Light and Momentary

Fifteen Years On

Fifteen years ago exactly, I was in the last semester of my MA program and writhing with anxiety about what came next. I had felt called into this field. After discerning the best I could I packed up my belongings and moved away from the big city to get a master's degree.

Next

I've started telling people that I'm going to graduate in the spring. If some kind of disaster strikes, like if I'm diagnosed with cancer next week or my husband runs off to live in Ixtapa with one of the interns, then I may curse my optimism. But I'm feeling confident enough that when people ask me how school is going, I tell them I should be done soon.

What they are supposed to say is, "Hey, that's great! What an accomplishment, finishing a PhD with five kids." Then I'm supposed to smile a satisfied but still modest smile and scuff my toe in the dirt.

What they could also consider saying is, "How's your dissertation coming together? I've always heard that last stage is tough." Then I could allow as to how it is pretty tough, between my recalcitrant stats program and the baby whose mission in life is to stuff into her mouth every choking hazard she can find.

What they actually say is, "What will you do next?" And I don't know yet.

I've been pretty calm about the whole thing, because I believe that God has a plan for me. I felt called into this program and I don't need to know the next step just yet. That, however, is a conversation I'm more likely to have with a good friend than with some random parent at soccer practice.

I have been hoping that something would open up at the university in my neighborhood, but I just didn't know how likely that was. I've been telling people that I would apply for a half-time post-doc that's based an hour away. I figured I could spend one long day on campus and work from home the rest of the time. I could also return to clinical work, even though that's not my favorite idea.

Today in between subsetting my data and making more pretty tables for chapter 3 of my dissertation, I looked at the faculty openings for the university in my neighborhood. There's nothing in the department that trains SLPs, but there are three tenure-track openings in a closely related department. On paper, at least, I have all the qualifications they're looking for.

No heavenly lights beamed down to spell out, "This is it, Jamie -- it's what you've been preparing for." I'm not sure if I should pursue a full-time appointment when baby Stella will still just be toddler Stella. (The department is a five-minute walk from my house, though, and I'm pretty sure a full-time load would be two courses per semester.)  

I'm going to apply, I think, and we'll see what happens next.

Garbage Out

The thing I love about computers is that they do exactly what you tell them to do.

The thing I hate about computers is that they do exactly what you tell them to do.

My dissertation dataset is much the largest set of numbers I've ever tried to wring any kind of meaning from, and the most basic number-crunching is proving to be a larger task than I'd anticipated. I came back from our vacation feeling relieved, because I'd been able to learn enough about multilevel modeling to take a stab at building models with my data. (It is revealing that one of the highlights of our vacation for me was the afternoon spent alone with a stats book and a sleeping baby, in which a couple of light bulbs finally flickered on. Thanks to my husband, who made it possible by wrangling four boys through a water park that sounds like his personal version of purgatory.)

I'm not going to be building models for a while. First I have to get my numbers to line up and stop giving each other noogies.

Last week I spent an embarrassing amount of time just getting some descriptive statistics into a usable format. Today I discovered that the format may be usable but the statistics are garbage.

Somehow when I said to R "take the year 2 tallies for low-frequency vocabulary and divide them by the year 2 utterance counts" -- somehow I actually said, "Please serve me up a heaping bowl of number salad." And R said, "Happy to oblige, and how about some exotic phthiraptera eggs on top?"

I cannot figure out what went awry. The numbers all look plausible but they are all wrong. And because computers do exactly what you tell them to do, it must be my own fault somehow. Argh and blast and FIE.

I Hate Personal Growth

Two years ago I was taking a class in regression analysis. During the week that my second son turned eight and I needed to produce Halloween costumes for four children, I also had a big stats assignment due. I was doing this assignment with a freebie student version of SAS when the computer on which it lived died a sudden and yet harrowing death.

Guess what I did today!

I finished coding transcripts for my dissertation. I'd been working on it 5 or 6 days a week since January. I started when the baby was just two weeks old, knowing that it would be harder when she was bigger than when she was tiny. It was a bigger undertaking than I thought it would be. More than once I kicked myself for being so ambitious in my prospectus.

But I finished.

Wow

Remember how I had resolved to slow down? Forty transcripts each week, no guilt no pressure? I did that for a bit and it was GREAT. Then the nudges started: finals are coming. You should speed up so you don't get behind while you're grading. If you just did one more transcript each day, you could get 46 done each week. That's more than 40, you know. More more more.

I am never sure how much detail to include about my dissertation, because I don't want to cause any glazed eyes or nodding-off-bashing-head injuries. So I'll try to be succinct. I'd been coding elaborated noun phrases. Kids start off saying "dog," and progress to "big dog," and eventually get to "the gigantic smelly dog that dug up my grandma's prize-winning tomato plants." A dearth of those more complex phrases is one of the things that distinguishes kids with weak language skills, especially when they're telling stories. I had also planned to do a syntax analysis on about half the transcripts. For about a third of them, one of my advisor's collaborators had already done the analysis, and a few dozen of them couldn't be analyzed because the kids didn't have enough to say. I would do the rest.

Oh, my.

Syntax is complicated. Kids can toss off a five-second utterance that takes me 10 minutes to sort out grammatically. (I hadn't used this approach since 1993, which is part of why it was so slow.) Even though I was picking up some speed as I went along, it was very very slow, painfully slow.

Yesterday I started thinking that maybe I should ask my advisor's former collaborator if he would consider helping me with this last analysis. She and I had discussed this possibility earlier in the process, but it just seemed like a crazy request. Hi! Want to do 50 hours' worth of work for me even though we hardly know each other? That's what it would take me: a bare bald minimum of 50 hours. Not just any 50 hours, either -- 50 hours of solid concentration. I find that I hit a wall with this kind of close analysis, too, so that even if I had no distractions whatsoever in my life, I couldn't possibly do it for more than four hours a day.

It seemed like a bold request, perhaps even a presumptuous request, but I felt this funny prompting to go ahead and make it. I asked for my advisor's input on the email I drafted to him, and I asked some friends to pray for me, and then I sent it off.

He said yes. YES!! He said he'd be GLAD to do it for me.

I am going to finish the year 3 syntax analyses, about another 30 of them, because I want to understand this approach. I still have to finish coding noun phrases for about 130 more transcripts. But! He is going to do the year 2 syntax analyses for me! Wow.

I feel like somebody just handed me a thousand dollars.

Do svidaniya, Vlad

I am done with morpheme analysis for all 301 transcripts. Huzzah! I am also done with some cleaning up that I didn't realize needed to happen with the first batch of transcripts. I am not done with the counting, which could require quite a bit more cleaning up, but I am going to be optimistic. It might just require a ten-minute chat with my advisor and a couple of lines of code. Whatever the rest of the week holds, it doesn't hold any more disambiguating.

Now for some nice refreshing washing of dishes and some even more refreshing grading. :-P

Halfway

Dun-da-da-DAH! Disambiguator Woman here in a pretend snazzy superhero outfit. I have just finished morpheme analysis for the 151st of my 301 transcripts. Now I am swooshing away in my imaginary purple cape to be Dishwasher Woman and Rosary Woman and Toothbrushing Woman and Very Tired Sleeping Woman.

Still Random But Happier

On Monday I started the year 1 transcripts and it became immediately evident that I would not meet my original goal of finishing morpheme analysis by the end of February. It was a relief, actually, to realize that I had to revise my schedule. Yesterday I had a phone conference with my advisor after the faculty meeting in which they review all the doc students' progress over the past year. I passed. She told me, in fact, that when my file came up, a few of the faculty joked that I should give out time management tips. (There are a couple of PhD students with one child, and one woman who has two children, but I am decidedly an oddity with my five. My only time management tips are to give away the TV and teach the kids to clean bathrooms and load the dishwasher.) We also talked for a while about what comes next. I am working on a timeline that she thinks is ambitious, but she also thought it was ambitious for me to try to get to candidacy before the baby arrived. I'd much rather be shooting for a goal that she thinks is optimistic than be told I needed to meet a deadline that seemed unrealistic to me. So that's good.

It's been almost three weeks since any baby pictures were posted on this blog, which is a situation that must be rectified. I call this one "Baby with Brothers and Laundry." (My MIL has been asking me to dress her in pink and take a picture. I never dress her in baby pink, you might remember, because I have always hated baby pink, but I finally obliged.) Here she is (still in pink! two whole pictures of her in pink!) with her oldest brother; here she is with the youngest. Joe and Pete made her a bed of stuffed animals, which she thought was intriguing and they thought was hilarious. Pete loves, LOVES to hold her and to be photographed holding her; she is sometimes uncertain that this is a good idea.

Alex is making sure that she'll know how to text-message before any of the other babies can do it. Pete's eyes are closed in this one but I love it anyway: he has his shirt hiked up too so he can nurse his baby, who shares his sister's name. I am realizing the 9yo is not in a single one of these recent pictures, which is another situation to rectify.

I was up late last night grading (late by postpartum standards, which is to say midnight), and that always jacks up the cranky potential for the next evening. I am going to go make onion soup, and fold laundry, and tidy up the living room. Which means I should end this post pronto. Catch you later.

A Whole Lot of a Good Thing

I have started the first of the four big analyses for my dissertation. In this one I'm looking at kids' use of derivational morphemes. A morpheme is the smallest unit of meaning in a language. In some cases, that's a whole word. In some cases, it's something smaller. When toddlers figure out that they can add -s to a word and it means "more than one," that's a grammatical morpheme. And it's a big step forward linguistically. (It gives me a geeky little rush, actually, to watch that skill unfurling. How do they know, at 18 months, how to do that? Why do some kids struggle so with it?)

We keep adding morphemes throughout childhood. They're a huge part of the vocabulary expansion that happens in the later grade school years. Quick becomes quickly. Attract can morph into unattractive. One of the questions I'm looking at is whether the control kids use more of those affixes than the kids with risk factors.

(Oops, diaper break. It's a good thing I don't count how many times I get interrupted writing a blog post or I might give it up in discouragement.)

For several of my analyses, I'm using a powerful but not very user-friendly program. The manual section that covers the morpheme analysis tool is pretty forbidding. This is for the serious user, it says, who's willing to put in a lot of time getting it right. Not quite "Abandon all hope ye who enter here," but still rather alarming given the learning curve for the other tools in this package.

But I've started, and I haven't thrown up my hands in frustration yet. It will require me to go through every line of my 301 transcripts and check whether the automated parser assigned the right function to each word -- and, in many cases, to parts of words. I have begun seeing morphology tags everywhere I look, involuntarily parsing the sentences in picture books.

The sentence that I'm typing now would look like this.
det|the n|sentence rel|that pro|I~aux|be&1S part|type-PROG adv|now aux|will&COND v|look prep|like pro:dem|this.

I find myself thinking of the parser as a living and not very bright entity. He reads "Bob and Mary" and calls Bob a verb (which, granted, is certainly possible) and Mary the adjective form of the verb "mar." I guess a mary table is one that's had too many glasses of cold water left sitting on it? Anyway, there are many errors to untangle. Many many.

I have moments of thinking,"I was born to do this," because I love grammar with a rather startling passion. In eighth grade I would take home my English book and diagram sentences for fun on the weekends. (Why, no, I wasn't very popular in junior high school -- why do you ask?) I also have moments of thinking, "This is a staggering amount of work." I have done 16 transcripts. Only 285 to go. If I'm scarce, blame it on the morphemes.

The Plan

All right, I have a dissertation to work on here. I was always vague about my early research project because I am the only person ever to look at that particular relationship, but my dissertation is less unique. (Which is good, because I don't have to argue so much about whether it's a reasonable research question.) I am analyzing three sets of language samples, collected a year apart, for two groups of kids -- one with a set of risk factors and one without. I am running nine computerized analyses on a total of about 160 transcripts. Four of the analyses are big and hard and will provide me with opportunities to grow in patience. (You know how I love opportunities to grow in patience.) The remaining five analyses are easier, but three of those require me to use results from one of the complicated measures.

So: last week when Pete went back to preschool, I pulled out the transcripts I'll be analyzing, converted them to the appropriate format, and fixed the errors that popped up in the conversion. Now I am going to knock off the two easy analyses before I roll up my sleeves to tackle one of the hard ones. My goal for the week is to get the two easy ones finished and get the program I need running on this computer by Friday. We'll see how it goes.

...and they said

Pass!

Roads were nasty in the morning but fine by the time I needed to leave. It was much less grueling than I expected. After it was over, my advisor asked me to meet with her in the lab for a few minutes, and when she opened the door there were shouts of "Surprise!" She and my labmates had arranged a little congratulations/welcome baby party. I was floored -- didn't expect that at all.

When I got home we went straight to Mass. Time now to sit down and eat some dinner. Thanks so much for all your prayers and good wishes!

Countdown

Prelim in less than eighteen hours. Precipitation expected, with temps straddling freezing. Here's hoping there's no ice on the roads. Must remember to take a beta blocker along in case my arrhythmia acts up under stress.

It is extremely unlikely that my committee will fail me, but I'm feeling decidedly what-iffy.

O Mary, conceived without sin, pray for us who have recourse to thee.

<- Look

Over there, in the left sidebar. See what I just did?

The people with feed readers are grumbling, "No, your sidebar doesn't show up in the feed reader, silly." So I'll tell you: I just sent off my first two dissertation chapters to my committee. Two weeks until my prelim.

And she lived to tell the tale.

I just spent two nights sleeping alone for the last time in the foreseeable future.

My professional organization just had its annual meeting, in Chicago this year, and I was presenting my early research project. I had been dreading it because people in my discipline have been consistently skeptical about my hypothesis, and I imagined a long line of naysayers. I have the numbers to support my hypothesis and a pair of plausible mechanisms to explain the numbers, but I still wasn't looking forward to it.

It turned out really well. I was supposed to be there for ninety minutes but people kept talking for more than two hours. By the end I was thinking, "If I just lay down on the floor and closed my eyes, would they figure out that it's over?" Several people said, "I saw this in the program book and thought it sounded crazy, but it makes a lot of sense," which was gratifying.

Only one guy was obnoxious. He was asking me for more information about one of the mechanisms I'm proposing and I mentioned that some of the contributing evidence comes from animal studies. He leaned in and said, "Well, you see, animals don't talk." In response I said, "There's compelling evidence from performance in other domains," though what I thought was rather less charitable. When my project director came back she rolled her eyes -- I hadn't connected the dots but this is a guy who hates her work with an unflagging passion and probably stopped by the poster especially because she was the second author. He has a long history of trashing her wherever possible, so I won't get bent out of shape about it.

I stayed in Hyde Park with friends from college and we had a lot of fun together. (I wondered if I might spot Obama but no dice.) The last time I took Amtrak, the train caught fire and I was delayed by five hours; this was a peaceful trip with no firefighters in view.

Tomorrow I am scheduled to send off my dissertation chapters to my committee; last night I dreamed about the final edits. In my dream I was horrified by my verbosity ("This sentence doesn't even make SENSE!") and by the scrawled unfamiliar citations I had to put in the bibliography ("Why am I citing Leakey and Leakey when this dissertation has nothing to do with anthropology?"). Here's hoping the actual undertaking goes more smoothly.

Submitted

I just sent off a draft of chapter one to my advisor. Lots more work to do, lots and lots and lots and lots, but it's a start.

Progress

Goal for the week: met.

Word count: 6586. That includes my bibliography, though. Hold on -- actual word count for chapter one: 5274.

I am going to tidy up my messy desk and go relax by folding a monster pile of laundry.

Ack!

I found out why my page count didn't increase like I thought it would: Google Documents ate part of what I was writing on Thursday!

Thankfully, the revision history feature allows you to go back through and restore missing pieces. Kind of a pain, but at least I didn't have to recreate it.

Back to work with me -- glad tomorrow is Sunday, when I will not even say the word "dissertation."

Ifs

...If I told God that I was content with a little bit of patience, would he stop sending me all these opportunities to become more patient?

...If I schedule my preliminary exam for the middle of my 39th week, what are the odds that I will deliver precipitously in the middle of it? (And would they pass me or make me reschedule?)

...If I were going to knit a soaker or two for this new little one (in my abundant free time, you know), what pattern should I use?

...If I have put in as many hours as I recall putting in today on my dissertation, why doesn't my page count look like it?

My first chapter is coming together. I'm enjoying it. It's a big job, though. My goal is to finish drafting the sections that explain my independent variables by Saturday evening, and then to draft the sections that explain my dependent variables over the following week. I'll still have gaps to fill in, but I'm planning to submit a draft of chapter one to my advisor by the 15th. Then I am planning to take a little nap before I start chapter two.

Can I really pull this off?

To write a dissertation in my field, you come up with an idea that will fill in a previously unmapped corner of the field, or that will add some useful detail to an existing map. You explain why it matters, and what previous cartographers have sketched out (chapter one, introduction). You write about where you'll be surveying and what you're going to measure (chapter two, methods).

Then you meet with a committee that tells you what's missing from your chapters, and what's inadequately supported in your arguments, and what measurement tools you really must employ that you somehow overlooked. When they agree that you can proceed with your study, you can call yourself a doctoral candidate. (Right now I'm just a doctoral student.)

Since I learned I was pregnant, my goal has been to get to candidacy before the baby comes. I just got an email from a developmental pediatrician whom I asked to be on my committee, saying he'd definitely be interested but he'd need to schedule the committee meeting now. He's already booking up in December.

So it's great that he's interested, but I'm wondering -- can I really be ready by the beginning of December? Am I confident enough to say, "Okay, then, let's set a date"? Also, I've been assuming all along that I would be late again, but what if the unexpected happens and baby is early?

I suppose we can set a date for the first week in December, when I won't quite be 38 weeks along. I will just push aside thoughts of the ice storms we've had in early December in recent years. Global warming, right? And you know, I haven't missed a deadline yet in this program, so I suppose it's reasonable to be sanguine about getting two chapters written by mid-November. I hope. Did I just jinx myself?