2.4.3: Comprehension

Comprehension of written texts

Finding a suitable text

First you will need to know what the child’s reading age should be and then select an appropriate passage. Assessing the reading age (R) of a passage is extremely difficult, and there have been many attempts at designing a formula for finding the reading age of a text.  

Here’s three examples of formulae for determining the reading age of a text:

1. FORECAST formula

$$ R = 25 – {N \over 10} $$

where N is the number of one-syllable words in a passage of 150 words.


2. FOG index

$$ R = {\frac{2}{5}} {\choose \frac{A}{n} + {\frac{100L}{A}}} $$

A = no. of words in passage
n = no. of sentences
L = no. of words containing 3 or more syllables (excluding the ‘-ing’ and ‘-ed’ endings).


3. FLESCH formula

$$ SCORE = 206.85 – (0.846 x S)-(1.015 x W) $$
S = total number of syllables in 100 words
W = average number of words in a sentence.

This score is not a reading age but can be converted to one by using the table below.

Score Reading Age
score > 70 $$  5 + {\frac{(150-score)}{10}} $$
60 < score < 70 $$ 5 + {\frac{(110-score)}{5}} $$
50 < score < 60 $$ 5 + {\frac{(93-score)}{3.33}} $$ 
score < 50 $$ 5 + {\frac{(140-score)}{6.66}} $$ 


We are basing what follows on the advice of Hornsby, and Leslie and Caldwell, in order to look for typical errors and interpret them. According to Hornsby, the passage must be relatively easy, but not too easy, so that you can base your conclusions on about 25 mistakes.

Evaluation of comprehension

Leslie and Caldwell used four soundings for examining reading comprehension:

  • Throughout school, children will come across stories which rely on sequences of events (narrative texts) or those which require understanding of deeper or more analytical concepts (expository texts). The first may be tested with questions like ‘And then what happened?’, and the second with questions like ‘How is the person in the story feeling now?’ The first are usually easier to recall.
  • Readers with prior knowledge of the text will remember and understand better than those without prior knowledge.
  • Readers with difficulties find it easier to read aloud rather than silently, and these readers will read aloud and silently at the same speed, betraying a lack of automaticity in word recognition. Good readers read faster when they read to themselves silently than when they read aloud.
  • Poor readers use less efficient strategies for finding information than good readers to overcome difficulties in comprehension.

Leslie and Caldwell suggest four steps to examine the comprehension of written texts:

This should be done in two ways:

  • By using a conceptual task: you ask four questions to do with background ideas concerning the text presented. Allot a mark from 0 – 3 for each answer depending on its quality: a good synonym, a relevant example, etc.
  • By using a predictive task: you say to the child ‘since the title of the passage is … and that it speaks about …, …, … and … (name the four concepts covered in the conceptual task), what do you think the text is going to be about? You count the number of verbs or nouns the child gives which are explicitly or implicitly related to the passage.

You can then score the ‘results’ by adding the marks for the conceptual task to the number of correct answers for the predictive task, in order to obtain an estimate of the child’s background knowledge.

In Leslie and Caldwell’s study children who obtained at least 55 % in this test got at least 85 % for the comprehension questions after reading the text. This shows that background knowledge is an important factor for the understanding of a text.

Give the child a text, inviting him or her to read it out loud on his or her own, without your help.

If the level of the passage is appropriate for the age, the pupil should be able to read it even if some mistakes are made. However, if the child is completely blocked, stop immediately and find a less difficult passage.

Note that some children, especially those with dyslexia, have to get used to reading, and have to find the courage to read out aloud; they will sometimes give the impression that the passage is too difficult after attempting only the first few lines. So do everything you can to put them at their ease and, if necessary, ask the child to read some other passage aloud before starting the actual test.

Try not to show, verbally or otherwise, that the child has made a mistake and avoid mouthing the word unless this is absolutely necessary. If the child refuses to read a word, encourage him or her to guess it.

Prepare a second copy of the passage on which to mark the mistakes although you may not be able to mark them all. It is strongly recommended to make a recording so that you can listen again, and notice the behaviour of the child during the test.

Finally code and analyse the mistakes (see below).

This is done in two ways:

  • Through recall:

    Ask the child to recall the passage in his or her own words as if telling the story to someone who has not read it. For this you will have made a list of the key elements of the passage beforehand which you can tick when the child remembers them.

    Once again, it is useful to record the reading so that you can listen to it later.

    This recall of the passage can provide important qualitative information and you can use it to answer the following:

    – in stories, does the recall keep to the basic structure: background-people-issue-events-outcome? Have the most important points been included?

    – in expository passages that go into more depth about the plot and characters involved, does the recall bring back the central idea and the details which support it? Have the most important points been included?

    – is the recall sequential?

    – is the recall precise?

  • Through questions:
    – explicit, for which the answer is clearly in the passage
    – implicit, for which the child must use hints in the text. This will examine the child’s abilities to make inferences.

When it is not possible to know if a mistaken answer results from problems of comprehension or memory, after asking questions and finding the level of understanding, you can present the text again and ask the child to find the missing answers or correct the mistakes. If the child can do this he or she has probably understood the text. If he or she still has difficulties, in order to see if the child can find the clue, you can give the right answer and ask the child to find where in the text this is hinted, or you can give the clue and then see if the child can arrive at the right answer. This procedure provides useful indications, given that children with reading difficulties find it harder than good readers to follow the clues.

Analysis of mistakes

The kinds of mistakes made provide important indicators of the strategies the child is using. Hornsby suggests qualifying the errors as positive, which indicate good strategies, or negative, which indicate inappropriate strategies:

  • Positive errors:

    These include:

    – omission, insertion or change of sequence, and inattention to punctuation provided the meaning is not altered

    – repetition of a word or part of word, which at least signals an attempt to get to the meaning

    – self-correction

Negative errors:

– omission, insertion or change of sequence, and inattention to punctuation by which the meaning is changed

– refusal to read a word, which signals no attempt to decode or contextual retrieval

In general it is also important to listen to the intonation. If it sounds natural it usually means the text is understood. If it is without tonal variation or stilted, then there may be a problem of comprehension.

Using the results of the comprehension test

It is vital to use what you have learnt from this examination to stimulate the positive strategies which are being underused.

These could be, for example, to:

  • – link more to prior knowledge
  • – pay more attention to the punctuation
  • – improve decoding strategies
  • – encourage self-checking and self-correction

However, be very careful to guide the child away from strategies which are being overexploited, typically those relying only on context and neglecting other strategies.

You may want to use one or more passages to test understanding in order to answer precise questions with variables, such as:

  • – comprehension of a scientific versus a narrative passage of the same level
  • – silent reading versus reading out loud
  • – comparison of performance of recall and comprehension of a text read by yourself with a text read by the child