Wednesday, February 17, 2010

Okay, Let's Talk Numbers A Bit

I've had some time to take a longer look at the Denver Post's List of Important Plays and there's some interesting stuff to look at here, particularly when you look at the cross-tabs. I'm really thankful to the fine folks at the Denver Post because they've done something that highlights part of the whole continuing discussion of statistics, bias and diversity that rarely gets highlighted. Think about this: what do these plays have in common?

American Buffalo
Arsenic and Old Lace
The Diary of Anne Frank
Fiddler on the Roof
God's Country
I Never Promised You A Rose Garden
The Kentucky Cycle
The Time of Your Life

Not a whole hell of a lot, at first blush, really. But these are the plays that scored as significantly more important with the male participants in the Denver Post's survey than with the women, some of them, like American Buffalo, having a gap of more than 100 places. (American Buffalo is 15th on the male lists, 148th on the female.) All have a gap of at least 50 places.

Of these, two are all male casts, one is set in a male-dominated subculture (God's Country is about white supremacists, not known real well for their feminism), seven are set in the past, when patriarchy was much the norm. None revolve around women characters, though Arsenic and Old Lace have a pair of great parts. One, Bent, has real historical significance, as one of the first depictions of the treatment of homosexuals under the Nazis. (That one, incidentally, was 67th on the male lists, 148th on the female.)

And the reverse? The plays that scored high on the womens' list and low on the men's?

The American Dream (Albee)
Awake and Sing
The Children's Hour
Fool for Love
The Heidi Chronicles
I Am My Own Wife
Polaroid Stories
South Pacific
Sweeney Todd

Now, the caveats. A) The women in the survey ranked fewer plays, almost fifty. B) This group was largely pulled from theatre professionals. And most importantly, C) the question at hand wasn't "What plays do you like," but "What plays are important." Oh, and D) the men outnumbered the women (68% of the respondents were men.)

So what do we see here? The women tended to rank plays that featured, well, women, in central roles. A couple of wild cards thrown in there, like Oleanna, make it interesting. But it does seem like, largely, the men have as hard a time connecting to plays about women as the women do, connecting to plays about men. Of course, though, when you look at the top ten:

1. Death of a Salesman
2. Angels in America
3. Streetcar Named Desire
4. Long Day's Journey Into Night
5. Who's Afraid of Virginia Woolf
6. Our Town
7. A Raisin in the Sun
8. The Crucible
9. Glass Menagerie
10. Fences


Art said...

I think your post is intriguing, but I'll admit you lost me a little bit at the end.

While American Buffalo would be "muscular", would Machinal be gentle and probing?

Would Arsenic and Old Lace be muscular and the Dutchman be gentle and probing?

And, just perusing the cross tabs there seem to be many examples that just don't add up.

Let's just take Mamet, for example:

While the "male, muscular" American Buffalo does indeed have a gap between male and female admirers, (as does Speed-the-Plow) how do we explain the almost dead equal admiration from both genders of the equally testosterone-fortified GlenGarry Glen Ross?

Ditto: True West

Not saying that looking at the data in the way you are isn't worth doing, but there is just a lot more work to do on it.

Art said...

Although, I will follow up to say that I understand that the survey asked for Most Important and not favorite or best.

That's the reason I'm interested in what you are bringing forward.

99 said...


The descriptions weren't meant to be tied to any specific plays, just generalized descriptors used for "male" and "female" plays.

There may be outliers such as True West and Glengarry, but that doesn't outweigh the larger trend and the larger indications that plays with central female characters aren't regarded as important by the men.

99 said...

As to the Important vs. "like," I think that's actually even more the point: what is regarded as important. I did acknowledge that this isn't a list of favorite plays and just because a play ranks highly that doesn't mean it's more "liked." More highly regarded is probably a better substitute for "important." But the larger point holds: men are more likely to list plays that are about men as important than they are plays about women. And there is an indication of the opposite being true, but not as greatly.

Art said...

Thanks for clarifying that.

Art said...

Like I said, it is intriguing. But in order for this to make sense to me, I would have to interpret what you are saying as follows:

The top of the list, say, the first 50 entries, consists of everybody, (men and women,) just kind of paying lip service to what they know should be the right answers. However, as we go further down the rankings we start to see what men and women respondents actually like and/or think of as meaningful or good.

Is that about right?

99 said...

I'm definitely NOT saying that. I'm saying that this is a list of what a group of men and women rank as the most important American plays and that we can extrapolate some things from where their rankings overlap and where they diverge. I personally don't believe that there is some standard of important play, so the whole thing is a bit subjective. Is A Raisin in the Sun less important than Death of A Saleman? On what points.

But, since this is, largely, a set of plays that most of us are familiar with, we can use our knowledge of these plays, their subject matter and their place in history (because some of the plays are indeed historic) to draw some conclusions about the way men and women view plays.

At the end of the post, I take that extrapolation and apply it to a real world situation.

That's it. I don't think anyone is paying lip service or holding back or anything. I can't know that. All I know is what's on the list. If the men in the survey rank American Buffalo as the 15th most important American play and the women rank it as 147th, that tells us something about how they each view that play. Then we can infer how they might view a play that's structurally or thematically similar. Or dissimilar.

Thomas Garvey said...

Are you sure this data doesn't show that women are more anti-Semitic than men? I notice men rated "Fiddler on the Roof" and "Diary of Anne Frank" much higher than women did. I find that very suggestive. Don't you agree?

99 said...

Good point, and I did notice that. However, I thought it was counter-acted by Brighton Beach Memoirs, Awake and Sing and The Heidi Chronicles, so it was probably not signal, but noise.

In terms of plays with overt Jewish themes, this list is actually pretty thin, in general.

Anonymous said...

God's Country was about anti-Semitism, too. Why didn't women like it? Shit, those bitches hate Jews.

Anonymous said...

God's Country was about anti-Semitism, too. Why didn't women like it? Shit, those bitches hate Jews.

Thomas Garvey said...

Whew! So it turns out women aren't structurally anti-Semitic. Thank goodness.

Or wait a minute - do women like "Awake and Sing" and "Heidi Chronicles" because they feature dominant women even though they also feature Jews???? Oh. My. GOD!

Surely you can still see the problem with your post - you're pushing your own world-view onto the data, which actually isn't even statistically-valid data anyhow. The populations aren't randomized, aren't equal, and seem to have ranked different numbers of plays! You can draw a general inference from surveys of this type that everybody agreed the top plays were, indeed, pretty good, but you can't work backward to draw inferences about the populations as their opinions diverge. Particularly given that many plays "map" to more than one interest group!

And sure, it's no surprise that people more easily relate to those like them on stage. Is anyone denying that? Is anyone saying "only minorities feel that way"? The point is that artistic quality ameliorates, and then overcomes, that bias. That's my belief, and this survey does seem to indicate it's true. Although the fact that almost all the plays people agree are great are older plays is suggestive of a troubling trend, particularly given that even young people seem to agree that the older plays are the great ones.

99 said...

No one has claimed that this is statistically-valid data or a statistically-valid survey. It's a small sample group, and all of your limitations are indeed correct. Does that mean one can't look at it and say there's a trend here? You're doing the same thing at the end of your comment, looking at the top plays as being older. The younger participants also rated less plays. It's totally valid to look at this and draw conclusions. Isn't it?

99 said...

In fact, looking at the cross-tabs, the argument about older plays runs into all the same problems you point out: the group skews significantly older, so a preference for older plays would surface in just the same way.

If the data is incomplete for one thing, it's incomplete for both.

My feeling is that this is suggestive data and worthy of discussion. You draw your conclusions. I draw mine.

Thomas Garvey said...

Well, unlike you, I'm not going to push this as a "conclusion," but just glancing through the data it's striking that while younger people list far fewer plays (as one might expect, as they weren't around for the original productions of many of these), and the plays they do mention skew toward the more recent, they nevertheless agree that the older plays were better. And it looks like, again just glancing through, the variations in ranking are lower on the age axis than on the gender axis. This would seem to indicate that the "classics," which are revived again and again, so that young people are able to see them, are indeed "classic." But again, because each group rated plenty of plays that no other group rated, this is only suggestive.

99 said...

The survey was not about "best" play or even "play I like." It's a survey of "important" plays, which is a stickier question that probably connects to some of those concepts, but not all. One can think or acknowledge a play is "important" without thinking it's better. I acknowledged that in my original post. Saying that this is a list of "best plays" is a little weasel word-ish.

That said, I definitely acknowledge that any trends drawn from this are merely suggestive. I haven't stated or pushed any definite conclusions. Look at what I wrote:
Looking over this list, there is a trend that certain plays split the genders (this is also true of age, but that's another set of calculations). One common factor I noticed is that plays that have strong connections to men and to the male world have a tendency to rank higher among men than they do among women and plays that have strong connections to women and the female world rank higher among women. However, since there are more men in this survey than women, many of the plays that ranked particularly poorly with women, but not as poorly with men, rose to the top. Which gives a somewhat skewed perception of their overall important to the society at large.

I drew two inferences from this, one I used in the post, one I'm making now: A) that leads the "general" list to be male drama heavy and give the impression that men write "better" plays or more important plays and B) that men are likely subject to a bias that's often attributed only to minorities, wanting to see their life on stage. And given B and the dominance of men in positions of power in industries like theatre and the film/television industry, that bias might be a reason for a lack of female writers and a lack of advocacy for female writers by other women.

What definitive conclusion am I pushing there?

As to your suggestion, I think you have a valid point that the younger respondents might be more familiar with plays that have had multiple revivals. The lack of responses I attribute to a smaller sample, in the same way that the women who responded responded to fewer plays. As with American Buffalo, I wonder if more young respondents would have pushed some of those older plays down and some newer plays up. The make-up of the respondent group is a factor in the final list.

Thomas Garvey said...

Okay, we won't call them "conclusions," we'll call them "insinuations."

One of your insinuations is:

"there is a trend that certain plays split the genders"

But this is really limited to only the handful you've selected. The big outlier is "American Buffalo," on which men and women are really split. Somewhat less polarizing are "Fiddler on the Roof" and "Arsenic and Old Lace." But you know, the odd lack-of-resemblance between these plays should tell you something:

That there's really no "trend."

Why? Well because only 12 people out of 177 actually voted on "American Buffalo." That's only 6% of the voters. Indeed, as you get below the top five, you realize the "rankings" are based on input from fewer than half a dozen people; once you go below 100th place, rankings are being divvied up by single votes.

But where there is a "trend" (albeit a small one), suddenly the genders cohere. In fact, if you take out that pesky "American Buffalo," you notice that the top 25 barely vary at all between the two genders, and things only get a little more choppy for the top 50.

As you go down the list, yes the rankings seem to diverge by gender a bit - although what really seems to happen is that things get random. And why not? Most of the bottom 200 plays are there because someone listed them as their eighth or ninth choice out of 10. Again, that doesn't seem like much of a trend.

And the leap from that to your comments on the television industry seems justified by absolutely nothing.

99 said...

I look at the list and see a consistent variation between the scoring of the men and the women. I chose the widest variations to look at. I drew some conclusions from that and extrapolated to a related circumstance.

You disagree with my conclusions. Fine. It's a big country. There's really nothing more for us to say to each other, is there?

Thanks for stopping by.

99 said...

Actually, I do have one other thing to say: I noted a consistent variation both ways. Several plays that the men rated as important, the women rated as less important. On the balance, more plays that the women rated as important, the men rated as less important. And I looked at both of those things and drew my conclusions.

Thomas Garvey said...

Of course you can draw your own conclusions, but I can also point out that you drew those conclusions based on faulty logic. I mean if you really want to see a "trend" in three guys over there liking "Fiddler on the Roof" and one woman over there liking "Ruined" and another one over THERE liking "Sexual Perversity in Chicago," ok, I guess, you can say whatever you want. But statistically, it's not a "trend," and I don't think it's a "trend" in the common parlance either.