Last month, I wrote an article summarizing 13 meta-analyses on writing instruction. I think this was a useful article, because it allowed the reader to quickly see what types of writing teaching methods had been well tested in research and which ones had a large impact on learning outcomes. I included with this article an interactive database that allowed the user to cross compare the individual sample, study size, and study quality when looking at this data. While I personally found this data interesting, it actually left me with more questions about teaching writing than answers and some of my most astute readers actually emailed me the same questions that I had. 

1. Which instructional methods support which writing outcomes?

2. How should writing instruction change based on grade?

3. How should writing instruction change based on the instructional tier?

In this article, I have dived deeper into the writing research and hope to provide some greater clarity on these questions. But before we dive deeper into the experimental research on this topic, I would like to explore some of the proposed theoretical models for writing instruction.One of the problems with looking at a mean effect size from a writing meta-analysis and comparing it with other mean effect sizes, is that writing instruction and assessments cover a diverse set of skills that are actually not well connected, including but not limited too: printing quality, spelling, sentence quality, writing length, and passage quality. Moreover, writing researchers are in particular interested in the intersection of academic outcomes and executive functions like: working memory, planning, self regulation, and cognitive flexibility. That said, for the purposes of this article, I will not be discussing executive functioning, because it is beyond the scope of my current discussion. 

The Five Pillars of Literacy Instruction

Within the context of reading instruction research, the term “the five pillars of literacy” are often used to describe phonemic awareness, phonics, vocabulary, fluency, and comprehension. These instructional methods are often referred to as the pillars for two reasons: 

1. They correlate with board skills that can be somewhat indepedently assessed from other skills. 

2. There are large bodies of scientific research that support the efficacy of instruction specifically targeting these skills. 

Many modern researchers would also suggest that morphology instruction be included as a 6th pillar and I personally agree with this assertion. That said, the concept of the 5 pillars is not without its critics, many of whom worry that oversimplifies literacy instruction. However, as a teacher I think it is a very helpful instructional model. It gives teachers core concepts that they must teach, well supported by scientific research.
 

Theoretical Frameworks for Writing Insturction
Several authors have proposed parallel theoretical models for writing. Well-known examples include The Writing Rope (Joan Sedita), The Simple View of Writing (Berninger & Amtmann, 2003), and The Not-So-Simple View of Writing (often attributed in the literature to subsequent elaborations). Each highlights essential components of writing development:

• The Writing Rope identifies Critical Thinking, Syntax, Text Structure, Writing Craft, and Transcription.
   

• The Simple View of Writing identifies Transcription, self-regulatory executive functions, text generation, and working memory.
   

• The Not-So-Simple View of Writing emphasizes Transcription, self-regulation, working memory, and text generation.
   

Despite differences in emphasis and terminology, these frameworks cluster around four recurring capacities: transcription, sentence/text generation, self-regulation, and working-memory constraints. My goal is not to replace these models but to translate them into classroom-measurable outcomes that teachers can teach to, assess, and monitor.

With deep respect for this prior work, I prefer to anchor the classroom lens in specific instructional methods that are easy to conceptualize, assess, and that have strong empirical support. At the risk of some hubris, the pillars I propose reflect that teacher-facing translation and will guide the practical recommendations that follow.

The Five Pillars of Writing Instruction
 

1. Legibility (Transcription & Speed)

2. Spelling

3. Fluency (Writing Production Fluency)

4. Sentence Quality (Syntax & Clarity)

5. Passage Quality (Structure & Development)

Each pillar includes multiple research-backed instructional approaches, and its outcomes can be assessed independently of the others.The app below provides definitions, high-yield methods, recommended grade emphases, and citations to relevant meta-analyses.