How did you learn to code? | Hacker News Reader

How did you learn to code?

caseorganic.com

46 pointsaaronpk37 comments

For improving the survey:

I don't have the sources (a few hundred papers / articles) anymore, rather this was for my own research in designing surveys for companies garnering feedback from IT Director (mostly) levels.

For buy-in, Saliency of Questions and Subject Matter is key to engagement (finishing the survey).

Assume that respondents will fade towards the end of the survey. Your most important bits of information that you want to get from respondents should be towards the beginning after a few "warm up" questions, so you get the data you want without the drop-off.

Assume each question that is not open ended is ~ 1 minute:1 question. This will help you determine length. Most engagement drops off ~10-15 minutes. These questions INCLUDE Name, email, etc. (also, you may want to drop the name req., as other commenters are pointing out).

Surveys of 20 minutes can be very "long" and have a much higher chance of not finishing the survey without interest-engagement (video, graphics, saliency of questions).

Research showed that length doesn't necessarily drop engagement, but poor questions (relevance / saliency) does.

You will want to tighten up the questions, reduce the # of open-ended questions (how are you going to compare all of those?)

You could instead make radio-text boxes (seeing you are using Google Forms), such as: What did you first make the computer do: Option 1,2,3,4,5, Other (as a text line).

What language could also be radio-boxes (to shorten respondents time to type).

You are also requiring random open-ended questions to be completed (red asterix), or the form will not submit. You will lose much engagement this way.

Also note why you are gathering this (I assume for schooling), which would perchance help with privacy concerns (though you don't need anyones real name).

Thank you. Initially I sent this to people I knew, and the responses were to be included in a book to be distributed to people interested in hearing stories from programmers. Some of this is mentioned on the survey itself.

I'll make a new survey with less questions that address your concerns and takes less time to fill out.

I did not answer this survey, for the following two reasons:

- there are a lot of open-ended questions. It's probably very long to fill in.

- required full name and email are a showstopper for me. Having a required email is ok of course, but why do you make it required ? And what about the full name ?

I believe that lots of people will share these concerns so I hope that it can help you tune your survey so that it's more successful.

I also stopped reading at the full name.

Thank you for your feedback! Made the full name optional and added an anonymous checkbox. All responses will be made anonymous by default.

Your anonymity checkboxes need to be converted to either a single checkbox or a set of radio buttons. Right now you can select either or both options, which doesn't make sense.

Thanks, good catch.

I entered my full name as i don't mind if you email me, but if published somewhere, anonymity to some regard would be nicer.

Names can be kept anonymous when the results are posted. It's definitely designed to get long-form answers. If you are interested in a shorter poll version I'd be curious to hear what you were expecting!

Surveys like this should always be fully anonymous. If the name isn't published, why is it collected? Asking hackers to provide their full name is like cryptonite.

As for a shorter survey - The form itself could be a lot more inviting. For instance, the question "Did you know anybody else your age that programmed" is a simple Yes or No. Or does it want me to give away the personal details of that person as well?

can be kept private. Ideally we can publish the results along with your name.

Think about this as questions that would end up in a book like Programmers at Work. http://www.amazon.com/gp/product/1590597141

Well, we do have Coders at Work.

http://www.codersatwork.com/

I filled out a few of the answers very thoroughly then realized it was just too much. With the red asterisk of requirement, I lost interest.

You have to scroll all the way to the end before you find out what the copyright terms are: a Creative Commons License that allows commercial use of your replies. The fact that this is a potentially for-profit enterprise should be clearly disclosed before you begin the survey.

Here, for convenience, are the license terms:

By submitting this story you license this work under a Creative Commons Attribution 3.0 Unported License.

http://creativecommons.org/licenses/by/3.0/deed.en_US

Which says:

   You are free:

    to Share — to copy, distribute and transmit the work
    to Remix — to adapt the work
    to make commercial use of the work

The survey is very focused on "firsts", like, "what was your first program?". It's weird, would you ask a marathon runner what their first steps were like?

A lot of people have dabbled with code, and the first programs are always the same. If you started in the BASIC era like me, it's

    10 PRINT "NEILK"
    20 GOTO 10

For others it might be making a button in VB or doing some little bit of JavaScript interactivity. My point is, there's no difference between the first program of someone who gives up after thirty seconds, and the first program of someone who sticks with it for thirty years.

I've seen people ask 'what was your first race', which I think is a better equivalent. Answers vary from "high school track" to "a 5k walk for breast cancer". Potential answers to "what was your first program" range from your BASIC answer to my own "a quadratic equation solver on my TI-83". I find it interesting, and when you are aiming to use the survey responses to tell stories, I think you'll get some good material out of these questions. Even reading the other comments here get you a lot of different 'first programs'.

There very well could be. The most likely difference between someone who tries gambling and gives up after 30 seconds and someone who becomes a gambler for life is that the latter is far more likely to have done well in their first game. That's where the notion of beginner's luck comes from — it's a selection bias where people who are lucky at first are more likely to continue gambling. I don't see any reason why a similar selection bias couldn't exist in programming. For example, perhaps people who start programming at a very early age are more likely to continue programming. This survey won't prove anything, but very well could point in the direction of some interesting correlations.

Showoff.

My first program was

10 PRINT "ART IS COOL"

My second program added

20 GOTO 10

I'm half as fast as you are. :(

I first learned to program in BASIC on an Apple ][+ when I was about 10, but never really followed through with it until I re-taught myself in my 20s, starting with "Learning Perl". So, I wasn't sure how to answer;)

My experience is really similar to yours. I just kind of worked both in there...

Requiring my full name and cities where I grew up caused me to become wary and leave. That information can be used too easily on stupid sites that have account "security" questions like that.

Personally, I never answer those account "security" questions with any kind of information someone could find out easily via a social network, random public documents or even social engineering attacks.

"What is your mothers maiden name?" "correct horse battery staple".

As a side bonus, it also makes support calls very amusing.

Survey asks too many personal questions for no reward or incentive.

My reward was remembering things I haven't thought of in years (I remembered halfway through about how satisfying moving the turtle in logo felt, or that I had almost completely forgotten about the hypercard stuff I played with in middle school).

Hah! I completely forgot about my high school programming stuff!

I'm with you, the reward was remembering all the cool shit I did (started programming at 4 on a C64), and the hopes that the info might be useful to people starting out. I definitely wish there was less of a barrier to becoming a programmer these days, although I think interpreted languages go some way here.

This.

Recounting bits of my history...well, I could go on for days. Writing software is just what I do and enjoy.

My first program was in Python. I was in grade 10 and math homework consisted of repetitive questions with the x/y/z changed.

I got tired of doing it on the calculator, so I put in the general equations and just updated the variables. It felt good, almost like I had a superpower.

I learned in Stunt Island! Best game ever!

http://www.youtube.com/watch?v=T0YcO25-tKU&noredirect=1

I actually forgot until recently that C in high school (self-taught) was not the first exposure I had to programming. It was HyperCard! Granted, about five years ago I actually found the original Mac I'd had and discovered I was a truly, truly awful programmer in my first few months, but it did bring back a lot of welcome nostalgia.

You know about this site, right? http://ilearnedtoprogram.com/

Wow! Exactly what happened to me :)

You realize that link returns a random response, right? I'm not sure what you're referring to.

I mean in general, there are about 500 responses there. You can get permalinks for all of them. (After doing a lot of page refreshes—and seeing a lot of amazing stories from interning at the Swedish Defence centre to being inspired by Gloria Steinem—I found mine http://ilearnedtoprogram.com/?id=4db7598635a76c8952000000 )

I actually enjoyed it. It was the first time I sat down and thought about that journey and I was quite happy to share it. I don't mind about sharing my name or other such information. Never bothered me. I find it easier sharing my real name than trying to come up with a clever username.

Too many personal Q's on the survey... so I laid it out here.

On a TRS-80 and the first program I wrote was a Pong knock-off in (GWBasic? can't remember)

PEEK & POKE anyone?

Heck, if I could remember even half of the things this survey asks.

...in HTML on GeoCities, because I wanted to make something cool.

Reply on news.ycombinator.com