- EphBlog - http://ephblog.com -

A Hypothesis on Yield

An anonymous comment awakened me to the fact that this is the year to test a prediction regarding matriculation at Williams that I have had for a while: that weather during Previews weekend is a significant factor in an admittee’s decision to attend. Bad weather will lower yield on admittees. This year is the first chance I have to test my hypothesis since 2001. Read on for details and musing.


The commenter wrote:

I can’t believe it was snowing a week before you took this video, Diana. Bad luck for Previews, but maybe it will weed out the less hardy.

At last! By my possibly erring tally, every Preview Weekend since 2001 has been uncannily good weather. Sensible though Williams admittees are, I think we’re all aware of the power that little, logically immaterial things have on our decisions: in choosing a job, a mate, and I propose a college as well. My prediction: Williams’ yield on its accepted students — that is, the percentage that choose to attend of the group that is accepted — will drop this year.

Sadly, I will have no way to statistically test if any drop is significant (all my statsd knowledge derives from PSYC 201; suggestons are welcome). Nor, of course, is there any way to discover whether this year’s drop would be due to the bad weather or, say, some effect of the new Anchor Housing. But it’s an interesting question (to me) and testable, if only I (or someone) decide to do the legwork to find and check historical weather records, Previews dates, and yield records.

Someone please let me know when the yield data for this year comes out, so I know whether it’s time to eat crow or not.

Facebooktwitter
Comments Disabled (Open | Close)

Comments Disabled To "A Hypothesis on Yield"

#1 Comment By hwc On April 23, 2007 @ 9:17 pm

The real yield numbers won’t be released until the Common Data Set filing comes out…usually in October.

It was seriously nasty weather. However, a completed Paresky Mall is more impressive than a muddy hole in the middle of campus.

It’s hard for yield to swing much because it is anchored by Early Decision.

#2 Comment By eph ’11 On April 23, 2007 @ 10:07 pm

I was in Williamstown for Previews, and even though I’m sure I’ll go to Williams next year and am therefore somewhat biased, I think the bad weather didn’t change many minds. It’s true that the area’s beauty didn’t show through as much as it might have, but the Admissions people did a great job in organizing up interesting activities and events (including a midnight hike, which was awesome, even in the rain/snow). Heck, even if nasty Previews weather weeds out the superficial and boring kids, that would be great!

#3 Comment By Jonathan Landsman ’05 On April 23, 2007 @ 10:33 pm

I hear you, eph ’11, and I’ve no doubt the weekend was fun. But of course they are every year. My claim is just that that kismet, that Williams magic, will be a little less widely felt in sleet than in spring sunshine, and there will be a marginal effect. Since 2001 there’s been fun in the sun, this year just fun.

#4 Comment By ’10 On April 23, 2007 @ 10:48 pm

I think the presence of Paresky, as opposed to a massive construction hole, will negate whatever marginal effect the weather may have had.

A fun study would be to ask current Williams students if and when they toured the campus, and what the weather was like at that time (and if they don’t know, look it up). Then, if the percentage of sunny days is higher than average (taking seasonality, etc. into account), you could draw a weather-based conclusion.

#5 Comment By MikeyD223 On April 23, 2007 @ 11:13 pm

Totally untestable hypothesis. How would you know what the yield at Williams would have been if the weather was nice instead of nasty?

And “HWC”, of course regular decision yield as opposed to overall yield would be the outcome var if such a study were conducted. Duh!!!

#6 Comment By hwc On April 24, 2007 @ 12:06 am

You also have to back out the waitlist admits. Overall yield is somewhat self centering. If your RD yield falls, you start calling people on the waitlist and asking if they will attend should an offer be made. If they say “yes”, you make the offer, thus the low yield is offset by larger numbers of 100% yield off the waitlist.

Because of these self-centering mechanisms (including 40% of the class being Early Decision, yield at schools like Williams usually doesn’t change more than a point year to year.

#7 Comment By hwc On April 24, 2007 @ 12:09 am

“A fun study would be to ask current Williams students if and when they toured the campus”

Oh, my daughter remembered. She went for a weekend alumni shindig in the 10th grade with her mother in January. There was six feet of snow on the ground and it never got above 0 degrees the whole weekend. It took three years of hard work on my part to get her to even consider Williams and go visit again — in the summer.

#8 Comment By Anonymous On April 24, 2007 @ 12:32 am

Speculating about how to measure yield is completely useless.

How could you possibly disassociate the impact of weather from the impact of bomb-makers or the Hitler posters?

As I said, it’s an untestable hypothesis.

P.S. There was never ever six feet of snow on the ground in Williamstown. Though your usual tendency to exaggerate is noted. Williams infrequently gets 6 feet of snow in an entire season (I think the average is around 5 feet per year). Furthermore, snow compacts and sublimates quickly in generally dry environments like Williamstown, making a snow pack of 6 feet highly unlikely (except at ski resorts with the best snow-making equipment) anywhere on the East Coast.

#9 Comment By frank uible On April 24, 2007 @ 12:40 am

On Monday, April 16, I encountered a previewing acceptee and her mother at Pappa Charlie’s. They were looking forward to skiing at Jiminy Peak during their preview visit.

#10 Comment By hwc On April 24, 2007 @ 2:53 am

Thanks for the lesson on snow, anon.

Having lived for five years in the Berkshires, I can assure you that it is not uncommon to have plowed snow banks as high as your head. The way it works is when they plow the roads and parking lots, they push the snow to one side into piles.

Stay tuned. Next week I’ll cover how to drive a car up RT-2 from the hairpin curve in a blizzard when cars are sliding backwards on the ice. Or down the one lane switchback road from Whitcomb Peak to Hoosac Tunnel — those aformentioned snowbanks are your friend, because the only way to stop the car on the icy grade is to drive into one, get out, push the nose of the car out of the snowbank, go for another 20 feet before putting it back in the snowbank, etc.

So, yeah. When you wanna talk snow in the Berkshires, bring it on.

#11 Comment By MikeyD223 On April 24, 2007 @ 3:12 am

ughhhh… It’s time to get a job, HWC

#12 Comment By MikeyD223 On April 24, 2007 @ 3:18 am

Plowed snow banks are not the same as “6 feet of snow” on the ground.

It’s possible to create a snow bank 6 feet tall with a snow depth of only 1 inch. And who said anything about Whitcomb peak? We’re talking about Williamstown.

Here is a link to seasonal snowfall totals in Williamstown. You’re welcome!

http://oit.williams.edu/weather/100_history.cfm?which_data=snowfall

#13 Comment By Jeff Z. On April 24, 2007 @ 6:35 am

It’s only a shame that, unlike students, alumni can not transfer their affiliation … then HWC could be a proud Swarthmore alum, and he’d be able to erase memories, Eternal Sunshine of the Spotted Mind fashion, of all the evil, drunk, dumb Williams jocks burying him in fifty-foot May snow drifts and subjecting him to forced readings of Adam Smith, and replace them with happier memories of debating philosophy into the wee hours with sober, sports-hating, future-Phd, Marxist social activists from Swat.

#14 Comment By David On April 24, 2007 @ 7:50 am

Hey!

Although good natured joshing is welcome and HWC does have a tendency to exaggerate [Not that there is anything wrong with that! — ed.], can we please maintain a positive, accepting attitude here at EphBlog? All voices are welcome. This is especially true of HWC, who does more than most of us to supply relevant statistics and background information to our discussions of various topics.

#15 Comment By Jeff Z. On April 24, 2007 @ 8:11 am

To use a play from your playbook David, is someone writing as “David” in place of the real David Kane here? I doubt HWC, an adult and a veteran of these sorts of discussion forums, takes any offense at my poking fun at his, in my view, occasional excesses of negativity concerning Williams (and I do of course appreciate his useful information and even agree with his critiques at times) … I guess, David, you’d rather I’d save my negativity for non-humorous stabs at prospectives and current students with no involvement whatsoever in posting on this site, a la certain posters here? I guess there is a d, a, v, i and d required to spell “hypocrisy” after all. To quote Rob Cordry, “come ONNNN.”

#16 Comment By David On April 24, 2007 @ 8:42 am

I challenge you to find anything that I have written in the last year which matches the vitriol you and MikeyD direct toward HWC above. Now, we are all sinners, of course, but would you rather that the quality of discourse on EphBlog go up or down in the future?

#17 Comment By Jeff Z. On April 24, 2007 @ 9:00 am

Are you really, truly, serious? First of all, I think my tone is obviously joking around. Second, you have to consider the target — again, an adult who directly engages in often very harsh critiques of Williams and certain elements of the Williams community (namely jocks), vs. high school students and authors of articles in the Record who have never even participated in this blog. Third, in any event, while I am not going to spend hours going through comments to past threads, I am 99 percent confident in my impression that you have employed far nastier / snarkier tone (and I’ll add, less clearly joking around) on multiple occasions in the past year. And I’m hardly alone in this impression. It’s not I who was called a rather nasty ibanker alum, or whatever it was, on college confidential. Fourth, I said some posters, certainly including, but not limited, to you. For example, Aidan, at whose altar you worship and who is certainly never called out for excessive negativity by you (at least not that I recall, I could be wrong), employs far more vitriol on a regular basis. All of that being said, if HWC feels offended in any way by anything I said, and I seriously doubt it, then I apologize 100 percent and will reiterate that it was meant as a joke.

#18 Comment By Rory On April 24, 2007 @ 10:30 am

wow…and i wasn’t involved in that back and forth at all:)

Anyway, as for the stats. It’s not that it is untestable in its entirety, its that causality is practically impossible to show without using an experimental set up. That said, with some funds and legwork, it’d be easy to set up a regression to predict yield before splitting preview weekends into two for a more rigorous causal study.

I can imagine a bunch of ways to do it. Consider comparing Williams yield to Amherst’s yield which should be very, very similar and consistent (and you can check if that is the case), but the two have different preview weekend dates I’d bet and thus different weather patterns on preview weekends, for one possible approach, but it relies on some assuptions.

That does only look at students who come during previews. It might be informative to get a representative sample of students from a variety of years and ask if they went to preview weekend. Compare across years. The problem there being the assumption that preview weekend’s effect hasn’t changed (shaky grounds in some ways).

Or, combine the two: compare sunny and wintry preview weekends at a nationally representative sample of institutions across the country.

It’d also be interesting to look at the population of students who came to previews and the yield from that population alone across time/schools.

just some preliminary thoughts on how to do it. (Can you tell I just finished a class on methodologies?)

It’d be easy to think up controls for all the concerns raised (weather is one, major construction another, etc.)

I toured the day after a big snowstorm. I remember being very jealous of the hot coffee or chocolate my tour guide had the entire time. BUt she was nice enough to keep us indoors for a long time, unlike the amherst tourguides the day before who I think enjoyed watching us suffer trekking through the weather. ick! :)