How do I "grade labs"? In class.

I had students write lab reports for the first six or so years that I taught.  I worked hard on getting students out of their write-as-much-random-bullcrap-as-they-can-shovel report voice, and into writing clear, concise, and precise prose.  But it took enormous amounts of time and political capital to turn students from writing procedures like this

We came into lab today in order to accomplish the goal of discovering the effects of a pendulum.  First we found our partners, mine was Willie.  He and I cut a sting precisely to multiple lengths, measured precisely.  We tied a lab mass which is called Bob to our precisely cut string and let the mass go from rest at a carefully and precisely measured angle that was not too big for Mr. Jacobs to holler at us.  We recorded all our data carefully in our lab notebook.  I forgot to say we measured the period of the pendulum, too.

into writing this

A pendulum of various lengths l, measured with a meterstick, was released from rest at a small angle from the vertical.  The period T was measured with a stopwatch by timing ten back-and-forth cycles and dividing by ten. 

I mean, it seems so simple... but it's not.  Convincing just one student to adopt the simple but appropriate style in the latter paragraph took multiple re-writes, several writing conferences, and too often several arguments with students, advisors, and parents.  ("What do you mean, I failed, I wrote lots of words about a pendulum, didn't I?  That's not worth even a C?  My English teacher gives me a C for that!  This is entirely unfair.")

Multiply the time and mental effort investment by the 50 students in my classes, and, well... it just wasn't worth it.

So what do I do if not lab reports?

Nowadays, we do most lab analysis in class. The students collect data collectively, but they graph or re-graph that data individually - especially if they are linearizing a graph.  Next, they draw best-fit lines and take slopes using far-separated points on the best-fit lines that aren't data points.  This is all on a worksheet that looks like an AP Physics 1 free response problem - a grid for graphing is provided, and several questions are asked with space included for responses.

Students are asked to describe their procedure in "no more than three sentences", including what they measured and how they measured it.  They're asked to explain the physical meaning of the slope of the best-fit using direct reference to the y = mx + b equation of a line.  And then they are asked to determine that physical meaning, with an uncertainty.

Because we're doing this in class, students come to my desk after each part for me to check.  If it's right, they get a stamp and move on.  If it's wrong, they get immediate feedback and do it again right there - with 15 peers working on the same thing, so that they can ask questions and get help without trouble.

This is so, so much better than having students write lab reports or even answer lab questions at home.  See, they tend to do the work at home for the sake of getting done, right or wrong be danged.  (If you think the grade is what motivates the student, you're sorta wrong.  B or C students do the work as quickly as possible, right or wrong.  A students spend way too much time stressing about perfection, still getting the answers generally wrong, they just get angry and uptight about it.  And D or F students don't do the work at all.)

By working in class, everyone focuses not just on finishing, but on doing it right - in fact, they can't finish unless it's right.  And then if someone is slow, they can finish at home... but they've listened to me give feedback to all of their classmates, and they've had feedback themselves on the first part of the lab sheet, so they're more likely to get it right than if I had just sent everyone home to write a lab report.