Statistics or real people – what convinces us?

Negative individual experiences influence us more than relevant statistics, even though we say the opposite. What consequences does that have for you, when you have to convince the CEO or Board of Directors?

Do you recognize these strong statements? If you do, do not worry. It is part of human nature to draw general conclusions based on individual experiences. And naturally, it also happens in executive offices.

Instead of judging these easily recognizable human qualities, this article is about how we assess and make decisions based on individual experiences, as well as what you can do to convince the decision-maker in your organization who requests statistics, calculations and data, but gets a far-away look in their eye when you give them a load of figures and graphs.

100 customers or Birgitte?

Imagine that you have to decide how good or bad a company is. You can choose one of the two following sources of information: Which do you choose?

• The results from a satisfaction survey with 100 previous customers
• The story of 51-year-old Birgitte’s experience as a customer

If you are like the majority, you chose the statistics with 100 previous customers. Why did you do that? Because you know that one customer’s experience does not necessarily say anything about the experiences of the other 99. That’s also how we usually work internally within companies. We come up with KPIs and measurements which give us data.

But does that mean that we are not influenced by individual events or our own individual experiences? Hardly. Just think about how we react to one bad dinner at a critically acclaimed restaurant or one bad review on Trustpilot. Or the other way around: when a colleague recommends a film or a good place to eat. That is also how decisions are made in the business world. We trust the most recent examples and our own knowledge. On a bad day it is called availability bias, because we make decisions based on flimsy foundations. On a good day we only use the one exact piece of relevant information we need to make a decision (and thereby save mental energy). 

This leads me to the really interesting aspect: how the things we say and the things we do contradict each other. When our preferences and behaviour do not agree with each other. Or put simply — when we say that we prefer a healthy balanced diet but end up coming home and devouring a bag of pick and mix instead of some fruit after a long day at work. Because that is what often happens when management make decisions. They ask for data but let themselves be influenced by examples and their own experiences. But to what extent?

People or figures?

The inspiration behind this article (and the results I present) come from the political scientist Asmus Leth Olsen. He researches human behaviour, primarily within the public sector. It might sound dry, but it is both interesting and relevant — also for private sector management. If you’re still reading this far, I have clearly convinced you.

Try to think about the classic public pillars in our society like hospitals, schools and the political system. Then try to think about examples where they have made bad, strange or outright wrong decisions about, for example, buying new IT systems or restructuring the organisation, where it came to light afterwards that there was a large amount of data showing that it was a bad decision. Are there similar examples in the private sector (read: humans often behave in a strange and logically inconsistent way)? But why does this happen when, in many cases, there are statistics and data which show us we are about to make a bad decision?

In the article “Human Interest or Hard Numbers” from 2017, Asmus Leth Olsen asked more than 1000 representative Danes about their assessment of hospitals: Do we prefer statistics or individual experiences when we decide how good or bad a hospital is? How do individual experiences versus statistics influence us when we assess a case? And how well do we remember examples and statistics?

The conclusions are very clear:

1. When we are asked what we consider important and what we would rather have, we say we prefer statistics over individual experiences.
2. When we actually have to evaluate the object (in this case, a hospital), we have a more negative opinion when we evaluate based on individual experiences rather than based on statistics.
3. Simply put, we just remember individual experiences better than statistics.

Intense pain after mistreatment

Think back to your answer to the question of customer satisfaction above. Now, I will present to you a patient’s experiences from a hospital. As you’ve probably guessed, it’s from Asmus Leth Olsen’s study.

Imagine that you know the following about a hospital:

Last year, 31-year-old Erik was admitted to hospital for a simple operation.
After the operation, Erik had intense pain due to mistreatment.

In light of this, how good do you think the hospital is?

How did you answer? It cannot be a good hospital if a patient like Erik is suffering with intense pain after a simple operation, or can it? It can be, if a clear majority of operations result in pain-free patients. However, Erik’s bad experience influences us proportionally more than a statistic, where he would just be one of many.

The graph below shows how individual experiences — respectively, “a patient’s” experience, Erik’s experience and Birgitte’s experience — result in lower assessment of a hospital than compared with a statistic where 1% of patients experienced the same “intense pain”. The personal examples had considerably more influence on our assessment of the hospital than the statistics.

To clarify the difference:

• Subjects who heard statistics stating that 1 in 100 patients (1%) had intense pain after treatment gave the hospital a score of just under 55 points out of 100.
• Conversely, those who heard the three personal examples (“a patient”, Birgitte and Erik), gave a collective score which was 17.8 points lower.

Most hospitals have more than 100 patients. That means that the individual experience example represents less than 1% of patients — probably a lot less. Therefore, it’s clear that individual experiences – compared with statistical data – makes us draw stronger conclusions. Olsen’s research also shows that the three individual experiences resulted in a much higher rate of emotional responses (anger, empathy, discomfort) than the statistics. Not only is our assessment more influenced, but we also become more engaged in the topic.

One individual experience can be one too many

Essentially, individual experiences have a greater effect on us, especially when they represent another real human being’s experience. Even via text, we can imagine how unpleasant it must have been to experience “intense pain due to mistreatment” (maybe you are even thinking about other similar stories you have heard, because this example has influenced you?). Imagine that you also saw a picture of Birgitte or Erik below a strongly biased headline. There you have the recipe for how individual personal stories in the media have such a strong effect on many of us. The same process can also occur in the boardroom if a CEO has a concrete personal experience related to the topic they are making a decision about.

It is worth noting that negative stories influence us more than positive (amongst other things, because they remind us of what we have to lose), and we should be aware of this. At the same time, the study emphasises that individual experiences rouse our emotions in a way in which statistics often cannot. For that reason, I would encourage you to use individual experiences which give a balanced view of the situation, if you are faced with a decision maker who is strongly influenced by their own experience. If you have to convince leadership who are influenced by emotions, individual examples should be met with individual examples (preferably combined with data, statistics and measurements), not just data and statistics alone.

Just one believable witness is enough

We began with 21^st Century hospitals, but end with jumping a long way back in time. Back more than 300 years before the birth of Christ, to one of Antiquity’s most essential thinkers: Aristotle. In his book on the art of speaking, Rhetoric, Aristotle describes, among other things, how a speaker (imagine you are a man from an affluent family in a Greek city state) should use examples to demonstrate his point. He discusses the importance of building arguments up in a way so that logical arguments are followed with and supported by personal examples. In this way personal examples take on the character of witness statements, and even just one believable witness is generally enough, according to Aristotle.

Things have not changed much between then and now, where a single example about another human being can also be enough to convince another person. With Aristotle’s rules and Asmus Leth Olsen’s research from the 21^st century in mind, relevant statistics can hardly be expected to have the same effect.

Practical advice you can take with you out into the real world

Next time you have to make a decision, consider a recommendation from your colleague, or convince other stakeholders to make an essential decision, make sure you are aware of how individual experiences can influence you, and use them with care. It is also just as important that you remember what Asmus Leth Olsen proved in his article: As humans we ask for statistics when we’re responsible for making decisions. However, individual experiences make us draw stronger conclusions which engage us emotionally, and we remember them better than statistics. So even though your stakeholders ask for data, they are more influenced by personal examples.

1. Do not ignore individual experiences, just because your statistics show a different (and potentially more accurate) picture.
2. Do not use statistics alone to counteract individual experiences, but instead talk about other individual experiences which give a more nuanced view of the situation. Substantiate your general points and statistics with personal examples which support your argument.
3. Negative individual experiences are very effective in influencing our opinions, because they appeal to our emotions and are easy to remember. Remember that, when you meet resistance — for example when you try to convince others to make a big decision, which can bring great change.

Read more here
Olsen, A. L. (2017). Human interest or hard numbers? Experiments on citizens’ selection, exposure, and recall of performance information. Public Administration Review, 77(3), 408-420.
See www.asmusolsen.com for a selection of Asmus Leth Olsen’s interesting research within the field of behavioral public administration.
The graphs above illustrate patient happiness and patient evaluations of hospital treatment, and come from the report Indblik i sundhedsvæsenets resultater 2016 (An insight into the Danish health service’s results in 2016), published by KL, Danske Regioner, The Danish Finance Ministry and the Ministry of Health and the Elderly.