Manipulation of Data Lives

I am hip deep in the final prep for Jaguar Bound, working furiously to fix a couple of minor issues that came up in the conversion process. Because of that, I frankly forgot this was Tuesday. That meant I started scrambling this morning before my first cup of coffee finished brewing to find a topic. In my search, I stumbled across an article over on The Bookseller that caught my eye. Then I looked deeper and realized they were following the same plan so much of the media does.

The article in question deals with the conclusion that much of the publishing industry is facing burnout. The magazine based this on a poll it conducted. So far, so good, right?

Not necessarily.

You see, the poll was announced by The Bookseller on May 5th. The poll closed at 11:59 pm May 6th. Basically, it lasted less than 48 hours. In the announcement, there is no information about other means word of the poll might have been spread or how the questions were selected. That’s problem number one. Problem number two is it limits–severely–the number of those in the industry who will see the announcement, much less have time to answer it. Problem number three, and the one that really resonates with me, is the poll itself is no longer available to check to see what questions were asked, how the questions were framed or even if there was anything that limited access to the poll to those who actually worked in the publishing industry.

And that is just scratching the surface of my concerns with the poll.

But let’s have a look at some of the so-called data it produced. Data that was correlated an reported two days after the poll closed.

According to the report, “more than” 230 people responded to the poll. Of those, 87% were listed as “publishing staffers”. According to The Bookseller, “89% of staffers responding to the survey had experienced stress during the course of their work over the last year, while 69% reported burnout.”

Okay, I don’t know about you, but that doesn’t seem all that out of line. How many employees don’t feel stress at some point during a work day, much less during a year of work? This is the problem with reporting poll results without giving the actual poll question and possible answers at the same time. Let’s see what else this so-called poll had to say.

They complained about having to work in an “always on” atmosphere, of having to work more hours than they were contracted for, and one even complained because they only got to take five days of “holidays” because there was no one to cover for them and they had to make up the time by working weekends.

Pardon me while I boo-hoo–not. Most workers do not have some assistant to cover for them while they take time off. They have to work to either get ahead of the game or know they have to make up for it when they get back. As for the “always on” atmosphere, welcome to the real world, boys and girls. Try working in the service industry for a bit. Then you will know what it means to be “on” all the time. When most of your money comes from tips, you don’t get to be “off”.

I don’t mean to make light of some problems that do exist in the industry. But these aren’t issues unique to publishing. More than that, knowing how many of my friends and family lost their jobs during the pandemic, I get a bit bitchy when I see others complaining because they have a job. Besides that, the additional hours aren’t all that unusual for the industry. I know any number of editors, assistants, etc., who have always put in the extra time because that is how they would get ahead.

But here’s the thing. The data is not a good indicator of what is happening in the industry as a whole. There is no scientific measure of how respondents were chosen. There is no way unless you grabbed a screen shot of the questions before they were taken off-line to see if there was any bias written into the questions and answers. The pool of respondents is too small. But the conclusions are being toted as being indicative of the industry a a whole.

Sounds kind of familiar, doesn’t it? We see this sort of thing all the time in the MSM when it comes to political or social issues. (And this is not an invitation to go into those polls and their results.)

And, frankly, considering I haven’t had more than two days off at a time in at least four years makes me cringe when someone complains about getting what is probably a paid vacation–something I haven’t had in years. I don’t write and publish, I don’t get paid.

So, what are your thoughts about the poll and its results?

Now for a bit of housekeeping and promo.

In the past week, I was called out by a reader for enforcing the rules of the blog. Here’s the thing. We have very few rules here. The main one is that we aren’t a political blog and, unless the OP opens the conversation up to politics, they don’t belong here. The second is that comments need to be related to the post or be a direct response to another comment. We have enough open floors where you can toss out comments and unrelated observations or you can email any of us with your thoughts. Along this line, one sure way to get a warning and possible ban is if you continue to use the comment section of the blog to flog your favorite topic, whether it is anti-Amazon or anything else, when that is not related to the day’s post. Trying to hijack the post and the comment thread will not be looked upon with anything but the hairy eyeball. If we warn you that you are going off the rails, please listen. We don’t block someone lightly and we don’t like having to put a reader into moderation.

Finally, the promo bit. I have a couple of new-ish books out and another coming out soon.


Foil of the Gods

Jaguar Bound — coming May 17th

Featured Image by Andreas Breitling from Pixabay

26 thoughts on “Manipulation of Data Lives

  1. “So, what are your thoughts about the poll and its results?”

    My first thought is that I haven’t worked just a 40 hour week in my entire career. Boo-fricking-hoo.

    1. Duty day starts when the pager goes off, unless it’s really going to [foul] up the schedule for the next crew. – Chief Pilot at Flying Company

      In theory, I was on 0500-1700, then a week off, then rotated back. In reality? We tried to stay close to that, but if a flight went long, or back-to-backs came up, well, “duty day started when the pager went off.” It was rare to work 200+ hours in a two week period, but during Trauma Season, sometimes it happened.

    2. I have.

      Sometimes people just won’t do what you need done in order for you to do more work. . . .

  2. Sounds like the usual FUD from a media outlet. Fear, Uncertainty, Doubt. You want clicks, FUD is a great way to get them.

    One step back, self-selected internet poll responses have essentially zero informational value. You don’t know who your sample is, plain and simple. There are also people who like to do/have time for surveys and people who don’t. A self-selected poll eliminates those who don’t.

    Two steps back, 230 people is not a large sample. Even if you knew who that was and they were all in your target group, it is not representative of a whole industry.

    So already the poll is invalid, but let’s kick it again anyway.

    Three steps, the poll seems to be composed of questions that elicit complaint. “Do you make enough money?” gets a “NO!” answer every time, no matter who you ask. Elon Musk says no to that one. “Do you want more holidays?” gets a “YES!!!” every time. I’m retired, I don’t get enough days off.

    Four steps, the poll represents the unsupported opinion of 230 people who felt the need to complete a survey. If you ask 230 people their opinion about UFOs, does it tell you anything useful about UFOs?

    I recognize all these little tricks from years of reading the medical journals. Study design, sample size, choosing the sample, those are the foundations of the study and determine the outcome. When bias is introduced at the study design level, the outcome is assured.

    Public Health is built on garbage studies like this. Almost any smoking study in the Public Health literature will boil down to a telephone or internet poll with loaded questions and zero sample control. If smoking were not so amazingly bad for your health we would never know. 90% of the studies out there reveal nothing.

    But people love to read bad news. So media has learned to give them Bad News by the shovel full. This is how you create Bad News when there isn’t any. You lie. But because it’s a “scientific poll with numbers and everything” they can pretend they’re not lying. This one appears to be so bad they don’t even know if they’re lying, it’s just a series of solicited complaints.

    Yeah kids, work sucks. You know what sucks more? Unemployment. >:D

    1. A lot of those unscientific polls are more informative than the scientific polls.

      I took an elective some odd decades ago called “Human Geography”. I thought it would provide fun insight into settlement patterns.
      What it wound up being, was a masterclass in how to “randomly select” a sample that would give the answers you wanted. (The only plausible deniability was in the catalog and syllabus. The professor was very forthright about the whole thing. And that most of the students in the course were coming over from a psych class focused on presenting questions to get desired answers, which the prof regarded as ham-handed, unnecessary, and often counter-productive.)

      Polls are very rarely intended to capture what the public thinks.
      They’re intended to influence what the public thinks, reinforced by authority and claimed peer pressure.
      The organization paying for the poll always has an agenda. The polling will reinforce that agenda, or the pollsters had better find a different line of work.

      I’m not exaggerating when I say that the nihilism and cynicism of that class put me in a depressive tailspin and permanently soured my opinion of college.
      This was not a search for Truth, but a mercenary effort to manufacture consent.

      1. Yes it absolutely is, Luke. I was fortunate to go to university in the 1970s when they were still teaching the Scientific Method and how to design a study to discover the Truth, capital T, of the thing you were looking at. My introduction to bullshit science was animal behavior. My Physical Anthropology prof was studying monkeys in Africa and their relatives exported to the island of St. Kitts in the Caribbean. He wanted to see if there was any differences in the behavior of the African vs. the Caribbean population, with an eye to identifying genetically determined behaviors. Pretty clever study design, taking advantage of that historical fluke. So we all read the studies done on St. Kitts monkeys.

        Which were -garbage-. It turns out that A) the jungles of St. Kitts where the monkeys live are utterly impenetrable to humans and B) they only come out at night. All observations were from some guy watching the monkeys steal food in the daytime, probably on the yacht club lawn. But they got published anyway, because publish or perish. Some academic on St. Kitts made a nice stipend publishing monkey stories in the little small-time St. Kitts university journal, aka Old Man’s Drinking Club.

        What I learned from that was that science is harder than most people think and journal articles are often crap. What I didn’t know at the time and only learned later in the medical biz is that people use journals to spread lies, for fun and profit. They aren’t even sneaky about it, they literally just make shit up and print it, with the full support of the journals.

        Fresh and still stinking examples can be found in the work done on hydroxychloroquine and Covid-19. It’s not an exaggeration to say that the science all happened before Donald Trump spoke approvingly of HCQ at that one press conference. All the studies ‘proving’ it didn’t work came after that. My favorite was the ‘doctor’ in California who was giving ventilator patients two (2) grams of HCQ. One study resulted in arrests for fraud, the results were being 100% manufactured by a former pr0nz actress and two other guys.

        If all you read is the abstract, 90% of those HCQ studies don’t pass the giggle test. Why do all that? Orange Man Bad, that’s why. And this is with people’s lives on the line, right? Imagine the crap they pull when it’s only money.

        Polls are -worse- than journal articles.

        1. High school debate class. Overwhelm the other side with material, and slip in undocumented or flat out false statements. If they don’t counter it, you win that point. Anyone conducting polls/studies for someone else’s money is likely doing the same. It’s helpful to be able to recognize BS when it flies across the screen.

          1. Heh. I’ve been saying this online and in real life for years about how debates prove whatever you want them to prove and never convince anyone who can look past the bafflegab. And I’ve been told for as long by the self-styled Great Minds that I’m just too dumb to know what I’m talking about. Good to see other folks got it too.

      2. Sounds like my high school statistics class.

        Create survey about what the favorite soda in school is. Stand next to Coke machine to find participants. Feign surprise at results.

    2. I like replying/filling out the polls sent to my work email. I’m the one that skews the pretty curve they’re looking for.

    3. When the result of your “poll” tells more about the person or persons conducting the “poll” than the “data” you recovered, You Have A Problem. On a larger scale, when the funding for supposed science always and only comes from politically minded sources, You Have A Bigger Problem.

      Humans being human, if there is a way to completely eradicate politics from the business of science I do not know it. But I do know that if we do not find a way to limit its influence over science, then the “science” that is produced will only be politics spoken from the mouths of men and women (formerly) of science.

      The fact that such “polls” are being produced does not surprise me. The fact that they do not receive a sound drubbing from any and all sane persons who read them is still disheartening, though.

    4. I once took the Internet Nerd test – 800 questions.
      If you took it, you were already identified as a Nerd – the only question is how high on the test you would score.

  3. I suspect you’re right, Amanda, that this survey tells us nothing. However, it does confirm my prior assumptions. I suspect that the publishing industry is full of just-out-of-college English majors who would put up with just about anything to be able to tell their former classmates, “Yes, I’m in Manhattan working in publishing!” and be able to look down on those doing marketing for a firm in Indianapolis. Why bother paying them much or improving working conditions when you know that, should any of them wise up and head to Minneapolis or Houston or someplace, there will be thousands of resumes from the next crop of graduates just begging to take their place? Add to that the fact that many of these are recent graduates who probably haven’t calibrated their expectations to the real world, and I suspect there is a lot of discontent and burn out.

    Like I said, those are the assumptions I have going in, and it’s probably dangerous to have those when looking at a poorly-designed survey that confirms those assumptions, but those are my thoughts. They’re worth what you paid for them.

  4. On the one hand, it does look like the poll was not constructed or run to illuminate.

    On the other hand, I’m seeing other indication that something has gone sideways across the board. Apparently there has been a global productivity drop, on the order of 5-7%. As in, it seems to be taking 5-7% more labor to produce the same thing we made a year ago.

    And in my corner of the world, it also feels like there’s been a drop off as well. People I used to be able to depend on aren’t as dependable any more, and I find myself having to sit on people more often to make sure things get done.

    Today I’m in office because we had a chunk of hardware that completed a test, but what I didn’t find out was it had failed the test, and needed approval to rework. Apparently it has been sitting there for a week. It critical path for our program, and its just been sitting doing nothing for a week and no-one even asked what we wanted them to do with it.

    Which means today, instead of working on the stuff I can’t delegate, I’m running down stuff that, in theory, another person was supposed to be running, and doesn’t seem to be.

    On the one hand, maybe it’s just me? But then I see the productivity numbers and wonder if this isn’t just me, but everyone?

    1. I know a lot of places that were having distance work went back to in person, even though there was an objective productivity boost when they were doing distance.

      I’ve also heard of a lot of people taking advantage of the labor market to find a better job– and a lot of their replacements are, ah, of inferior quality.

      There’s also the stress of COVID mitigation attempts.

      Just… a lot of possible vectors to make things messier.

      1. That’s the weird thing, we have haven’t done the forces back to work, and they’ve finally dropped the mask rules, but the issues I’m I’m seeing are both factory side, who never went remote, and office side, who still are

    2. Talking to my water guy is always illuminating, Harry. Two tales from him today illustrate what you might be facing.

      First, local chicken farmer had to re-do the insulation and paneling of his barn. Basic job, plain plywood sheets, plain insulation. So his -material- cost is $70,000 dollars for plywood and insulation. He built the entire barn, foundation and all, for less than that. Note that this is Canada, the land of forest products.

      Second, he was telling me about a guy who has ten boom trucks. The guy has work for all ten, all the time. He’s desperate to get the work done. Any given day, up to nine of those trucks are sitting in his lot because his licensed crane operators don’t come in to work.

      Those two items are connected. If you need to do anything that involves a crane, it’ll be hard to get it done. And expensive. The difficulty and expense accumulates and is added to the cost of -everything-.

      I’m not surprised to hear that productivity is down. You spend two years terrifying the population of a whole country, things that used to run smoothly stop running. People start to question if they’re really willing to risk death to go do that job.

      1. But, why? It seems like not showing up to work would be a good way to become a former licensed crane operator.

        What are they doing when they aren’t showing up? Working other jobs that pay more? Fishing on their savings?

        That just seems really really weird to have people just intermittently stop doing the day job, especially if they’re getting paid by the job and aren’t salaried?

        1. I suspect that it’s WuFlu “isolation” requirements, and a “positive” test is certain to come up if you try enough. After that, you’d have to look at group preferences etc.

        2. I’m seeing something similar at Day Job, and I wonder if it’s in part the uncertainty and constant background stress everyone is dealing with. Sort of like trying to swim in corn syrup instead of water. (There are other elements at Day Job.)

        3. This apparently is happening across all industries and businesses in Ontario. People just don’t show up for work.

          Some of it is getting government money to stay home, some of it is Covid-19 protocols, but most of it is they just call and say they’re not coming in.

          I’ve never understood how other people operate, most of them do things that make zero sense to me. But if I had to guess, I’d say that the Normie Universe got a very big and long-overdue existential reality check. All those people hated their jobs, and now they can feeeeel that they only get one life. I surmise they’re thinking it’d be stupid to waste it working for some company they hate.

  5. my exquisitely tiny violin just can’t be brought to play for them at this time.

    Want to see stress in a job? Go sit in at pretty much any PSAP. Understaffed? Last time I looked the national average was down near 70% staffed. Mandatory overtime. Talking people through CPR, gunshot and stabbing wounds, domestic violence, childbirth, etc. And you rarely know the outcome of what’s happening because once the officer/fire fighter/ambulance crew arrive it’s now their responsibility and you’ve got four more phone calls coming in. Cry me a river.

    Or working on a fishing trawler. Farmers, and ranchers, don’t get any days off. You might end up working 36 or more hours straight.

  6. No hijacking? This isn’t the Miami airport in the 1970s-80s?

    Burnout does happen, but I share your doubts about the poll. When I was working in aviation, we worked as long as we could, because we did have hard duty limits (aircarrier is like truckers). The spray pilots worked from can’t see until the wind got too strong, then flew in the evenings until it was too dark to see, then started over the next day at 0430 or so. You work when there’s work. Writing’s like that. If I have time to write, I need to write. Yes, some things take more of the starch out of me (academic writing), but you work until you are done.

  7. I don’t believe any poll.

    a) Most of them are badly designed.
    b) People lie to pollsters all the time (I certainly do!)
    c) They’re never random, no matter what the pollsters claim because only people who want to answer, do.
    d) Do pollsters throw out answers that don’t fit the questions in open-ended polls? You bet they do.
    e) Too many polls are push polls. That is, a respondent can only answer one way. If you don’t answer that way, it doesn’t count.
    f) Finally, a regular at Ecosophia admitted to conducting polls as his day job. He told us that polls get outsourced to smaller firms all the time (no oversight) and that when the closing time for the poll looms, but enough answers had not been accumulated, the pollsters were told to fill in the data sheets themselves. An anecdote to be sure, but it sure feels true.

    1. That wouldn’t surprise me, the “make up data so we get paid” part. Sort of like “find a way to call it a best-seller” and “find a way to keep it off the best-seller list” methods used by Trad Pub and their accountants.

Comments are closed.

Up ↑

%d bloggers like this: