or decreased in the recent past. What possible of these situations would stop you from buying?

The answer is that neither of that information is relevant. Whether you would invest in the stock market for the next five years, the recent changes in the price are irrelevant.

However, since we have that information now, it anchors us. It puts us in a particular perspective.

The Anchoring

  • Always view a problem from different perspectives.
  • Look for information from different sources.
  • Be aware of anchors in negotiations.

The Sunk-Cost

  • Cultivate an organisation culture that is not afraid of failure.
  • Seek the view of the people that were not involved in the early decision.
  • Establish rules that facilitate getting out of unproductive investments.

What should we do before we make critical decisions?

As stated in the class, there are a couple of situations that are especially common in organisations, such as the examples described below.

We take credit for the positive outcome while attributing the negative outcome to external factors.

As a team, usually compete for limited resources and create projects to emphasize the positive aspect of it, however, usually, the same team doesn’t adjust the original estimates to account for inevitable problems. Besides, we usually do not consider our competitors’ moves in the same market as we develop our estimates. And since the most promises proposal receives the resources, we tend to make over-optimistic forecasts.

There’s also the fact the most managers reward optimism and interpret pessimist as disloyalty. Keeping the positive-vibe just for the sake of having the impression of a positive company might affect your decisions.

Organisation Strategies

Although the checklist is in hand to help us to make better decisions, the most important approach should be developing organisation strategies, process and practices for different decision-making situations.

For instance, create a process for budgeting, paying attention to anchoring as we always think in percentages in these cases. Other processes such as for Merges & Acquisition, or to avoid Sunk-Cost traps and even in the hiring process. Hiring new members for the team, we need to pay attention to Halo Effect and Representativeness Bias.

It may seem that all these processes are for complex decisions only that envolve huge outcomes, such an acquisition. However, we should also pay attention to the biases in our routine decisions as they are as much as important.

Principles of effective organisational decision-making

To conclude this module, we are left with a set of a top 3 principles of effective decision-making, recommend by professor Huseyin Leblebici.

Managers must have at least two alternatives when dealing with critical strategic decisions.

Managers must discuss their assumptions rather than predictions to avoid confirmation bias.

Managers must establish clear sunset rules in order to avoid the sunk-cost trap.

If you made until here, congratulations. I really appreciate that you took your time to read all this article. I hope that this information is useful for you and can help you taking better decisions in de future.