

When we use overgeneralizations by labeling all the members of a group with the characteristics seen in some, or when we assign a label to someone or something based on the inferred character of that person or thing. These inflexible expectations for how things will turn out before they happen often prevent us from taking action. For example, when we tell ourselves there is no point in starting a diet because we will break it anyway, or when we don’t even attempt to participate in a competition because we are certain there is no way we will win. When we predict an outcome before we have enough evidence. The employee starts looking away or appears to be uncomfortable, and the manager infers that they must be responsible for the file’s disappearance, even though there are other potential explanations, such as the employee feeling unjustly accused, or them knowing about the person responsible but not wanting to rat them out. For instance, a manager asks an employee if they know about a file that has disappeared. When we infer a person’s probable thoughts from their behaviour and nonverbal communication. While we often jump to conclusions with good intentions, there are three main cognitive distortions that account for most inference-observation confusions.

The three types of inference-observation confusion The gerbil finally gets out of the woman’s pant leg, and she tries to explain what really happened. Another passerby sees the struggle, and assumes the first passerby is an attacker, so he punches him the face. A passerby thinks the woman is having a seizure, so he approaches her and wraps his arms around her to help. The woman pulls over, gets out of the car, and begins to jump up and down, shaking her leg to get rid of the gerbil. The gerbil escapes and starts crawling up her leg inside her pants. In the story, a woman is driving to her son’s show-and-tell session at school, with a pet gerbil in a box on the passenger seat. The story of the “Gerbil-Caused Accident”, described by Jan Harold Brunvand in his Encyclopedia of Urban Legends, is a perfect illustration of well-intended inference-observation confusion. In The Lost Art of Listening, Dr Michael Nichols explains how such assumptions may achieve the opposite of our goal: instead of sounding supportive, we may come across as dismissive. To sound compassionate and invested in what someone is telling us, we may interrupt them by saying “wow”, “what a shame”, or “I know what you’re going to say!” When, in fact, we have no idea how the person wants us to feel nor what they are going to say next. Often, jumping to conclusions comes from good intentions. As a result, we reach unwarranted conclusions without having all the facts. The inference-observation confusion happens when we fail to distinguish between what we observed first hand from what we have only assumed. The psychological term for jumping to conclusions is “inference-observation confusion”, which is defined as when people make an inference but fail to label it as one, ignoring the risk involved.

Why do we do it, and how can we avoid jumping to conclusions? In fact, doctors themselves often jump to conclusions: “Most incorrect diagnoses are due to physicians’ misconceptions of their patients, not technical mistakes like a faulty lab test,” writes Dr Jerome Groopman in How Doctors Think. Do you know someone who always seems to jump to conclusions? While this behaviour may be more obvious in some people than in others, we are all prone to it.
