“It wad frae monie a blunder free us,” wrote Robert Burns in his poem To A Louse. How do others see us, and if we could see ourselves as they do, from what blunders would that free us? If I sound as though I’m talking about public relations, image or brand management, then I probably am.

Imagine saying to someone at a party “Hi, I’m a statistician!” Such encounters have led to many jokes and comparisons with accountants or other supposedly boring professionals. In my experience they often bring back the other person’s memories of learning statistics, eliciting comments such as, “I never did understand the statistics in my psychology course,” or, “I hated statistics in college,” occasionally “I really liked statistics when I was at school (but went on to become a rocket scientist or a brain surgeon).”

Many of us will have helped people with their statistics. For example, over the years I’ve had many colleagues from other fields come to me when they found themselves unable to publish a paper on their work until they satisfied a referee on some statistical matter with which they were unfamiliar. Often a grad student or I was able to help. As long as we gave what was needed, we can expect that most of these people will go away from their encounter with us having a better informed and more positive view of statisticians and statistics than they did initially. They had a need, and we helped fill that need. I’ve always felt that these were very positive moments for statistics.

Our long-term scientific collaborators will be among our best informed and most faithful supporters. They know first-hand that our discipline is not simply the routine application of cook-book procedures. They have seen us puzzle, sweat, agonize, feel defeated, detour, think, learn, sweat some more, have ideas, and ultimately rise from the ashes to do something half-reasonable with their data. Unlike the lady in Burns’ poem, they see us as we would like to be seen.

Lately I’ve been reading books and attending conferences about Big Data, places where people like us do not figure very prominently. I’ve been wondering why not. At a meeting on Big Data in Health Informatics I asked professor from a medical school why she thought there were so few statisticians present. She replied that statisticians “don’t deal with risk, with uncertainty,” that we’re “too absolute, we do p-values, confidence intervals, definite things like that.” She, and others at this meeting, felt that something different was needed in their field: data visualization, pattern recognition, learning, prediction, modelling, simulation, multivariate testing, separating signal from noise. They didn’t see these activities as part of our world, as things statisticians do. We can smile (or grimace) at this misunderstanding, but again and again I hear or read that we deal with p-values, that we show how to do tests correctly, and that we can help calculate sample sizes for carrying out tests. I also see it in books on Big Data or Analytics. We’re not mentioned very often, but when we are, it’s usually in relation to testing.

Another view of statisticians: we tend to “raise arcane concerns about mathematical methods.” New York City’s first Director of Analytics is quoted as saying that, as well as, “I had no interest in very experienced statisticians,” and “I wasn’t even thinking about what model I was going to use. I wanted actionable insight, and that was all I cared about.”

What about the editors of journals such as those from Nature Publishing Group: how do they see us? They want us to help them reduce their irreproducibility, and so I suspect that they would agree with the view I saw on a recent statistics blog: “It’s our job to keep people honest.”

This is a sample of ways that others see us. Some are downright negative, others lukewarm, while others still (e.g. the policing role) are not likely to endear us to people. I don’t want to be seen as a policeman, or as a person who computes p-values correctly. Perhaps only our collaborators hold positive views, of us doing what we like to do and do best. If only more people knew us as our collaborators know us. Let’s get more of them!

What blunders can be avoided? I do think we need to talk much more about data visualization, pattern recognition, learning, prediction, modelling, simulation, and much less about testing. At parties we have an opportunity to explain that statistics is more understandable, more interesting and rewarding, indeed more fun than others might think; grasp it. Can you imagine ever saying “Simply: I make beautiful things with data”? Try it some time, and thank Hilary Mason.