Clashing norms of deference

I posted this on FB. “How to be deferential but not excessively deferential: If you have a scheduled appointment with your professor and you can tell she is talking to someone else, knock or stick your head in so you are sure she knows you are there, then back up apologetically and say “I’ll be happy to wait.” Quietly waiting without letting her know you are there is a problem because she may prefer to get rid of the person in her office and stick to her schedule rather than run late with you, and she should be the one who gets to decide this.”

In my office configuration I cannot see the hall from my desk and I have OFTEN been chatting aimlessly with someone, telling them “I’m expecting a student soon” and then even “I wonder where my 3pm appointment is, did he forget?” while, unbeknownst to me, the student is sitting or standing quietly and patiently outside the door, never announcing their presence. This drives me crazy, as it seems going way overboard in the deference direction when you have an actual scheduled appointment with someone not to announce that you have arrived for it. Thus, when given the opportunity, I instruct students (as above) about how one can simultaneously exhibit politeness and deference while also honoring schedules. However,  former students (who are now professors themselves) confirm that their own sense of deference would lead them NEVER to interrupt a conversation a professor was involved with.

Is there any hope for this culture clash? I obviously need to return to the sign on my door that says “please tell me if you are waiting for me.” But even when I used to have that sign on the door, I’d have students who either would not notice the sign or not think it applied to them.

Coauthoring Norms 1: Assisting and Junior Authoring

My goal is to improve the culture of publication and coauthoring in my department. Although some of our students do great on this, others languish, and many of our students complain that they do not get enough mentoring about publishing. I have identified as one problem that many faculty consider it “exploitative” to involve students in their research if they are not being paid. Another problem is wide variation in opinions about the level of involvement that merits a coauthorship. What I want to do is to develop a set of normative guidelines for apprentice-like experiences that do not involve payment, as well as guidelines for those that do. I am working up a draft of this and would appreciate comments and reports on good and bad experiences and practices in other programs. So here is my draft. Comments, please. NOTE: This is explicitly a model for the hierarchical situation in which a professor takes the lead in defining a project and the graduate student is an apprentice. It is NOT a model for the more egalitarian relations that develop organically. Continue reading “Coauthoring Norms 1: Assisting and Junior Authoring”

Letting Go

The publisher emailed me last week to ask whether it isn’t time to just give up and admit the book is never going to get written. It’s not that I have done no work. I’ve done tons of work multiple times for nearly 10 years. I’ve generated hundreds of pages of text, hundreds of regression analyses, and more than 10,000 graphs. Four years ago I dug down, worked hard, and sent off a partial MS saying “I don’t know if this is a book. If it isn’t, just tell me so and I’ll let it go.” I was encouraged to go on but asked a question that led me to another round of analysis and a whole new set of findings and a totally different understanding of the main story. Two years ago there was another crisis: the publisher said  it has to be finished by September or we’ll pull the plug. I dug down, did another revision of another partial MS and sent it off, again saying, I don’t know if this is a book. Maybe we should just give up. Nearly a year ago we had a “book meeting.” The basic response was: This isn’t a book. There is too much information, we don’t want all those details.  It was pretty frustrating to hear that four years after I’d said that. I said I’d think about whether I could reshape things to pull the main narrative out. There is a narrative in there, but it is  hard to see how it hangs together into a single simple story. And I’ve done this so many times, I just don’t think I can do it again. I’m tired of it.

I think it’s time to pull the plug, to salvage the fragments of what I’ve done and put them out in other ways. But it is hard to do. It is hard to decide that I’ve wasted so much of my time for the past 10 years working on a project that will never see the light of day. I can’t make myself send the email. I have to sit with this for a while.

Lessons learned.

For the publisher: don’t give a contract for a book to someone who has never written a book unless there is already a set of articles to build on or a good draft manuscript and outline.  I’ve got (or used to have) a good track record as an article writer, but I have no track record as a book writer. My one book is a collection of articles and that was delivered five years late. Books and articles are different kinds of products and being good at writing one of them does not make you good at writing the other.

For the writer: don’t accept a contract for a book unless you know you can deliver it. It’s been clear for at least 8 years that I had no clear conception of what “the book” would be, and trying to write a book without a clear conception of the product is a recipe for disaster.

And another lesson. Academic work has to be shared and communicated to be meaningful.


Staff Relations

I’ve gotten involved with a committee on “staff climate” issues at my university. I’d thought this would be mostly about how rude faculty and students are to staff. And this is a real issue. Staff are often treated as part of the furniture (i.e. ignored as people) or as the targets of abuse by faculty or students who want something done and don’t  care about what else the person has to do.

But I’ve also learned that a huge issue is the relation between department staff and central administration staff. Department staff have many different kinds of paperwork to do, do some of them only rarely, and make mistakes. Central administrators are seen as failing to recognize their own inconsistency in how they want things done and as being unhelpful, hostile and even abusive toward department staff. I chatted with a friend about this and learned that he (as a dean) says these issues are endemic, but at his institution they pay central administrative staff well to fix mistakes made by department staff. There is an underlying structural cause of this conflict: faculty and students want decentralized staff who are available to meet our needs in a personalized way, and department staff typically prefer jobs that involve a lot of variety and human interaction.

Another staff issue at our place is frozen [low] salaries and limited career paths.

What about your school? Do you know how the staff feel about their working conditions? And specifically department vs central administration — do you know how that works at your school? Are the relations good or bad? Do you recognize structural/organizational features that help or hurt the situation?

Article Equivalents

I went to a reception yesterday for outstanding women of color at the university. This was a lovely event except that we all had to stand for an hour of awards presentations and keynote. The award winners had all done jaw-dropping amounts of service. The keynoter was a Native American professor whose first career was in journalism. She used the occasion to criticize the academy for failing to give adequate credit for service. She said that diversity is not just a matter of getting darker skins in the place, it is a matter of getting people from different communities who have different priorities. She was arguing that diversifying the institution must include giving greater weight to service in the tenure process, making the “three legs” of the academic stool (teaching, research, service) more evenly balanced. For her and for most women of color, she said, what you do is not just about yourself but about what you contribute to your community. I was reminded of other things I’ve been reading/hearing that confirm the difference between the individualism of White professionals and the family and community focus of other groups. Few communities of color need another article in a peer reviewed journal, she said. Then she said something like: “Each board or committee or community project or group of students mentored is another article or book chapter you don’t have time to write.” There really is a finite amount of time and if you are doing a lot of service you have less time to do research and write. You cannot really diversify the institution unless you change the reward structure to acknowledge the importance of service.

(This in turn reminded me of a brief conversation I had years ago with a couple of very prominent woman sociologists. People had exchanged information on the order of “I’m dealing with children now, you know how that is,” and grunts of acknowledgment. Then one woman said, “I was talking about this to X [prominent male sociologist] and he said that each child he had cost him an article.” Eye rolling, exasperated sighs. One article, right. We wish. “Five or six articles at least,” muttered one woman.)

To clarify: I don’t think institutions can or should reward time spent in child-rearing, although they should accommodate it. But institutions can and probably should better reward time spent in community service. How to do this is a hard issue.

Reading Tables

Reviewing articles makes me realize that people (including people who appear to be otherwise quite sophisticated in their methods) don’t know how to read tables for error and instability.  Obviously, I just found a zinger. Details suppressed in the interest of the integrity of the peer review process. But if the author had really looked carefully at the tables instead of just coming up with stories to explain the coefficients, s/he should have realized something was amiss.

When you are comparing different model specifications on the same data, don’t just look at what is significant, and don’t just look at the variables you are interested in. Pay attention to whether the coefficient on each variable is relatively stable across models or fluctuates with the addition or subtraction of other variables.  The coefficients on the same variables on the same sample normally stay pretty similar as other variables come and go from alternate specifications.  If the coefficients are relatively stable (roughly the same magnitude, roughly the same standard error) in different models, this is good. They may go in and out of statistical significance depending on what else is in the model, but if the effect size stays about the same and the standard error stays about the same, that’s stable, that’s good.

If they are not stable, you need to know why before you mail the article off to the journal. In the worst case, unstable coefficients change between significantly positive and significantly negative, or between close to zero and large in either the positive or negative direction.  But also pay attention if they keep the same sign but get a lot bigger or smaller.

What if coefficients are not stable? If the coefficient of variable X changes when you add other variables, one of three things is true: (1) the other variables correlate with X and overlap or interact with it in explaining the dependent variable, or (2) the sample is different in the two models, or (3) you made a mistake in running the models or copying the tables.

Some correlations or interactions among independent variables are substantively meaningful or otherwise unproblematic. It is normal for the coefficients of each of a set of correlated variables like income and education to be smaller when they are together in a model. Sometimes the whole point of an article is that a coefficient goes to zero or changes from zero to significant when something else is controlled. Similarly, sometimes the point is that some factor is salient only for a subset of the sample.

But before you hang your whole theory or interpretation on a fluctuating coefficient, you want to make sure it isn’t just a mistake. Make sure there are no typos in the code that produced the results. Make sure the table is copied properly. Check the sample sizes to be sure cases were not dropped for some unexpected reason. And especially check for specification error: explicitly test whether coefficients bounce with minor changes in model specification. Very often, you will see that the explanatory power of a model does not change at all when you add more variables, even though the coefficients change. This is a symptom that your sample is too small to make the distinctions you are trying to make. This is especially likely in fields where samples are necessarily relatively small, as is often true in research on organizations or political units or annual time series. Do your variables of interest have strong bivariate effects without controls? If not, exactly which control variables are needed in the model for the variable to have a significant effect? At what point do you stop adding explained variance and just change coefficients? In particular, watch out for pairs of correlated variables like income and education that take opposite signs in models with lots of other independent variables: this is frequently an artifact.

The problem of ignoring coefficient fluctuations is especially likely when the coefficients for “control variables” are suppressed. I have reviewed quite a few articles in which coefficients on control variables fluctuate quite suspiciously with nary a mention from the author, and am never happy when control variable coefficients are omitted entirely. (If they are going to be suppressed in the interest of space and readability from the main table, I still want to see them in an appendix as a reviewer, even if they appendix will end up on a web site instead of in print.)

Also pay attention to the number of cases in each model, to be sure you are not losing cases unexpectedly to missing data or other anomaly. If patterns of missing data are not a problem, the coefficients will stay pretty stable despite sample size fluctuations. But if a coefficient changes markedly when the sample size changes, that’s another sign of trouble.


Originally posted at Scatterplot. I just stumbled across the matter of Jared Diamond’s New Yorker article last year telling the story of vengeance fights in Papua New Guinea based on stories told to him by his driver.  As Diamond told the story, the driver Daniel Wemp and other real people whom he named by name and attributed to a specific tribal group — bragged about murdering and raping people in an ongoing vengeance war. The short version of what appears to be true (the case is still in process) is that Daniel Wemp told Diamond the stories when they were driving around in 1999-2002, but  Diamond did not take notes on them at the time but rather reconstructed them from memory and a follow up interview in 2006, and got the facts all wrong about who did what to whom and when and why and what tribes were involved, as described in meticulous detail byRhonda Roland Shearer at The original article was pulled from electronic archives last year. The driver and purported murderer, Daniel Wemp, and one of his purported victims, Henep Isum (who was not, in fact, paralyzed by Wemp or anyone else), filed suit April 20 in New York for libel and defamation of character, asking for $10 million in damages. Continue reading “facts”