My Own N=1 Design

Those people lucky (!) enough to know me are fully aware that when they start making proclamations about how people act or why people do what they do, they’d better have some data backing their arguments up.  I tend to be a bit of a blow hard, calling folks on when they attribute behavior to the rest world which in fact  is based entirely on their own experiences.  (The criticism I have is that each of us is so unique that we cannot possibly expect that our experiences or even those experiences of friends and family combined generalize to many other people in this world.)

So imagine Dave’s surprise when this week I declared myself a N=1 design.  Such a good N=1 design, my colleague and I agreed that she could use me as an example in her research methods class.

First, N=1 means that there is only one person in the research.  (Which Dave immediately pounced on as “bad” since I have drilled that into his head lo these many years.)

Ah, yes, that’s true.  N=1 is generally bad.  But a N=1 design can be good!  So what is a N=1 design?  It’s a situation in which you essentially study one person’s behavior intently over two (or more) levels of intervention.  Usually one level of the intervention is “control” in which nothing happens and one level is “test” in which one administers a treatment of some sort or another.  In both cases, you examine changes in some outcome, quite often behavior.  It’s a good design because you’re examining the same person in two different circumstances (test/control) and trying to determine if changes in behavior can be linked to whether the person is in the “test” or the “control” state.  Good N=1 designs will even move back and forth between test and control states to see if behavior changes accordingly. 

Still with me thus far???  Not bored out of your mind?  It’s ok.  I’m used to those glazed looks when I start talking about research methods.

So, here’s the scoop.

I believe that my basal body temperature (BBT) charting is a good example of a N=1 design.  I have many “control” months in which I’m not pregnant and I’m not even trying to conceive. Then we have the "potential test” months in which we, ahem, engage in an intervention.  (You can clearly see why Dave thinks I’m one red hot mama).  Let’s go engage in an intervention, big boy.  Intervene me!  I say potential test month because it’s only a real “test” month if I’m pregnant, not just TTC. 

There is a problem with this design.  I have tons of control months.  So I have a good idea of what my temperature does when I’m not pregnant.  But I only have one test month.   As a researcher, I’d really rather see several charted pregnancies for the same person (me) to get an idea of 1) if the pattern of temperature changes during pregnancy and 2) if so, how.  However, I’m not going to commit to 6 kids just to prove that my N=1 analogy of charting works. 

The problem is that being a researcher or simply being a wee bit compulsive I’m data obsessed and thus BBT chart obsessed.  It’s the only objective data I have in trying to figure out whether I’m pregnant or not.  For example, with Conor’s chart, I was triphasic.  If my waking temperature currently rises, let’s say to some odd temperature way out of the ordinary for me, 99.2, I’m totally scanning back to the one past positive chart and all the control charts to see if this has any precedence.  If my temperature drops back down to normal 98.6, I then scan back over Conor’s chart and all my other charts to see if I’ve had a drop like that before. 

And then with all the possible subjective symptoms, it’s making me nutty.  -er.  Nuttier.  And it’s driving my husband crazy.  I had really bad cramps when I got pg with Conor.  I’m constantly thinking about my belly now.  Do I feel anything?  No.  Booo.  Do I feel something?  Yes!  A cramp!  No, wait a minute.  That was just gas.  There’s an ache.  I definitely feel a mild, dull ache.  Is that a cramp or do I so want to feel a cramp that I am making my body perceive there’s a cramp when there’s nothing going on but me thinking about my stomach too much.  AHHHHHHH!  

So temperature is objective.  The numbers are objective anyway.  The patterns, especially trying to generalize off one real test month? Well, they are way too subjective.  And it’s not good for my sanity either.

But stop?  You have GOT to be kidding me.

8:28:38 PM