SAFMEDS instructions

Terrific post on Leitner boxes and flash cards at A Progressive Case for Classical Education. I’m inspired.

. . . as I prep lesson plans, I note the vocabulary, procedures (yes, you can have Ss recite the steps for long division, for example), and other facts they absolutely must know. On the first day of the unit, I have them write the item down on an index card and tuck it into the daily recitation section for them to practice. 

You can assign retrieval practice for homework, as a bell-ringer, as a time-filler for students who finish fast–whenever it works for your classroom schedule.

Retrieval for the Busy Teacher

Starting Thursday, I’m going to add SAFMEDS (not Leitner boxes) to my composition class. I have no idea whether Leitner boxes work better or more efficiently than SAFMEDS. They may. But if only because my class meets twice a week, having just 3 decks of flash cards (SAFMEDS), all of which every student does every session, seems more manageable than 5 decks of cards that require a spaced-repetition schedule.

I’m finally beginning to use SAFMEDS myself, by the way. It’s taken a while for me to get to it–it’s taken years–because until this very afternoon, I did not have a set of directions I could actually follow. Writing directions is hard.

This procedure from Rogue ABA is the most usable not to mention most logical procedure I’ve seen, and it’s what I’ll be using.

SAFMEDS = Say All Fast a Minute Every Day, Shuffled

EGGs and NEGGs

A core principle we were taught at Morningside Academy’s Summer School Institute: whenever you introduce a concept, you must always provide not only examples, but nonexamples, as well, especially what Kent Johnson called “close-in nonexamples.”

A close-in nonexample is close but no cigar. 

EGGs & NEGGs: books

I mentioned a while back that precision teaching spends a lot of time “training the inspector”: teaching students how to tell a good performance from a bad one.

Which you do by giving students as many examples and nonexamples as they need to learn the discrimination.

EGGs and NEGGs. Examples and nonexamples.

Thursday I spent a couple of hours in an actual bricks-and-mortar bookstore (bring them back ! ), where I discovered EGGs and NEGGs books for French and Spanish.

Happy day.

Now I need the same for listening and talking.

EGGs and NEGGs: teachers need bad student sentences

A core principle in precision teaching — in any effective teaching — is that instruction must begin by teaching the novice to discriminate good performance from bad.

Correct from incorrect.

This principle is true of conceptual learning as well as procedural: in order to know what a concept is, the student must also know what the concept is not.

That’s where you start.

“Yes/no” knowledge comes first because we can’t know whether we’ve done something well if we don’t know what doing something well looks like. We all have an internal inspector who judges performance, and our internal inspector must be trained.

To learn the difference between correct and incorrect, students must be given examples and nonexamples. EGGs and NEGGs in precision-teaching parlance.

EGGs alone won’t do.

When it comes to teaching writing, the need for NEGGs is a problem because examples of bad student sentences are surprisingly hard to come by. I’ve spent hours scouring the web, looking for the genuine article. There’s not much out there.

Sure, your own students write bad sentences, but using students’ own work to illustrate bad sentence writing is rude. At least, it would be rude for me; I can imagine there are instructors out there who could pull it off with humor and esprit de corps.

And even if you do use your own students’ sentences to illustrate what bad sentences need like, you still need an organizing principle.

What different kinds of bad are there?

Bad student sentences in creative writing

After yet another hour this afternoon, I’ve come up with this list of “worst student sentences,” reportedly saved by a professor of creative writing:

The sentences on this list sound like they were written by real students to me. I’m sure they were if only because it’s quite difficult to write a bad student sentence on purpose.

I can certainly write bad sentences of my own. Everyone can.

But I don’t make the same mistakes students do, and I don’t understand their mistakes well enough to be able to imitate them.

What makes a bad sentence bad?

Another problem: this particular list doesn’t really include the type of bad student sentence we instructors see in nonfiction college writing. I’m thinking of sentences that start off fine, but then go off the rails as the word count adds up. During my first semester teaching, I took to calling these constructions train-wreck sentences, a metaphor that had no instructional value whatsoever. Very frustrating.

We need a robust, teachable collection of bad student sentences, and we need a corpus linguist to analyze them for us.

Writing instructors need a taxonomy of student error.

In the wake of today’s Google expedition, I see that there exists a field of written-language study called “error analysis, as well as a body of work on automated scoring of writing. They may have something useful to offer.

But if so, it’s going to take many more hours to ferret out.

ACT results: 82nd percentile to 96th

	4/29/2017: Timed practice test	6/9/2017 ACT	9/8/2017 ACT
English	27 – 85th %	34 – 98th %	35 – 99th %
Reading	23 – 66th %	30 – 88th %	31 – 91st %
Comp. 1	26 (est.) – 82nd %	29 – 92nd %	31 – 96th %

National Distributions of Cumulative Percents for ACT Test Scores ACT-Tested High School Graduates from 2015, 2016 and 2017

^{1. The composite score includes all four sections of the ACT: English, reading, math, science. ↩}

20 hours to fluency

Until I re-read our rationale for the writing curriculum, I had completely forgotten that fluency in sentence composition can be reached in as little as twenty hours.

I’m pretty sure I got that figure from Kent Johnson’s Response to Intervention and Precision Teaching: Creating Synergy in the Classroom, but since we’ve moved and all of my books are still packed, I’m not going to be fact-checking myself any time soon.

I need my books!

Precision teaching, fluency, and training the “inspector”

A section of the rationale Katharine & I wrote for Oxford explaining what we wanted to do:

Fluent performance means we can perform a skill quickly, accurately, smoothly — and automatically, with a minimum of conscious effort.

Fluency is the hallmark of expertise in any realm, physical or cognitive. In an academic discipline or profession, fluency requires years to develop. But fluency in the more basic skills that underlie complex tasks — composing sentences, in the case of writing — can be acquired much more quickly. Fluency in sentence composition, for instance, can be reached in as little as twenty hours of practice. #faf5f1

Good writers are fluent in at least three essential skills:

• Instantly identifying (or “discriminating,” as learning theorists would say) grammatical errors in their own or others’ writing, often “by ear”
• Instantly identifying (or discriminating) many stylistic flaws in their own or others’ writing (dangling modifiers, long chains of prepositional phrases, unclear pronoun antecedents, etc.), often by ear
• Instantly writing grammatically correct sentences. Copy editing and revision can consume many hours, but the process of simply getting a single grammatically correct sentence down on paper is automatic. Fluent writers never have to consciously ask themselves, for example, “Where does my dependent clause go in relation to the independent clause?”

It is often thought that students can learn to write simply by writing a lot and/or by reading a lot. In fact, this approach rarely works. The reason most students do not learn to write by writing is that they cannot ‘hear’ what they write (or what they read). Their ability to discriminate a good prose sentence from a bad one has not been trained.

Fluent discrimination is important because all performance depends upon our internal “inspection” of results. When we read words out loud, for instance, we are actually doing two things: reading out loud and listening to ourselves read out loud, inspecting our performance for error.

Of course, if we are uncertain what the words on the page sound like, we can’t function as effective monitors of our own performance. By the same token, students who have difficulty distinguishing a fragment from a complete sentence cannot inspect their writing for mistakes in grammar, style, or cohesion – at least not efficiently.

In particular, students who have had little exposure to academic prose (most students, it seems) can’t hear themselves as they write—nor can they tell whether readers will understand what they’ve said.

“Precision teaching” is a method that develops fluency and trains the inspector.

The Supplement gives students a sequence of exercises that develops the inspector and trains them to write sentences and paragraphs fluently. Once students reach fluency in sentence and paragraph composition, they will be prepared to move on to the next step: marshaling evidence and pursuing an argument throughout an entire college essay. #faf5f1

Precision teaching and writing

A few summers back, I attended Morningside Academy’s Summer School Institute, which pretty much changed my life. A slight exaggeration, but still.

Morningside is the only school I’m aware of that guarantees its work:

The school promises that your child will make two years’ progress in one year’s time, in his or her most challenging subject, or tuition is refunded.

Their guarantee covers reading (and writing) as well as math, which brings up one of life’s mysteries: why aren’t charter schools, which tend to produce better results in math than reading,¹ beating a path to Morningside’s door?

Why isn’t anyone beating a path to Morningside’s door, for that matter?

If you were ever inclined to think that success breeds imitation in the education sector, Morningside is proof that it doesn’t.

It’s worse than that, actually: successful techniques like sentence combining and sentence-level rhetorics have gone missing.

Our writing curriculum was inspired, in part, by the precision teaching I watched and practiced during my two weeks at Morningside. I want to stress the word “inspired“; the lessons aren’t a proper precision teaching curriculum complete with slices, fluency aims, and celeration charts. But the philosophy of teaching a component skill — sentence writing — to fluency is key.

Here’s how we explained our project to Oxford.

^{1. [I don’t know whether this situation has changed with the publication of Doug Lemov’s Reading Reconsidered and the headway Core Knowledge has made in convincing at least some charters that knowledge is key to reading comprehension.]↩}

How can you tell when a student has mastered a skill?

How can you tell whether someone has truly mastered a skill? What is the measurable indicator that a person really knows how to do something? These questions should be at the heart of every teaching decision . . . and every evaluation we make about the success of an educational program. Yet for many educators, and certainly for most parents, answers to these questions are anything but clear. Most of us have grown up in a “percentage correct world” where 100% correct is the best anyone can do. But is perfect accuracy the definition of mastery? . . . In fact, we see many children and adults who can perform skills and demonstrate knowledge accurately enough – given unlimited time to do so. But the real difference that we see in expert performers is that they behave fluently – both accurately and quickly, without hesitation.