Farewell, Daisy Shipton

This is more of a quick, explanatory “heads-up” post than anything else.

On March 31st 2018, I started an experiment: I created a new Stack Overflow user called “Daisy Shipton” with no picture and a profile that just read “Love coding in C#” (or similar). I wanted to see how a new user presenting with a traditionally-female name would be treated, while posting the same content that I normally would. This experiment was only a small part of my thinking around the culture of Stack Overflow, and I expect to write more on that subject, touching on the experience of “Daisy”, at another time.

I let a few people in on the secret as I went along – people who I fully expected to recognize my writing style fairly quickly. A single person emailed me to ask whether Daisy and I were the same person – well done to them for spotting it. (Once someone had the idea, the evidence was pretty compelling – the “Jon Skeet” account went into a decline in posting answers at the same time that the “Daisy Shipton” account was created, and Daisy just happened to post about C#, Noda Time, Protocol Buffers, time zones and Google Cloud Platform client libraries for .NET. I really wasn’t trying to cover my tracks.)

As Daisy reached a rep of about 12,000 points, there became little point in continuing the experiment, so I asked for “her” account to be merged into my regular one. So if you see comments on my posts referring to @DaisyShipton, that’s why.

There’s one aspect of experimentation that never happened: Daisy never asked a question. Next time I want to ask a question on Stack Overflow, I’ll probably create another account to see how a question I think is good is received when posted from a 1-rep account.

It’s been fun, but it’ll also be nice to only have one account to manage now…

26 thoughts on “Farewell, Daisy Shipton”

  1. “I wanted to see how a new user presenting with a traditionally-female name would be treated”
    What was you conclusion?


  2. I’m looking forward to seeing your final conclusions on this. Can you commit to any kind of time frame on when this fuller post might appear?


    1. No, I’m not going to commit to anything. Between regular work, conferences, user groups and finishing the 4th edition of C# in Depth, I’m extremely busy – and a blog post like this takes a lot of effort to get right. I might hope to have it done before the end of 2018, but I’m not going to commit to that.

      Liked by 1 person

  3. What were you trying to achieve/prove? I personally never bothered with user’s name/gender… I just want to solve my problem… whoever helps me get my vote…

    Liked by 1 person

    1. I was trying to experiment as to whether your approach is the one of most users. (Note that even if we’re not conscious of bias, it can still creep in subconsciously. I’m not saying you do treat people differently on SO, but that it’s possible. Unconscious bias can be really hard to fight.)

      Liked by 2 people

  4. Hey Jon, you seem like a good guy and you seem to mean well. This post isn’t very useful though sorry. Perhaps you should do a more scientific study if you aim to learn something. Cheers.


  5. Obviously we won’t get to see your conclusions for a while, but I think for this to be at all scientific, you should have also made a new “Eric Johnson” or something as a control to compare against. I’d imagine that most people in the C# SO community know who you are (the banner displayed across the website for a week or two when you broke 1m rep may have helped :P) which likely influences the way people respond to you (assuming you’re using yourself as the male control).


  6. Oh ! once I argued with Daisy for pinpointing a small error in my answer retrospectively (1+ year old) and down voting it. Finally corrected the mistake to get Daisy revert her down vote and from that day onwards I am wary of when Daisy will strike again, but it never happened and now i Know :)

    Liked by 2 people

  7. Experiments are usually designed to test a hypothesis, so what hypothesis are you aiming to test here? Presumably, something broad like “Female-identified posters are treated differently” or even something narrower?

    Liked by 1 person

    1. Female-identified, and also low rep. Experiment definitely makes it sound more rigorous than intended, but outside a scientific context it’s the right word. (People experiment in cooking for example, without ever considering a hypothesis to test…)

      Liked by 3 people

  8. I have had horrible experience asking question with low rep in StackOverflow. I only asked serious well formed question and yet got the seal for asking questions too poor to let me ask more. It was genuinely inexplicable to me. And completely unacceptable. Completely.

    Well, that’s the past. I have taken that as a challenge and started https://sharedmem.com.


  9. Hey, we’re still waiting for that “in the broader context of a larger post later” 3 years later. What happened? Did the tech world not turn out as horrible as you had thought? That’s a good thing you know, it’s worth sharing too.


    1. Yup, unfortunately I don’t think that larger post is ever going to come. I did write one… but discussing it with the Stack Overflow team, it was too easy for folks to read into the post whatever they wanted to see.

      I’ll say that I didn’t experience any absolutely blatant sexism when posting as Daisy, but I don’t want to draw too many conclusions from that.


  10. I can only imagine the response that a white man posting “there isn’t much sexism there” would receive, whether or not he was right.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s