Tuesday, April 13, 2010

Queries, Darn Queries, and Statistics

I majored in Mathematics Education and Economics in college, so I spent a lot of time with numbers. And as much as it pains me to admit it, some of my favorite classes--please don’t think less of me when I write this--were statistics classes. Honestly, there is nothing more satisfying than doing a little research, running a regression, and seeing those coefficients, with their respective p-values, pop up on the screen. I swear, it’s almost magic. Something very much like finishing that final FINAL draft:)

No regression analysis here today (I wouldn’t want to scare too many of you away), but I do have a lot of good, old-fashioned graphs and charts. Not the graphs and charts I’d like to have, mind you--Excel isn’t quite as robust as it ought to be--but they’ll work well enough. I just hope none of my old professors happen to drop in.

First, a disclaimer (in case any of you happen to be closet statisticians): My bar graphs should be dot plots. I know, I know, I know. But Excel doesn’t do dot plots, so bar graphs it is. And Mr. Gates was too busy, apparently, to program Excel with an actual calendar, so all my calculations assume 12 uniform, 30-day months instead of 12 28- to 31-day months. AND Excel insists on starting at one day instead of zero days (yeah, some agents are THAT fast), so everything is one day off. But everything was already a little off, thanks to those 30-day months, so…yeah.

You should also know that I sent every query by e-mail (or online form). And when I say query, I mean my initial contact with the agent. It doesn’t matter if that initial contact involved a one-page letter, or a one-page letter, five-page synopsis, and first three chapters--I consider all of those to be queries. (Actually, one of the agents I queried wanted to see the entire manuscript in the initial contact.)

All right, here’s the first graph, Response Times (Rejections). This graph depicts the frequency of query response times (how many agents responded in zero days, how many agents responded in one day, and so on), but only for those agents who responded via rejection.


I realize it's a little fuzzy, but the shape of the graph is more important than the numbers themselves, anyway. Encouraging, isn’t it? For the most part, agents respond to e-mail queries rather quickly, especially with rejections:) I cut the graph off at 90 days, as I officially catalog it as a non-response after three months. But I do go back and fill in dates if a rejection comes in after the fact, so there are a few outliers not included in this graph: at 116 days, 122 days, 239 days, and, the big winner, 263 days. Wowsers.

Now for the positive responses, the partial and full requests. This graph, Response Times (Requests), depicts the frequency of query response times for the agents who requested more material.


Now, obviously, there are a lot less data points, but the overall shape remains the same. Again, most agents respond to e-mail queries within a couple of weeks or, at most, a month.

To get an even clearer idea of what’s going on, check out this chart. Here, I’ve broken the data into quartiles. The minimums and maximums are exactly what they sound like--the lowest and highest data points, respectively. The median is the data point in the very middle of the data; 50 percent of the data points are below it, and the other 50 percent above. The quartile Q1 divides the data between the minimum and median in half, and the quartile Q3 does the same thing for the median and maximum.


So how should you interpret this? Well, if you look in the Combined column, you see that the median is 13. So 50 percent of all 59 agents who responded to my query got back to me within 13 days. In that same column, the third quartile, Q3, is 34. That means three-quarters of the agents who responded to my query, or 75 percent of them, responded within 34 days. That really isn’t that long. And it shows that the outliers really are outliers; only once in a very long while will an agent leave you hanging for 263 days.

I’m sure some of you are interested in how long I waited to hear back on my requested partials and fulls. I didn’t create a graph because the data points are much more spread out--hearing back on requested material follows a much less predictable pattern, apparently. Also, I didn’t break it down into partial requests and full requests because that would really involve very few data points. So here’s this chart.

I bet that’s not as bad as you thought it would be. The longest, absolute longest, turnaround time on requested material was (only) four months. And since Q3 is 62, I heard back on 75 percent of my partials and fulls within 62 days, or just over two months.

A few more notes about this project, since I feel like sharing them. Its title is SEE THE SAMELINGS (have I mentioned that before?), and it’s a young adult urban fantasy. Its main character is Eva George (she’s a cabbie), and it’s set in pretty much the only place a book about a cabbie could be set: the quintessential NYC.

Wow. Writing this post has been really…cathartic. Or maybe just nostalgic. Or maybe a little of both. And hopefully it’s been informative for you.

16 comments:

Stina said...

LOL. I have a MSc, and I love analyzing things, but I've never thought of doing what you did. Probably won't either. Maybe I could do a study on the heart and respiratory responses of agents when they find a query this fall in love. And compare it to the response of the writer who lands that request for more. ;)

Great post!

Charity Bradford said...

Wow, just WOW. Thanks for sharing this information. It is really good to see the information laid out like this. Now when I start querying I will have an idea what to expect.

Kayeleen Hamblin said...

I love that I knew exactly what you were talking about. I was a psych major, so I did a lot of stats. If you have to have math, at least you can apply stats in a meaningful way, right? I miss being able to use SPSS. The graphs, etc. were awesome.

Thanks for the breakdown. That's actually really comforting!

Myrna Foster said...

Thanks for doing this post. Yesterday, I was going to write a nostalgic post that featured my favorite writing style book and an old linguistics professor, but I squelched it when I came across that contest. Maybe I'll write it after all. I enjoyed yours.

Unknown said...

As a fellow statistics nerd, I LOVE this post! Also, you wouldn't have scared me away with a regression analysis. Now I want to see the data points for your stress levels throughout those 90 days! Thanks for sharing this with us. :)

Anonymous said...

WOWSAH...my head is spinning. I'm not very mathy, so I'm stunned and awed at all the shiny shiny numbers...LOL. Good work!

Shannon said...

Holy Stats flashback!

I shuddered a bit when I looked at the graphs (thoughts of doing ANOVAS by hand make me quiver), but once the room stopped spinning and I was able to look at your post, it's actually pretty cool.

Hrm...another distraction to keep me from working on revisions....

Kathi Oram Peterson said...

Impressive! I don't know what else to say, which is odd for a writer. I'm still pondering the love of statistics. ;)

Kelly Bryson said...

This was really encouraging. Thanks for sharing, Krista! I never would have thought to use statistics;)

Krista Van Dolzer said...

Thanks, everyone. I'm glad you enjoyed this.

(Sorry for not responding personally. My internet's been out since about noon yesterday, so I'm a little behind with everything...)

A.L. Sonnichsen said...

Wow, I'm speechless. This is amazing. I admire people who have both sides of their brains functioning at maximum capacity. You, clearly, are one of those kinds of people!

Krista Van Dolzer said...

Only on some days, Amy. Actually, I think being a stay-at-home mom is slowly sapping my intelligence:) Oh, the things we sacrifice for our kids...

Erin Edwards said...

I'm sure I will do *exactly* the same thing when I get around to querying, so being able to compare to your statistic will help, I hope, to keep my expectations realistic. :)

Thanks!

Krista Van Dolzer said...

Erin, you're welcome.

Adam Heine said...

And oh my gosh, if we weren't already friends, this would seal the deal. I am in love with this post.

Krista Van Dolzer said...

Adam, glad I'm not the only statistics nerd hanging around these parts:)