Sweet Jesus crackers, it’s data time!

Over the past month I’ve been collecting data off of Craigslist; specifically, data from the Craigslist personals. This is mainly because statistics is my crack I like to flex my analytical muscles as often as possible and I’ve been in a data drought for far too long.

Anyway, enough blabbering. ONWARDS!

Data: Craigslist Personals from all 415 individual Craigslist listings* within the United States, divided into the four main categories:
WFW: women seeking women
WFM: women seeking men
MFW: men seeking women
MFM: men seeking men

The data were collected over a 13-day span (March 1 through March 13). I recorded the number of personals under each listing that were posted within the last 15 days of whatever day in the 13-day span on which I was doing the collecting. So keep in mind there might have been some post-Valentine’s Day angst-driven posts for some of the listings.

Results of interest:
1. (via extrapolation) Approximately how many Craigslist personal ads (in the United States) are posted yearly?

2. (via correlation) How highly correlated are the number of ads per state and the population of the state? In other words, do more populous states have more personal ads?

3. (via graphs) Where in the United States do homosexual personal ads (WFW, MFM) outnumber heterosexual personal ads (WFM, MFW), and vice-versa?

4. (via way-too-meticulous-digging-through-individual-listings) Are there any individual listings that can be considered “unnecessary” due to lack of posts? Are there any individual listings that should be further divided due to way too many posts?

Cool? Cool.

Let’s do this!

1. Approximately how many Craigslist personal ads in the United States are posted yearly?
Within 15 days, there were a total of 297,141 personals posted. That’s almost .01% of the US population.

Assuming there’s a fairly uniform number of personals being posted year-round, that would be a total of 7,230,431 ads per year (about 2.3% of the US population). That’s a lot of Criagslistin’.

2. Do more populous states tend to have more personal ads?
This’ll probably end up as  a “duh,” but it’s worth checking out. Maybe EVERYONE in Wyoming is posting because they can’t find one another, while everyone in Cali is shying away from personals because they’re so sick of being around people.

Food for thought.

Anyway.

Here’s a quick little graph just to give you the idea of the range we’re talking about here.

State with the fewest number of ads: North Dakota (201 ads in 15 days)

State with the most number of ads: California (46,016 ads in 15 days)

To check if there’s a correlation between state population and number of ads, I ranked the states by the number of ads and also by the population, then ran a Spearman rank correlation on the two rankings (non-parametric statistics FTW).

r_sp = .837

That’s a pretty high correlation, I don’t care who you are. So yes, the higher a state’s population, the more ads they are likely to have on Craigslist. Durh.

3. Where in the United States do homosexual personal ads (WFW, MFM) outnumber heterosexual personal ads (WFM, MFW), and vice-versa?
This was an interesting one that didn’t quite turn out as I expected. For one, there were more ads in the MFM section than any of the other three sections for pretty much every single listing. This brought the total of the homosexual ads well above the total of heterosexual ads for most listings. That alone was surprising to me.

What’s even more surprising, though, is the pattern of homosexual- and heterosexual-dominated ads by state. Here’s a map that breaks the states down by the ratio of homosexual ads to heterosexual ads.

In order to make keying this thing easier, I centered the ratios at zero, where zero indicates a ratio of 1:1, negative values indicate a ratio of more than one heterosexual posting for every homosexual posting, and positive values indicate a ratio of more than one homosexual posting for every heterosexual posting. I color-coded the map by creating six intervals on either side of zero, with each interval increasingly more imbalanced (fewer/more homosexual postings per heterosexual posting). Therefore, the more intense the colors get, the more imbalanced that state is in terms of the ratio of homosexual to heterosexual postings. I’m dumb and lost the original ratios, but they ranged from .292:1 (.292 homosexual posts for every heterosexual post; South Dakota) to 3:1 (3 homosexual posts for every heterosexual post; Washington, D.C.). States that have more homosexual ads are a deeper red; states that have more heterosexual ads are a deeper blue. States that have a near 1:1 ratio are white.

Can any of you dudes see any sort of demographic that this pattern follows? I was thinking that maybe the ratios followed the red/blue states, but that doesn’t appear to be the case. I also thought that maybe it would vary by general geographic region, but that doesn’t appear to be the case either (except for the Northwest, which is pretty “neutral” overall). Interesting stuff.

4. Are there any individual listings that can be considered “unnecessary” due to lack of posts? Are there any individual listings that should be further divided due to way too many posts?
This wasn’t as tedious as I thought it’d be…just basically involved going back through the data for the individual listings to see if there were any that had HUGE amounts of ads or any that had virtually none.

Some points of interest:
The average number of ads posted per listing was exactly 716.
Pierre, SD had only three ads posted.
New York City had 23,122.
WFW had the fewest ads overall (7,923 for the whole country), while MFM had the most (191,753).

Cool stuff, eh?

*when I say “listing” I mean things like “Pullman/Moscow” under Washington’s state or “Rockford” under Illinois…all the individual cities/towns/regions. When I say “personal” or “personal ad” I mean things like “Good Man Wanted” under Tippecanoe’s WFW section or “SEXCAPADES” under Boulder’s MFM.