Behind the Scenes at Atheist Shoes

Last week, the makers of the Berlin-based Atheist Shoes reported that packages marked with “atheist” packing tape weren’t getting to their American destinations as quickly as packages without the tape.

Before that study came out, eNtR berlin got a sneak preview of the Packing Tape Experiment. They also spent some time behind-the-scenes with the company and you can watch the second video to find out how they operate:

About Hemant Mehta

Hemant Mehta is the editor of Friendly Atheist, appears on the Atheist Voice channel on YouTube, and co-hosts the uniquely-named Friendly Atheist Podcast. You can read much more about him here.

  • DougI

    The picture shows the tape was placed even poorly than I had previously imagined, right near the address. No wonder the machines weren’t able to read the address. They didn’t even have the return address in the correct location. At least the company admitted their research methodology was highly flawed.

    Simple to follow guidelines from USPS:
    “Tape the opening of your box and
    reinforce all seams with 2-inch-wide
    tape. Use clear or brown packaging
    tape, reinforced packing tape, or
    paper tape. Do not use cord, string,
    twine, masking or cellophane tape.
    Place a strip of clear packaging
    tape over your label to prevent
    the address from smearing.”

  • AtheistShoes

    Hey dougl, we ran the packages past the a post professional before shipping them – they assured us the lettering was too big to interfere with the machines’ reading of the addresses… and the successful on-time delivery of many of these packages would indicate that there way no systematic problem in reading addresses. Use of lettered tape is also pretty widespread in commercial shipping, despite the USPS guidelines… more detail on what we did and why we did it here:

  • AtheistShoes

    And our research was certainly not “highly flawed”, with several academics and statisticians stating that our research design was in fact robust. Again, read the comments… there were certainly ways in which the study could have been improved with a couple more controls, but our findings are valid and we stand by our conclusion of a likely explanation for the evidenced bias being attitudinal discrimination by some workers of USPS.

  • AtheistShoes

    Thanks for posting Hemant!

  • DougI

    And yet you said you shipped them DHL and you said when you used the correct tape, per the guidelines, you didn’t have a problem. So what’s more probable, the machines didn’t read them or there’s a massive conspiracy to slow down your mail by programming computers to delay packages with the word ‘atheist’? Never in my USPS training did they mention to misdirect Atheist mail.

  • DougI

    That is utter bullshit. Who are these professionals you talk to? They clearly can’t design a proper study since you clearly couldn’t. You never controlled for discrimination between the type of word on the label or wording in general. It’s such a utterly huge flaw in your study that you can’t possibly conclude that the delays were a result of prejudice. It’s no wonder your “study” was an infographic and not published in a journal, it would have been easily rejected for publication. The only result that can be concluded is that you’re simply projecting your own biases of Americans as being a bunch of backward, intolerant, religious hicks.

  • AtheistShoes

    I don’t think I follow. We have had no problems in Europe with ATHEIST tape, all packages arrive at the same time as those with neutral tape. Also, given we’re not in the USA, we have to ship initially with DHL who then hand over to USPS (no delays or disappearances in the DHL leg of the journey in over 600 shipments sent to the US). And nor are we claiming there is a policy at USPS to delay or “lose” atheist packages – instead we believe it’s highly motivated individuals operating against USPS policy. If you’re the same Douglas who has already commented extensively on our study before today, as well as emailed us about it, then we’re going over old ground somewhat. I understand, as a USPS employee, that you might find our conclusions uncomfortable, but I again refer you to the testimony of other USPS employees, in response to our study, who have claimed discrimination is not only the most likely explanation, but that they have seen such attitudinally-motivated tampering, by their colleagues, happen before.

  • Gus Snarp

    Sorry, I agree with dougl on this. Your research WAS highly flawed, as you missed a very important control. That missing control invalidates your results, no matter what your several academics and statisticians said. People make mistakes. Your study looked impressive at first glance, but upon review some subject matter experts raised an important concern about a potential confounding variable that you did not adequately control for. And now you’re hiding behind special pleading. I’d have a lot more respect for you if you could just admit that you missed something. Now it’s certainly still possible that bias is at play, but we simply don’t know. It’s an unfounded assumption that could have been easily ruled out with adequate controls.

  • DougI

    Your opinions aren’t fact and they aren’t something that you can conclude from your highly flawed study which appears to be more of a way to get some free publicity.

    So riddle me this, in facilities that operate 24/7 with multiple shifts you’d have to conclude that there must be some coordination to delay your packages. Doesn’t that sound absurd to you? You’re just a conspiracy nut on par with the 9/11 truthers.

  • AtheistShoes

    Thank you for remaining polite and balanced in your views.The academic write-up of this study is in progress (it takes time when you submit for publication in a peer reviewed) and the academic I am speaking of are social and cognitive psychologists, in Europe and America, one of them a professor at a leading US university. And they are very well placed to comment on the legitimacy of the study. Our own backgrounds are also in social science, and the statistician on our study has a PhD and conducts research like this, for a living, on a daily basis.

  • DougI

    Funny, I have a background in psychology and I managed to catch your flaw. Maybe you ought to find better academics, or is the “argument from authority” your only defense now?

  • AtheistShoes

    We have said we would like to replicate the study with the control in place, and that this would improve the research. But there’s no question the evidencing of the bias is robust. The explanation for what causes the bias is more open to question. And, given the controls in Europe and the huge unlikelihood that a sorting machine’s possible differential processing of packages with worded tape would lead to their absolute elimination (we’re not just talking about delays) discrimination and employee tampering remains the most likely explanation… in fact the only reasonable explanation anyone has put forward to date. The absence of the control does not invalidate our findings, and it’s sensationalist to say that the study was “highly flawed”. But we do look forward to replication.

  • AtheistShoes

    Again, you’re misreading what we said and jumping to the conclusion that we’re talking about a miscreant policy on the part of USPS – we’re not – we’re instead concluding a likelihood of a few individuals operating alone and in isolation from each other.

  • AtheistShoes

    None of the academics I’m talking about have made the sweeping and absolutist statements you have, nor erroneously insisted that we’re concluding an institutional policy of discrimination, when we’re not… rather they’ve been more balanced and, thankfully, independent in their assessments of our research, and have read our conclusions correctly.

  • DougI

    Yeah, I know about your confirmation bias but in a system with redundancies it’s highly unlikely. You’ll have to find a processing center filled to the brim with people who so utterly hate anything with the word Atheist on it that they’d be willing to risk their job and be punished with a felony charge. I’d just be shocked that an employee would take the time to read the envelope, usually we’re too busy putting them on the belt.

    Alas, since you didn’t do a proper study there’s no valid reason to form the conclusion that you did. Clear bias and a flawed study, I wonder who these “academics” are that would stain their name by attaching it to such a study.

  • DougI

    So who are these academics who can’t notice such a flawed study in front of their face?

  • AtheistShoes

    Well, like I say, other USPS employees have said that they consider it quite likely that their colleagues would tamper with post. So we have to accept your view on the matter is not universal at USPS.

  • DougI

    Oh, I said all million employees were in agreement? Looks like you have another case of seeing what you want to see. I thought you were doing an objective, well-controlled study and not suffering from a severe case of confirmation bias. This “study” appears to be nothing more than a very successful viral ad campaign. Good job.

  • zeggman

    Obviously a tardy April Fool’s joke. That “German” shoemaker doesn’t sound at all like Colonel Klink.

  • abb3w

    Explicit coordination isn’t required when common behavioral expression of endemic ethnocentrism will suffice. Or in plain English — you don’t need religious bigots working together, you just need a lot of religious bigots abusing the power of their jobs in similar ways.

    Delaying or discarding mail does not seem to require much imagination for a postal worker to come up with. The penalties if caught are fierce, which acts as a disincentive; contrariwise, hostile attitudes toward atheists by the religious are pretty widespread, and part of the fierceness is to make up for how difficult it is for breach of the rule to be proven.

  • abb3w

    Have the results been reported as a formal complaint to the U.S. Postal Service Office of Inspector General?

  • AtheistShoes

    Yes, we have reported them to the Inspector General of the USPS, but we await a response.

  • AtheistShoes

    Worth also saying that we’ve asked them to help us replicate the study, on a more expansive basis… if replication is possible given the publicity… but we’re hopeful they’ll see getting involved as a positive opportunity.

  • DougI

    It would require three different people over three different shifts, or maybe more if a supervisor happens to walk by. Keep in mind this has to be in multiple cities throughout the nation. It’s either this, or the machine had a problem reading the package because of the tape which is a known problem that’s why the USPS asks for certain tape to be put on the packages.

    It’s a matter of plausibility and the conspiracy theory doesn’t sound plausible especially when the “study” says when they used the correct tape the problems decreased.

  • Gus Snarp

    No, the postal scanners are a reasonable explanation. Your conclusions are hasty, overly broad, and based on a serious methodological flaw. You didn’t control for an obvious variable. From a scientific viewpoint, that’s highly flawed and there’s nothing sensationalist about that. For a study done by a shoe company trying to figure something out, it was a great job, but for scientific research, it’s highly flawed and no conclusion can reasonably be reached concerning discrimination, bias against atheist branded packages (as oppose to just branded packages in general), or tampering with the post, all things that you allege in your study graphic. What you’ve found is that packages shipped from Germany with lettered tape through DHL with USPS local delivery had more shipping problems than packages without lettered tape. This shows a bias against lettered tape, not necessarily against atheist tape.

    I don’t think, based on this research, that you should be leveling accusations of “USPS Discrimination Against Athesim” as the title of your infographic does (don’t get me started on the question mark as legal disclaimer, that doesn’t fly with me). You should edit the graphic to accurately reflect what you’ve found, without the sensationalism you’ve used, and to acknowledge the flaw in the research.

    You say you want to redo the study, so I suggest you do so if you want to make such serious allegations. I would recommend some changes to control your costs. I would find a U.S. partner to do the shipping and eliminate DHL, since you’ve got a pretty good case that they’re not at fault (not perfect, but I’m trying to reduce the cost). Then have your U.S. partner send empty envelopes, or envelopes padded out with a bit of bubble wrap, with atheist tape and tape that says something like “ALTHLIS” (a meaningless word with similar appearance to simulate the effect on a scanner) in the same font, type size, and placement on the package through the USPS.

    I understand this is still rather expensive, but you really either need to stop promoting your sensationalist interpretation, or spend the money to fix the study.

  • baal

    No, they are a real company. Your doubt is misplaced.

  • baal

    While there are always better study designs, you’re a relatively small company with limited resources and have a reasonable basis for further study (or not). I don’t understand why some of the commenters here are so emotional over the question. In the greater context of advertisements and lying corps, this case looks like small beans (if beans at all).

  • Kevin_Of_Bangor

    They look like bowling shoes.

  • AtheistShoes

    Gus, you really should read the comments we’ve already made on this page Extensive controls in Europe allow us to eliminate any cause of bias outside of the US. 600 cases of shipments to the US where only 1% of packages were delayed by US Customs allows us to isolate the problem as happening while packages were in the custody of USPS. The fact that we have had zero delays or disappearances of ATHEIST taped packages vs neutral taped packages in the EU would suggest that the mere presence of wording on the tape is not an explanatory variable for delays (although admittedly we have had to assume that US postal sorting processes are not radically different to European). And I await your explanation of how alleged selective processing by postal scanners have removed 9 packages from the postal system entirely, with no sign of them for 4, nearly 5 months after the study was conducted. We’re happy to admit the experiment was not perfect and that it could have been better designed had we greater financial resources, but there were more controls in place than you are taking into account (our fault we only list detail of these in our comments, we couldn’t cram them all into the infographic, but we look forward to reporting them when we do write this up properly) and our conclusion of a likely explanation (which is not “discrimination by usps”) is far from sensationalist when you take all these factors into account.

  • AtheistShoes

    to be clear re my last point – we are specific in our write-up that we believe the most likely explanation is discrimination by individuals in the employ of USPS, and not a policy of discrimination by the institution. i apologise if our headline suggests we are concluding the latter.

  • Marco Conti

    Ok, so the tape might have interfered with the address reading machines and so forth.,

    Do me a favor: the next control see if you can find a packing tape that says “JESUS” in approximately the same size font as the one saying “ATHEIST” and use that as a control. Make sure the tape is applied in the same exact fashion and let’s repeat the experiment.

    Also, let’s not forget what prompted the experiment in the first place: Boxes with Atheist on it were arriving late or not at all. That wasn’t them imagining it and I presume they recognized the pattern due to their experience in shipping other products.

  • Gus Snarp

    Discrimination by individuals within the USPS is still a hasty conclusion not supported by this evidence until you experimentally rule out other variables. It’s conjecture.

  • Gus Snarp

    Obviously we disagree. I stand by my position as stated previously, even in light of everything else you’ve said. Obviously, like me, most people aren’t going to scroll down to even see that there are comments on the page, so a few comments at the bottom hardly makes up for the sensationalist headline still being there. But I recognize that infographics aren’t cheap either.

    I would also point out that while your sample size is adequate for evaluating delays, with only 9 missing packages I don’t think we can reasonably rule out randomness. Not that we have to, increased hand processing can lead to increased package losses as well as delays, so there’s nothing unexplainable about the lost packages.

    Since you’re in the mood for answering questions today, could you tell me what you know of processing equipment in the U.S. versus other countries? Do they all use the same automated scanning systems? Are the same ones used even across the USPS? If we don’t know that, then the fact that they suffered no delays elsewhere still doesn’t rule out printed tape as the cause.

    Also, you mention other controls, but you’ve never, anywhere, given any indication that you used any other printed tape, and you didn’t in the video. That’s the control that’s needed here, unless you have others that are somehow relevant to that question and eliminate the need for that control?

  • agnosticscroller

    scrolling agnostic guy with science background here. whilst not published in a peer reviewed journal, the experimental design does seem robust (statistically speaking). of course youbare entitled to your opinion.

  • chicago dyke

    good gawd, Dougl, relax. you’re making this website look bad. it’s a shoe company, a youtube, and a couple of blog posts. if you insist on peer reviewed university standards, go read the Journal of Statistics or something. sheesh.

    we should be welcoming and assisting our European atheist friends and allies, not nitpicking them to death like trollish children. and insulting them with childish language, hmmm. i don’t remember that being part of a scientific analysis. grow up already.

  • Feminerd

    I had been skeptical of the study design, given the doubts raised by people about non-plain tape. AtheistShoes, thanks for the additional information about how you designed your study and what steps you used to control many of the potential confounding variables. While I, like many here, would love to see additional studies to verify what you’ve found, with control tape saying JESUS or RANDOM or something, I’m impressed by what you’ve done so far. I also understand that these are expensive studies for you to do.

    I’ll not say I’m convinced, not yet. But the evidence is pretty damning so far, and if replicated, I’ll go over to “convinced”. I’m just sad that I have to come to that conclusion; it says such unpleasant things about some of my fellow countrymembers.

  • Kevin Kirkpatrick

    First of all, I’ll side with the critics in saying that the study’s flaw jumped out at me the moment I saw the plain white package next to the “ATHEIST” labeled package. How did it not occur to you to put innocuous labeling on your control – preferably some unoffensive* / meaningless anagram of ATHEIST like “HI STATE”? Seriously, if anyone held up two packages, one covered in labels (any label!) and the other plain, and asked, “statistically, which is going to ship faster?”, the answer seems obvious – so much so that (sorry) the whole thing smelled like a pre-loaded experiment designed to get desired results.

    That said, even if (based on the difference between European and American results) we do grant deliberate intervention, I strongly suspect that any such intervention is more likely due to counter-terrorism/drug-dection policies. Particularly after the anthrax scare, it is my understanding that US postal service employees are routinely drilled to be on the lookout for suspicious packages in a way that European postage workers are not.

    From the USPS website:

    Though certainly not 100% rational, it seems entirely plausible that an employee might see a package marked “ATHEIST” over-and-over-and-over in dramatic fashion and think, “Hmmm, I wonder if [some Catholic nutter / Islamic Terrorist] is using the postage system to exact revenge on [PZ Myers for harming a cracker / Newspaper editor for drawing Mohammed]? Think I’ll report it; better safe than sorry….”.

    *TIL ATHEIST is an anagram of “EAT SHIT”

  • DougI

    I see, ignore scientific standards when we want to promote an agenda. Thanks for that bit of advice.

  • abb3w

    I’m not seeing what in the nature of post office operations would give rise to delay of a single particular package requiring multiple people.

  • Molly

    Really, because my Stat class, just the other day, used your study as an example of a poor experiment design. Too many variables, no effective control, no looking for other causes, and assuming that correlation=causation.

    For example, you said there was a 37 day delay in Michigan. You didn’t account for the fact that the postal services in Michigan only have a couple processing centers for the entire state, combine that with weather conditions during different times of the year and other factors, mail tends to be late. No big deal. Also, you taped up your packages in a way that isn’t advised by UPS.