Monday, 2 October 2017

The Right(est) Way To Teach Spelling (part 1)

Why words ought to be spelled correctly is an opinion piece best left for another time (or never at all). For now, let's assume that we want children to be able to spell correctly for no other reason than to be able to communicate efficiently in the written form. Yes, language evolves over time, but not usually as the result of the odd child sitting in a primary school who sometimes overcompensates by adding an 'h' into 'went'.

One aim of the teaching and learning of spelling is that children's encoding skills become more fluent, thus allowing them to focus more on composition when they are writing. For the same reason, most teachers care about children having good handwriting (let's not get into that one here, either). In order for children to write interesting, thought-provoking, engaging pieces of writing they need not to be hindered by atroshus poor spelling skills. And if you agree with me up to this point, you'll probably agree that teachers need to do something about the teaching spelling - for most, the ability to spell well does not come naturally.

Most teachers will have some ideas about what good spelling instruction looks like but 'there is limited evidence about what constitutes effective approaches to teaching spelling.' And for a blog post all about what the research shows about teaching spelling, that's highly inconvenient. But we will push on.

The EEF report on improving Literacy in KS1 does have this to say: 'Some approaches do have some evidence to support them, especially when evaluated on the basis of improvements to the spelling of individual words. It is less clear which approaches lead to better spelling in the context of pupils’ composition of full texts.' So, if you were hoping for a silver bullet to get children to spell correctly when they're actually composing pieces of writing, then perhaps you'd better stop reading now! However, surely the best bet in this case is to at least ensure they can spell individual words with a hope that this will eventually feed into their longer written pieces? I think so, especially if there's no obvious other way.

Choosing Developmentally-Appropriate Spellings

In their work 'American Spelling Instruction: What History Tells Us' Schlagal and Trathen (1998) concluded that providing levelled spellings was particularly effective in improving the skills of low and mid-level ability spellers. In an article by Bear and Templeton entitled 'Explorations in developmental spelling: Foundations for learning and teaching phonics, spelling, and vocabulary' they report that one of three important structural practices for spellings is that 'students should be grouped appropriately for spelling and word study'.

With the National Curriculum's prescribed word lists for each year group (I can find absolutely no information about how these spelling lists were generated) it is very tempting for teachers to give children spelling lists based on their age, rather than their current level of spelling achievement. With other research showing that setting children by level of attainment is not good for their development, many teachers are beginning to 'teach to the top', including in spelling, meaning that children are potentially being provided with spelling lists of words that are not suitable for their level.

It is best for teachers to assess children's spelling level in order to inform the kinds of words they are then asked to learn. Bear and Templeton are keen to point out that this is not 'for creating a label' but that it is to serve as a starting point for spelling instruction. Children who are seen to forget how to spell words, even if they spelled them correctly in a test after a week of teaching, are usually ones for whom the spellings have been pitched too high.

Schlagal and Trathen make sense of why children don't learn spellings when they are pitched too high:'some children may have insufficiently developed word knowledge for a given level of words'. If we think of spelling being like a building, children need foundations on which to build and these phases of building, according to the research, cannot be skipped.

Selecting Known Words

Bear and Templeton also propose that 'students should examine known words' and the EEF report on improving Literacy in KS1 suggests that 'the teaching of spelling is likely to work best when the spellings are related to the current content being studied in school and when teachers encourage active use of any new spellings in pupils’ writing.' However, at this point it is worth remembering that words selected from reading books, topics or children's spoken vocabulary should also be developmentally appropriate. Bear and Templeton point out that 'if theme is the sole criterion for selecting words... then students are reduced to learning how to spell one word at a time, with no opportunity to discover or explore the spelling patterns that apply to many words.' (more on spelling patterns later)

Testing As A Memory Aid

So, from this (which is based on a whole tonne of research and has been written about an awful lot) we might surmise that testing is a key part of the learning process when it comes to spelling.
  • young spellers studied high frequency words;
  • students corrected their own spelling (under teacher supervision);
  • teachers used the pretest-teach-test method of delivery and assessment; and
  • spelling was allotted between 60 and 75 minutes of instructional time per week.
So, in between the pre-test and the test, teachers should be teaching children to look for and identify spelling patterns in the selected words - this is called word study. Bear and Templeton outline different stages of spelling and suggest a progression of focuses for word study based on each developmental stage (these can be found on pages 225-226 in the article).
This could be summarised into three main areas around which our explanations, instruction and any learning activities should be based upon:
  • word origin and history (etymology);
  • syllable patterns and units of meaning (morphology); 
  • letter patterns (phonics).
A simple timetable can sum up what we've seen so far from the research on teaching spelling. This structure can be applied to children working at different developmental levels of spelling with different lists of spellings:

How Do We Learn?

Let's look at learning in a broader sense and ask ourselves, how do we learn? Without going into too much detail, one of the most effective techniques is to work on the recall of information from our long term memories. In Clare Sealy's super-helpful blog post 'Memory Not Memories - Teaching For Long Term Memory' she summarises research that shows that 'we can strengthen our ability to recall long-term memories by retrieving them.' This is called 'the retrieval effect' or 'the testing effect' and is where testing becomes a learning tool rather than an assessment tool. Clare goes on to explain that 'the more times we try and retrieve something, the stronger the memory gets. But it is the struggle that is important. If we reteach content instead of getting children to try and retrieve stuff they’ve probably forgotten, the memory does not get strengthened in the same way.' 

In his 2006 literature review 'Characteristics of Effective Spelling Instruction' in Reading Horizons, Randall R. Wallace, based on the work of Fitzsimmons and Loomer (1978), reports that spelling lessons offered in a word list format were effective when teachers followed the following guidelines:
In 'Improving Students’ Learning With Effective Learning Techniques: Promising Directions From Cognitive and Educational Psychology' by Dunlosky et al, it is reported that, based on much research, low stakes practice testing (p26) 'enhances learning and retention': pre-testing is an example of this practice testing.

Focusing in on the middle two points made by Wallace and linking that to what we know about recall of information from the long term memory, it would make sense that during a week children are pre-tested, taught and then tested again. To make this practical, we might say that they are pre-tested on their spellings on, say, a Monday, and that on that day they correct their spelling mistakes whilst self-marking their test. On the Friday of the same week, after being taught the spellings in between, the children are then tested again. If the process is left at this point then children are likely to forget the spellings learned in that particular week. It seems that the key to aid retention of the spellings from any given week would be to keep including past spellings in future tests, particularly those spellings which children find tricky.

Teaching Spelling Explicitly
    So far we've discussed selecting words and using testing at either end of a week. But what happens in between? Do we just send lists home and get children to somehow magically learn them? No. The EEF guidance simply says 'Spelling should be explicitly taught'.

    The third of Bear and Templeton's three important structural practices is that 'students should be guided towards discovering patterns and generalisations among the words they examine'. The EEF in their guidance for improving literacy in KS1 are a little less sure: 'there is some evidence to suggest that teaching word patterns may help with spelling.'

    Examining Spelling Patterns

    In a language notorious for having many exceptions to the rules, what sort of patterns are we looking for?

    In their 2008 article 'How Words Cast Their Spell' Joshi and Moats write 'good spellers develop insights into how words are spelled based on sound/letter correspondences, meaningful parts of words (like the root 'bio' and the suffix 'logy'), and word origins and history. This knowledge, in turn, supports a specialized memory system - memory for letters in words. The technical term for this is “orthographic memory,” and it’s developed in tandem with awareness of a word’s internal structure—its sounds, syllables, meaningful parts, oddities, history, and so forth. Therefore, explicit instruction in language structure, and especially sound structure, is essential to learning to spell.'

    When it comes to using etymology as a strategy for teaching spelling Bear and Templeton suggest that this only comes when children are at an advanced stage of development reading and writing (aged 10 and up). This should not stop etymology being a part of vocabulary instruction earlier on, indeed, it is likely that this practice is a possible stepping stone for using etymology as a spelling strategy.

    Goodwin's meta-analysis (2010) of morphological interventions for spelling shows that this method of teaching spelling is successful. A simple definition of morphology can be found on wikipedia:
    'the study of words, how they are formed, and their relationship to other words in the same language.It analyzes the structure of words and parts of words, such as stems, root words, prefixes, and suffixes.' The National Curriculum appendix gives a good breakdown of the spelling patterns that children should be learning. This seems to be the area that most research agrees on as being the main focus for spelling instruction.

    Looking at letter patterns is helpful to a point but Fitzsimmons and Loomer reported that heavily depending on phonic rules is ineffective and intuitively we know that to be true - many misspellings we come across in children's work are as a result of spelling phonically without applying knowledge of other rules.

    In Summary (So Far)


    In the next blog post we will look at research-based, practical ways of teaching spelling on the days between the tests and we will add to our weekly timetable.

    No comments:

    Post a Comment