Do You Hate Statistics as Much as Everyone Else?

December 15, 2008  |  Statistics

Photo by Darwin Bell

It happened again. I told someone I study statistics. He told me that he hated statistics in college. It doesn't annoy me like it used to - I've come to expect it - but why do so many people have this beef with stat? Is it really that boring? Confusing? What is it about statistics that turns people off? So I reach out to all of you:

What is it that makes statistics so uninteresting?

I'm going to assume that the icky factor is less for FlowingData readers (obviously), but still, I implore you - tell me why statistics sucks. I must know.

36 Comments

  • my guess is that it forces you to put your things in order. And, I should note, that is not plain statistics that people hate, is the method they hate. Remeber that we all love to throw a percentage or two in our discussion, in order to make our arguments more bullett proof. So, we love the results, that more often, we do not care to understand properly, and we hate the method that shows us that we are shallow. I studied statistics in school and loved it, but is also the merit of my two teachers I had, by the way, my BA is in Political Theory and Comparative Politics. It is imposible to study Political Science and not at least understand statistics, if not love them, it may be a too strong word.

  • first off, I think you’re asking the wrong audience. But here’s my guess on factors that lead in to this response:

    1. They hate math. Hatred of math tends to stem from difficulty with math and bad teachers. I had both in school and I could see that most of the people that had these problems hated math. I didn’t hate math, because I could see the value in it. What I did was hate facility in math. I’m still not the best in Math but I have tried to improve my capabilities over the years.

    2. They hate statistics for the same reasons as hating math, difficulty grasping the subject, and bad teachers.

    3. Overflow of part 1 into part 2, that is to say they hate math and see statistics as a subbranch of what they hate.

    4. The response also relates to lack of knowing how to respond to your branch of study. If I say what do you do, and you say I do X I am then obligated by the norms of social discourse to respond about what you do. If you say I study medicine I feel somewhat capable of responding with things about medicine I have learned from the TV. But if you say statistics then someone might not be able to respond knowledgeably about the subject, they might feel that they can come off somewhat stupid in an exchange therefore in order to forestall the appearance of stupidity they take a stance of animosity to the subject. You might get more extreme versions of this animosity in people who are more commonly understood ‘Humanities’ or ‘Creative’ subjects, where they will lead you to an understanding that they are interested in art and aesthetic subjects that are much more important than dry technical subjects.

    Anyway that’s my theory from seeing these kinds of responses.
    Maybe you could ask some follow-up studies to see if anyone falls into one of these groups of responses. :)

  • I think the key to answering this lies in the nature of the response itself (duh!). What I mean is that the response is emotional, probably somewhat irrational – contrary to the subject it is judging.

    I would take a guess that it probably has something to do with a natural (call it human?) dislike for logic-related subjects – people perceive them as cold and impenetrable.

    I can think of a few subjects. The first is maths – possibly the most hated, and possibly the most logical, or structured of the subjects I’ll mention. Statistics is second; there’s probably more interpretation involoved, but with that also comes a hearty distrust. Both maths and statistics also come with a heady dose of Latin/greek symbols which probably doesn’t help the penetrability aspect. The same can be said of economics and, perhaps more relevant here, econometrics. I’ve also found people have a dislike for other philosophical subjects which involve similar, if differently applied, forms of reasoning, such as ethics, epistemology, and the formal study of logic in philosophy. Even getting people to basically analyse and disect the truth of premises they make in everyday arguments was anathema to many students I knew (and eventually helped teach).

    I’ve studied all these subjects, but I loved maths but could never quite grasp statistics, even though I love data and want to learn more. I think the seeming absence of practical, everyday use of statistics is another of the reasons why people shy away from it.

  • Hey Nathan,

    Right now seems like a good time to plug a fun way to learn statistics: the “Manga Guide to Statistics”, available at http://oreilly.com/catalog/9781593271893/.

    I think statistics is uninteresting because it does not provide certainty. Forcing the decision making process through a set of tools that doesn’t tell you what is true, only what is unlikely, is the equivalent to torture for senior executives.

    Later,
    Jeff

  • my undergrad was in math/stats and my problem with stats (or at least, the use of stats by people who don’t understand stats [a great number!]) can be summed up here:

    http://www.guardian.co.uk/education/2008/sep/02/highereducation.maths

    http://www.edge.org/3rd_culture/taleb08/taleb08_index.html

    For any non-statistician who thinks stats are boring, I’d recommend the two books by the author of that second link, Fooled By Randomness and Black Swan.

  • A frighteningly high percentage of people are fundementally innumerate. I’ve been teaching 17 year olds about matters to do with money, and simple arithmetic is a real limiter. And this is kids who are going to university in September!
    It’s most worrying.

  • I liked stats and I taught it too to psychology students.

    It seemed that many students had not mastered even middle school algebra. Replacing X with a number would be was an issue.

    Then there are the examples. Teaching probablity with cards and dice requires conceptualization when these artefacts are cultural novelties. I’ve found that even illiterate (though bright) people understand probability and can explain it when we use examples from their own life (for example our peasant farmers planted three times to catch the possibility of early, middle and late rains).

    Then there is the demand to be logical. People with high F rather than T in their MBTI profile find that offensive. The reverse is also true.

    And then there are teachers not teaching at the right level for the students. People good at stats teach stats. Take them out and replace them with high school teachers who were good with the slower streams and we may get a different result.

    And then there is a change sweeping across all classrooms. Gen Y don’t treat truth objectively. They want to know how they are going to use something now – this minute.

    PS I’ve seen MBA papers where people are asked to do a t – test. That kind of tuition gives a subject a bad name. The problem needs to be couched in terms of the discipline being studied. We have seen a difference in numbers. Should we be concerned with the difference and why?

    Having said all that, this is a good thing. It leaves a role for you!

    I am a psychologist who works in the HRM space. Both are conversations stoppers too – and usually for good reason. When we stop to listen, we make good friends.

  • Many (most) people are uncomfortable with abstract thinking. Solving problems that exist only as ideas in your head requires disciplined and organized thinking, which many people avoid in the same way they avoid physical exercise.

    Personally, I blame the entire school system, with it’s “empty vessle” approach to learning — students are empty vessles and the teacher must fill them with knowledge. Students are spoon fed information and expect to be told what is important rather than determining it for themselves.

  • 1 – intro statistics has those “how many marbles in a cup” probability problems, which seem incredibly obvious to most people. it is so obvious, in fact, that i would guess that many people miss entirely the methodology you were supposed to use to arrive at the answer. so when the problems get slightly more complicated, intuition doesn’t work as well, and it seems mystifying.

    2 – statistics cares a lot about minutiae. most people don’t. knowing if the conditions of the mean value theorem hold true or not doesn’t interest most people. i think the uninteresting part here is just verifying conditions that have very little to do with the answer itself (from a layman’s perspective).

    3 – i have seen very few truly interesting statistics “textbook” problems. they are usually so rote and boring, it’s no wonder people turn away in horror.

    4 – it seems to me that stats need to be taught in context of something else in order to be seen as useful. for example, figuring out various population characteristics in biology is interesting, but not because the statistics is. or calculating probabilities of election results — the statistics isn’t what makes it interesting, but it does get you to the interesting results. even most data visualization is just a more user-friendly mask on some statistics.

  • It really gets down to presentation. So much statistical information is impenetrable because of it being poorly communicated.

    That’s certainly true for the way many textbooks are written and classes are taught. It’s also the case for much academic research, where the inclination to write only for peers puts the research in the language (formulae and words) of those with a presumed level of statistical training, and thereby limits the audience by design.

    To broaden the appreciation for and knowledge of statistics, the purveyors need to use more (lots of academic research, for example, doesn’t have any) and better graphics, more analogies, etc. The information needs to be accurately portrayed in ways that you don’t need to be an insider to understand.

  • I don’t think people dislike statistics because they are bad at math (though they may be bad at math).

    I don’t think the uncertainty is the reason, or the order it imposes.

    I think the major reason people dislike statistics is that it was poorly taught in whatever classes they took. Perhaps the instructor didn’t get it, or didn’t do the examples well.

    A related reason that people don’t like statistics is that any examples they ever saw were not relevant to something they understood or cared about.

    I wasn’t wild about the classroom statistics I had, but what I’ve learned since then has been interesting.

  • People recognize that, at times, organizations abuse statistics to promote hidden agendas but people fail to recognize the nature of the abuse, as a result statistics evoke a sense of confusion in those people and people dislike feeling confused.

    Statistics, like all fields, attracts a few great thinkers and a rash of pikers. In the hands of pikers, statistics seem to diminish existence, to shrink the world (rather than make it larger and more filled with detail which statistics in the hands of talent can do). People do not like to have their existence diminished.

    Statistics often analyzes and rarely synthesizes. To analyze, a border must be drawn. The border causes some people see statistics as an activity that argues parts represent wholes. People intuitively know parts do not represent wholes and blame statistics for their misunderstanding.

    To avoid any of the above when working with statistics requires much effort and intense thought. Most people prefer to avoid effort and thought.

    At which point we need to ask, except for those who must collect data and present interpretations in the way some people must paint or play violin, who would not hate statistics?

  • I, for one, hate statistics for the following reasons:
    - It’s pseudomathematics. It dresses up as a concise set of theories and methods, when these would more properly be referred to as cookbooks.
    - It’s simplistic. It gives a false sense of understanding about complex systems where no understanding exists. It prevents people from searching for mechanistic explanations that could indeed provide valuable insights.
    - It’s self-adulatory. Its practitioners have the courage to call every little possible way to plot data a “tool” or a “method”.
    - It’s too widespread. Most college programs that lack the most basic mathematics have their statistics courses (humanities and sciences), which helps spread misconceptions and misuse.

  • Also not to disregard is the connections in the public opinion’s mind between statistics and politics and maybe even worse, with marketing ;) Selling and/or lying. Please note that the most read book on statistics is appropriately named How To Lie With Statistics, I mean, what’s not to like in a discipline with such an exquisite sense of humor.

  • I wouldn’t know. I love statistics.

  • The media has made us numb to statistics. Also, most people arent
    familiar with evaluating statisitics.

  • Hey,
    I believe, like most things, it is how statistics is taught. I was taught stats by someone that really loved it and lectured with a passion at high-school. However, after a few years of non-use, I needed a refresher to complete my studies at University. All this managed to do was confuse the hell out of me. It was so badly taught that all it made the class do was hate stats with a passion.

    However, people need statistics to rationalize and understand large groups of data. Without stats, information can be meaningless. However, as Fabian has mentioned above here, the media has tended to dumb stats down to the point where people only comprehend percentages.

  • I’m going to jump in with the “How it’s taught crowd”. I had a statistics class where the professor spent most of the time deriving formulas on the chalkboard. Since then I’ve read a number of statistics books that used examples liberally throughout the text. The latter approach was more interesting and stuck with me much better.

    If statistics is taught by first showing WHAT you can do with it, then it becomes more enjoyable. Most people are not able to learn math without first seeing it’s application first. That goes for many (all?) of the sciences.

  • I did hate statistics in college as well. That has nothing to do with the science itself, but with the way it was taught. I was planning on doing a business degree, so statistics was a requirement. The professor who taught it was trying to fail us. We started with about 120 students in the class and after the first exam, only about half didn’t drop that class. I didn’t realize until much later how useful statistics is. Now I am reading the textbook again just to remind myself of the important concepts that I have to deal with at work all the time.

  • At least the Western culture, from what I can tell, retorts, “Good enough is good enough, even at the cost of future adequacy.” This is the presumption that leads to reserving rigorous math courses for later advanced levels rather than introducing (competent) proof structures during a student’s more formative years. Take that foundation, and (as someone else noted, roughly), toss on anything that smells of equivocation (the mishy-mashy greyness of probability), and the uneasiness is only compounded.

    Mathematics in general, and statistics thereby more specifically (for this discussion), is mystified precisely because we treat it as such, as a culture. Does this echo from centuries-old actual mysticism, the sort that made the inclusion of zero as a number a point of political and cultural contention? Maybe.

  • It’s a valuable tool, but I believe at it’s heart it is purely subjective. What is the best measure of central tendency for a data set, the mean, median or mode? It is arguably safe to say that it depends on the data. Likewise, if you’re model a time series, what is the best model to use? An ARMA(1, 1) can have an auto-correlation function that looks just like an AR(3) (for example). The tools we have to determine what model to use are not definitive, and so human judgment always comes into play. We only know in hindsight whether a model was correct or not.

    Since there is a necessity for human judgment, it is not as pure as other branches of mathematics. No human judgment is needed to prove the Pythagorean Theorem. Every moderately advanced civilization has come to the same conclusion. It’s true. Statistics is fuzzy. We’re leaving it up to someone else (statisticians and economists) to digest a large data set for us and to discern whether there’s anything important in there. People don’t like being told what to think, and there will always be a perception that statistics can hide meaningful information.

    Just my $.02.

  • I hated stats when I couldn’t visualize them. With websites like your we can interpret and manipulate the results to our needs!

    I used the wonderful immigration (stacked line) in class today. I teach Urban Studies and they fully appreciated!

    Thanks for your work and others who contribute.

  • I’ve wondered this myself over the years. I have an electrical engineer degree from a good school, top of the class sort of student, and I’ve been writing software for years. However, I have never read a book on probability or statistics that explained it to me very well.

    My impression is that most people, myself included, will tend to learn and reason visually or through a more narrative style. Hard-core statistics, however, is very difficult to discuss properly except by using equations. How many times do you read a statistics statement and think to yourself, ‘well… that’s not quite the right wording.’ Full correct statistics therefore does not lend itself to visual or narrative interpretation, unlike even calculus, frequency-domain analysis, or even multidimensional topology.

    That’s always been my best guess.

  • It’s been said already, but concisely:

    boring teachers

    It’s the same reason most people hate math/programming. Ask anyone what they thought of their CS lecturers:
    “One or two awesome, mostly sub par”

  • @Xavier – actually all of my CS professors were pretty dynamic. It’s the projects I didn’t like :)

  • I hate statistics for a number of reasons:
    - My intro professor was without a doubt the worst professor I have ever had. This was essentially intro to statistics for non-statisticians and she took powerpoint slides right from the textbook and threw them up on a screen. Needless to say, it was absolutely useless. Then, during the lab session, she was trying to teach us R without giving us a good background on the concepts. Thankfully, I found a book that barely got me through the class and gave me a great appreciation for some of the concepts. The worst professors are those who lecture for 90 minutes, then say “Any questions.” At which point you don’t even know where to start because s/he lost you in minute two and didn’t care. This was stats for me.

  • If more people were aware of the work of Edward Tufte (and those like you, his next-gen successors), I think they’d feel differently. Also though, coming from a literature/writing background, I think one thing that can make anyone who relates well to stories uncomfortable is the elimination of nuance–which numbers often trade for precision.

  • People get caught up in the mind-numbing exercises of their introductory classes and fail to see the forest through the trees. If people could understand the power of multivariate regression from day 1, there would be many more stat heads like us.

  • As a high school Algebra and Statistics teacher, this has been a great pleasure to read.

    I can relate to many of the above posts. One of the things that I see often is that by the time students have reached statistics, they have a relationship with math. Most every student who dislikes math can tell me what grade and what teacher did it in for them. It’s not statistics per se, it’s often emotional.

  • I feel compelled to add my vote to “boring and ineffective professor.” Our textbook was written by him. Our lectures consisted of watching him write out the steps that are written out in his textbook. For our exams we were allowed to use our book and any old exams, and new exams differed from those only in nouns and numbers.

    It was so bad that our queuing theory professor realized after two lectures that no student understood probability and statistics enough to understand queuing theory, so he spent two weeks of his own class teaching us what we should have already known.

  • Oddly, millions of people who don’t think of it this way devote significant portions of their time and money to recreational statistics. Some use only a few phenomenological metrics to buttress their intution, others draw on extremely sophisticated statistical machinery and domain-specific methods; but all are engaged in quantitative operational forecasting.

    These ‘Fantasy Sports’ enthusiasts constitute a major industry and in some cases (most notably baseball) have changed the way the sport itself is played.

  • i am not a statistician, nor am i a fan. i am, however, a grad student who was driven mad by what i viewed as lazy, sloppy, non-committed teaching on the undergrad level. therefore, i believe those who believe that statistics is the be-all greatest invention since termite control have only themselves to blame for the abysmal pr.

    secondly, are the textbooks. while an extension of classroom practices described above, textbooks are a living testimony to a number of misapplied conclusions of statisticians. first is this overarching assumption that statistics are relevant when it is my belief that their over-application in every facet of life, opinion making and goofy poll has rendered them trivial.

    secondly, a student viewing a textbook must wonder precisely who is the textbook being written for. one senses the author is beaming with pride at how clever they are when using marbles, or another silly toy, that shares no commonality with a human life. suffice to say, they must be writing for amusement and their fellow statisticians/academics who perhaps have long forgotten what it is like to be a student. the presentation and materials are red-zone convoluted, and not necessarily for human consumption.

    you don’t get it because once one reaches a certain acumen concerning a subject there is a danger of becoming biased and chauvinistic. you don’t understand because the world turns upside down: there are those that do and those that don’t, and obviously the ones that don’t are the ones with the problem. those that do merely become entrenched without a notion, or intention i might add, to look towards themselves as perhaps being the major cause of the problem. no, that can’t be.

    look, one size don’t fit all; and when you are talking about a subject as complicated and nuanced as statistics, the axiom described above becomes even more prevalent. it can be viewed by the consumer as “garbage in – garbage out”. because if you are betting on your appreciation for statistics to translate into an overall productive experience for all then you need to conduct some studies. some would go like this:

    1. how many college-level statistics instructors are actually qualified to teach?

    2. how many college-level statistics instructors are simply in the classroom for a paycheck?

    3. are students qualified to critique textbooks and, if so or not, are the gawds of statistics even interested?

    4. could those that teach statistics even handle the criticism, and do they even care?

    perhaps statistical analysis has already taken place concerning the cost-benefit of presenting statistics for mass consumption. maybe, through your prism a determination was made to do some “weeding out”, because surely many of those seemingly incapable of understanding statistical concepts have a deficiency in abstract thinking. of what use are they?

    it doesn’t take statistics to understand the shoddy nature by which it is presented in the classroom and texts. it requires an inward look at why this chosen approach is maintained. but that requires an admission of sins on the part of academics and the will to navigate their overall abysmal teaching in a new direction.

  • Statistics does suck. It is useless garbage that I will NEVER use. I am 54 years old and I have NEVER used it at work or even running my own business for 18+ years, so what the hell do I need if for now? I have to take it to graduate with my degree.

    I bore two boys, raised them, I have undergone open heart surgery and I have NEVER experienced the level of frustration and pain as I have had in this statistics class.

    The textbook is POORLY written and the online venue? DON’T have anything to do with Pearson!
    I would rather eat glass, drive a pencil through my eye AND walk on coals then to put up with this crap.

    There has been nothing my whole life, that could not be figured out by using just addtion, subtraction, multiplying and dividing. The plus? No STUPID rules, that if this happens, then use this or if there is this do this. PLAAEEEEZE! Who thought this junk up????

  • man, where are these comments coming from all of a sudden on this old post?

  • There are a variety of things I’ll never use in my life. It does not follow that they suck. But, really, here’s something to think about:

    There are people who do use statistics to understand the social behavioral groups to which you belong, and therefrom decide how better to appeal to you, even manipulate you. I know; I’m one of them. You could worse than to understand statistics as a means of knowing how these organizations operate.

    Beyond that, though, it doesn’t follow that a poorly written textbook in a badly taught class about a given field means the field itself is worthless. If so, then every field of endeavor is worthless.

Copyright © 2007-2014 FlowingData. All rights reserved. Hosted by Linode.