NUF Survey Results - Introduction

Part 1 of the NUF Survey results. Table of contents at the bottom.

INTRODUCTION
These are the results of my analysis of survey data collected from NUF users. The survey consisted of 80 questions on a variety of topics; when the survey was initially posted there were a couple of additional questions, however, I deleted them about 5 minutes in after deciding that they could not provide me with usable data and none of the responses to those questions, if there were any, were recorded. I'll discuss some of the mistakes I made with the question design in the relevant sections. A couple of the questions were inspired by the OkCupid stats blog. It's honestly pretty amazing and I'd recommend reading it if you've got a casual interest in social statistics, especially the posts written by Christian Rudder. There was also a little metadata provided by the site mostly regarding the respondents' OS and browsers, but seeing as that wasn't something I was initially aware of or had informed respondents was being collected I'm not going to refer to any specific figures, though I will bring it up very briefly for some specific trends in the Meta-Survey section. Otherwise, the primary take away from it seems to be that Windows and Android are still much more popular among NUF users than Apple.

In terms of statistical figures I'm including in this post, the most complicated I'm going to get is including some percentages. Sure, that wouldn't fly if I were writing a research paper, but I'm not a scientist, I'm a bored member of the housing industry. As such, while I'll try to make sure my use of the words "significant" and "correlation" at least somewhat follow the accepted definitions (though I did occasionally change from using 99% confidence to 95% confidence or even 90% confidence because I’m a right thug…well, because I wanted more interesting correlations to put in), my use of SPSS is like a newly blind man going into the Guggenheim Museum to use the bathroom: I'm desperately flailing around an oddly shaped building trying to find one specific function without the means or desire to appreciate the majority of that building's fixtures. In any case, if I say “strongly correlated” I’m talking about the significance (>99%) rather than the degree of correlation. When relevant, I try to compare the survey results to results of similar polls conducted for the general global population, however, in many cases, the US is the only or the largest country for which accurate polling data exists. In addition to that, I myself was born and am currently living in the US, so even if I spent the better part of a decade living in Asia, I’ll have to preemptively ask for forgiveness for any regional chauvinism that may show up.

I don't have an exact figure for how big the active NUF community is so it's a little difficult to estimate the accuracy of the data, but I think I've gotten a large enough number of responses to at least make it somewhat representative. The major problem I can think of is that exposure was limited primarily to the General Chat subforum, so perhaps this would more accurately be described as a survey of the General Chat users than of NUF users. *shrug* I'll try to write some reflections about the limitations and possible improvements to the survey later.


It seems like the full results were too long to be included in a single blog post, so here's a handy table of contents to which I'll be adding links once I make the posts.

• Introduction (You are here)
• Demographics
  • Location & Language
  • Age
  • Gender, Sexuality, & Romance
  • Education & Employment
  • Religion

• Meta-Survey
  • Useless Questions
  • Bad Faith
  • Declining to Answer

• Lifestyle
  • Novel Preferences
  • Recreation & Exercise
  • Family
  • Friends & Pets
  • (Yet Another) NU Section

• Personality & Society
  • Happiness
  • Self-Harm & Psychiatry
  • Self-Evaluation
  • Social Views
  • Bonus Round: NUF Dating
  • Experiences
  • Preferences

• Survey Reflection