Reading Assessment Techniques
We
have created this summary of the different types of assessment that
can be used for measuring development in reading skills in the hopes
that teachers will better understand how single skills can be assessed
by multiple measures. This description of the various assessment
techniques may also help teachers to design their own classroom
assessments, and may help teachers to better understand the district
or campus assessments that are already being used with their students.
Each of the elements of the framework is briefly described, and
descriptions of various forms of assessments that could be used
for that framework element are provided.
 Reading
comprehension assessments are the most common type of published
reading test that is available. The most common reading comprehension
assessment involves asking a child to read a passage of text that
is leveled appropriately for the child, and then asking some explicit,
detailed questions about the content of the text (often these are
called IRIs). There are some variations on reading comprehension
assessments, however. For example, instead of explicit questions
about facts directly presented in the text, the child could be asked
to answer inferential questions about information which was implied
by the text, or the child’s comprehension might be tested by
his or her ability to retell the story in the child’s own words
or to summarize the main idea or the moral of the story. Another
common reading comprehension assessment is called a "cloze"
task — words are omitted from the passage, and the child is
asked to fill in the blanks with appropriate words. Also, young
children’s reading comprehension can be assessed by asking
them to read and follow simple instructions, such as, "Stand
up" or, "Go look out the window."
Reading comprehension should not be confused with reading accuracy,
another very common form of reading assessment. In a reading accuracy
assessment, a child is asked to read a passage of text clearly,
without making any mistakes. The mistakes that the child does make
are analyzed to find clues about the child’s decoding strategies
(not comprehension strategies). Very often, an assessment combines
these two different assessments into one assessment — the child
reads a passage out loud while the teacher makes note of errors
the child makes (sometimes called a "running record"),
and then the child is asked some comprehension questions about the
passage. However, it is worth noting that a beginning reader’s
comprehension usually suffers when he or she is asked to read a
passage of text out loud. When children read orally, they usually
concentrate on reading accurately, and do not pay as much attention
to comprehension of the content. Oral reading accuracy does give
insights into decoding skills and strategies, but that is a separate
test. A reading comprehension test is most accurate if the child
is not reading for an audience.
Find published reading
assessments that test Reading Comprehension.
 Because
comprehension is what is being measured, language comprehension
can be assessed in basically the same way reading comprehension
is assessed. With language comprehension assessment, however, the
child should not be expected to read any text. Everything from the
instructions to the comprehension questions should be presented
verbally to the child.
It is also worth noting that a child’s listening comprehension
"level" is usually considerably higher than her reading
comprehension "level." A child that is not able to read
and understand a passage of text usually has no difficulty understanding
the text if somebody else reads it to her. For most young children
learning to read, their ability to read and understand text is limited
by their decoding skills, not by their comprehension skills (That
is not to say that most children have "good" comprehension
skills or that comprehension skills are not a reading teacher’s
concern. The point here is that even when a child’s comprehension
skills are poor, their decoding skills are usually worse.). However,
sometimes teachers find that a child who can not read and understand
a passage of text also does not understand it when the teacher reads
it to the child. It is always worthwhile to compare a child’s
language comprehension with her reading comprehension to be sure
that her ability to understand text is not being limited by her
ability to understand language.
Find published reading
assessments that test Language Comprehension.
 As
mentioned earlier, oral reading accuracy is one form of decoding
assessment, but it is not a very "clean" assessment. Teachers
need to be aware that, in their early attempts to acquire reading
skills, children apply many different strategies, some of which
are hard to detect. Children often attempt to guess words based
on the context or on clues provided by pictures — most of the
time, a child’s guesses are inaccurate, and their difficulties
with decoding are revealed, but sometimes the child guesses correctly,
making the teacher believe that the child accurately decoded the
word. Teachers who use oral reading as a decoding assessment need
to pay careful attention to the child as they read, and need to
be aware that the child may know some words because those words
are in the child’s sight-vocabulary, and the child may know
other words because she is guessing.
Typically, decoding skill is measured through the child’s
ability to read words out of context. Isolated words are presented
to the child one at a time, and the child is asked to say the word
aloud (this is not a vocabulary test, so children should not be
expected to provide meanings for the word). The words selected for
a decoding test should be words that are within the child’s
spoken vocabulary, and should contain a mix of phonetically regular
and phonetically irregular words.
A child can be tested on their accuracy (Is each word pronounced
correctly?), their fluency (How much does the child struggle with
word naming?), or their "level" — Leveled lists of
words are provided by many publishers, and the child can be assessed
as to her ability to decode words that are of varying difficulties.
Sometimes teachers test children’s ability to "recognize"
sight words as a test of decoding skill, but "recognizing"
words is not the same as decoding them. Decoding is a strategy that
readers can use on all words, even words they’ve never seen
before. Sight-word reading has to do with memorizing the "image"
of a word or a specific feature of a word, and with this strategy,
only a select few words are learned. All children go through a stage
as they learn to read where they memorize a few sight words, and
sometimes they are even encouraged by teachers who use Dolch word
lists and frequency indexes to focus the child’s attention
on the most useful sight words. However, memorizing sight words
does not help a child to learn how to decode words, and testing
the child’s knowledge of specific, well-practiced sight words
does not provide a measure of her decoding skill.
Find published reading
assessments that test Decoding.
 There
are many assessments on the market that measure a child’s general
knowledge of facts about the world. Usually some estimation is made
of what one could reasonably expect children in the first grade
to know (e.g. birds build nests in trees, or bicycles have two wheels),
and the child is asked to answer these simple "fact" questions
(similar to what would be found on the old intelligence tests).
However, the assessment we are suggesting is a measure of the child’s
relevant background knowledge, and by "relevant"
we mean "related to the task at hand." For example, if
a child is expected to listen to and understand the story Charlotte’s
Web, the child should have some background knowledge about farm
animals and spiders. Children know a lot of things; children raised
in the city know about public transit, taxis, traffic jams, shopping
malls, and sky scrapers. Children raised in other settings know
about other things. A particular child may not know much about a
particular topic, and it is worthwhile to assess a child’s
relevant background knowledge before expecting a child to be able
to accomplish a task.
Find published reading
assessments that test Relevant Background Knowledge.
 Linguistic
Knowledge is the synthesis of three more basic cognitive elements
-- phonology, semantics, and syntax. Linguistic knowledge is more
than the sum of its parts, but it does not lend itself to explicit
assessment. A child may have a grasp on the more basic cognitive
elements, but still have trouble blending these elements together
into a stable linguistic structure. If a child appears to have a
grasp of the more basic cognitive elements, but is still having
difficulty expressing themselves or understanding others, it is
likely that the child has not yet managed to synthesize those elements.
Find published reading
assessments that test Linguistic Knowledge.
 The
most common assessment for phonology involves discriminating between
two words that sound similar. In this assessment, the child is asked
to listen to the teacher say pairs of words and decide if they are
the same word repeated twice (which sometimes, they should be),
or if they are different words. When pairs of different words are
presented, they should only differ by one phoneme (and they should
be similar phonemes, such as /sh/ and /s/ or /d/ and /g/). Also,
when pairs of different words are presented, the location of the
difference within the words should be varied (Sometimes the difference
should be at the beginning, as in RHYME-LIME sometimes in the middle,
as in MUD-MADE and sometimes at the end of the word, as in RIP-RIB).
Also, attention should be paid to both vowels and consonants.
In a phonology test, the pairs of words do not have to be real
words the child is familiar with. If a teacher wishes to make up
a phonology test, she might find it easier to use made-up words.
There is some merit to this approach because the child’s attention
is focused on the words themselves, and not on the meanings of the
words.
Find published reading
assessments that test Phonology.
 Semantics
is a general term that just refers to "meaning." Vocabulary
specifically refers to the meaning of isolated words, and morphology
specifically refers to the meaning of word parts, but semantics
can generally be applied to the meaning of word parts, whole words,
sentences and discourse.
There are several ways to assess semantics at each of these levels,
but one common thread involves the question of whether the items
on the test are presented in written form. If the child is expected
to read the items, the test becomes more of a decoding test than
a test of semantics.
Although the items should not involve printed text, it is very
common to use pictures in semantics assessments. A child might be
asked to provide a name for pictures as a test of expressive vocabulary,
or to match spoken words with pictures as a test of receptive vocabulary.
A test of semantics at the larger-than-word level may involve asking
a child to arrange a series of pictures to reflect a logical sequence
of events.
Another common assessment involves asking a child to provide a
word that best matches a definition presented (verbally) by the
teacher as a test of expressive vocabulary, or to ask a child to
provide a definition to a word as a test of receptive vocabulary.
Similarly, a test of vocabulary knowledge could require that the
child be familiar with several words in order to answer each item
correctly. For example, the child could be asked to select a word
which does not belong in a group of words (e.g. THREAD, STRING,
ROPE, KNOT). In this sort of assessment, the child must know the
meaning of most if not all of the words in each item in order to
be successful. Similarly, a child might be asked to provide a synonym
or an antonym for words, which is a test of both receptive and expressive
vocabulary. Again, in this case, more than one vocabulary word is
being tested at a time — the child must know the meaning of
the test item, and must know a word which either has the same meaning
or an opposite meaning.
Morphology assessments often involve asking a child to describe
how a word’s meaning changes as parts of the words are changed.
For example, a child could be asked to break compound words into
their component parts and to describe the meaning of those component
parts (e.g. DAY-BREAK, BASE-BALL, HEAD-ACHE). Or, a child could
be asked to describe what happens when affixes are added to words
(as in SKIP versus SKIPPED) and to explain those affixes (What do
UNWRAP and UNTIE have in common?). Similarly, a child’s appreciation
of morphology can be assessed by asking the child to describe how
words with similar parts are related (e.g. EARACHE, EARRING, EARDRUM).
Semantics assessments at the larger-than-word level usually
depend on identifying words or sentences that do not make sense
in the context (e.g. "Billy had a dog. He loved his dog. His
fish was orange. His dog could fetch a ball."), or they depend
on the child identifying logical inconsistencies (e.g. "Billy’s
dog could fetch and he could roll over. He was a good dog, but he
didn’t know any tricks.").
Find published reading
assessments that test Vocabulary.
 It
is difficult to assess a child’s syntactic knowledge without
using printed text, but as was the case with semantics, if the child
must process text to take the test, then the test becomes more of
a decoding skills test than a test of syntax. It is possible to
make some estimations about the child’s productive syntactic
knowledge by listening to the sentences that the child forms when
he or she is talking.
Further, a child can be asked to identify sentences (spoken) that
are syntactically incorrect (e.g. Jane and her dog the hill they
climbed it.), and additionally, the child can be asked to restructure
the sentence correctly.
Another common syntax test involves presenting the child with sentences
which have one word omitted, and asking the child to suggest words
that could fill the blanks. In this case, the meaning the word is
not what is being evaluated, but instead, the child is graded based
upon the syntactic appropriateness of the word. So, for example,
the child may fill in the sentence, "Mary fell off of the ___"
with any of several obviously appropriate words such as "chair"
"house" or "wagon," but credit should be given
for any noun that the child supplies (e.g. "leaf" "brain"
or "mop") because they are all syntactically correct.
Another assessment of syntax that could be used involves the child’s
ability to combine simple sentences into complex sentences, and
to add modifiers appropriately. So for example, "Brownies taste
good" could be combined with, "Mary likes to eat brownies"
to make the complex sentence, "Mary likes to eat brownies because
they taste good." Similarly, children could be asked to appropriately
insert the words, "brown" "big" and "quickly"
into the sentence, "The spider ran up the wall" to make
a more complex sentence such as "The big, brown spider quickly
ran up the wall."
Similarly, a child’s syntax can be assessed through a test
of their ability to change tense and modifiers of sentences. A child
could be asked to restructure, "I went to the store" to
the future tense.
Find published reading
assessments that test Syntax.
 The
best test of the child’s ability to sound out regular words
is to ask them to name isolated (out of context) words that they
are not familiar with. This insures that they are decoding them,
and not just recognizing them or guessing based on contextual cues.
Some tests simply use real, regular words that are so rare that
it is unlikely that the words are familiar to the child (e.g. PUN,
MOCK, LOOT), but some tests use invented or made-up words (called
pseudowords) to insure that the child does not have any prior experience
with the test items (e.g. PARD, ORT, SERT). Some tests attempt to
make the task more authentic by asking the child to read aloud a
list of people’s names (to pretend they are "calling roll").
The names are spelled phonetically and are not difficult to pronounce
(e.g. WANDA BOLTON, WILLIAM BECKER, VICTOR CONRAD).
For young readers who are not quite able to sound out regular words,
it is appropriate to test their basic letter-sound knowledge (which
is a precursor to regular word reading). Common letter-sound knowledge
assessments ask a child to identify a letter that could represent
a speech sound (e.g. Identify the letter that makes the /s/ sound.),
or ask the child to generate a sound or several sounds that could
be represented by a letter (e.g. What sound(s) do(es) the letter
"S" make?). In some assessments of letter-sound knowledge,
the letters are embedded in the context of words, and the child
is asked to identify the first letter (or sometimes the vowel sound)
of a spoken word.
Find published reading
assessments that test Cipher Knowledge.
 First
a child learns to sound out words; then the child learns that when
certain words are sounded out, they do not make sense. As we grow,
and as we are exposed to more and more text, we learn new irregular
words. As a child, you learned words like ONE, SHOE, and PEOPLE.
As you read more, you learned words like CHOIR, COLONEL and ISLAND.
Later still, you learned words like GEYSER, FEIGN, and BUREAU, and
if you don’t already know them, in the future, you may learn
such words as SYNECDOCHE, BAREGE, and CACHET. In short, a person’s
ability to correctly read irregular words is directly related to
their exposure to those words combined with information about the
correct pronunciation (It is not uncommon for people to have read
a word dozens of times, to know the meaning of that word, but to
not know it’s correct pronunciation.).
Whereas a test of regular word reading is strengthened by using
words that the child is not familiar with, any test of irregular
word reading should use words that one could reasonably expect the
child to be familiar with. This can be accomplished using leveled
word lists, or by using words from previous vocabulary lessons,
or from the leveled texts that the children are reading from.
Once the words are chosen for an appropriate test of irregular
word reading, the test can take one of several formats. The child
can be asked to simply read the words aloud, or the child could
be asked to find a word from a set that does not belong with the
others (e.g. LINT, MINT, PINT). Similarly, the test might ask the
student to find a rhyme for each regular word from a set (e.g. find
the word that rhymes with CHOIR — CHAIR, CHORE, WIRE). Also,
a test of irregular word reading could ask children to match pairs
of words that have the same letter sounds (e.g. REIGN, SIGN, MINE).
Find published reading
assessments that test Reading Irregular Words.
 Phonological
awareness is a general term, and phoneme awareness is a specific
term which is covered by the phonological awareness umbrella. As
such, there are many tests that can be described as phonological
awareness tests, but only a few of those tests are specific enough
to also be called phoneme awareness tests.
Specifically, phonological awareness tests are tests which reflect
the child's knowledge that words are made up of sounds (linguists
call this a "metalinguistic" skill), while phoneme awareness tests
are tests which reflect the child's specific knowledge that words
are made up of phonemes.
So, to test phonological awareness, one could ask the child to
rhyme words (expressive) or to pick words that rhyme out of a set
(receptive). The child's ability to rhyme reflects an appreciation
of the sounds within words, and an implicit understanding that words
are made up of sounds.
Similarly, the child's appreciation of alliteration (words that
start with the same sound) can be tested. The child's ability to
produce words that start with the same sound (e.g. what word starts
with the same sound as the word MILK?), or the child's ability to
match words based on alliteration (e.g. which words start with the
same sound - MAN, MORE, FISH) also reflect the child's understanding
that words are made up of sounds.
Children's awareness of the fact that words are made up of sounds
can also be assessed through word length comparisons - a child is
(verbally) presented with two words, and is asked to determine which
word is longer. This assessment is especially effective for young
children if the phonemes of one word are contained within the second
word (e.g. KING and KINGDOM or PIE and SPY - note that PIE and SPY
have the same number of letters, but SPY has more phonemes.).
Another test of phonological awareness involves the child's ability
to break spoken words up into parts - the child would say the word
out loud, but would pause after saying each part. This type of task
is called a "segmentation" task, and it can be used in a variety
of ways. First, a child could be asked to segment compound words
into their parts (as in "BASE (*pause*) BALL"). Similarly, a child
can be asked to segment words into syllables (e.g. "PEN (*pause*)
CIL"). Also, a child can segment the onset of the word (the sounds
before the vowel) and the rest of the word (sometimes called the
"rime" - not to be confused with "rhyme"). In an onset-rime segmentation
task, the words are almost always monosyllabic, and the child would
say each word with a pause after the onset (e.g. "M (*pause*) OON")
The opposite of segmentation is blending, and every test of phonological
awareness that involves segmentation can be reversed and used as
a blending test. In a blending test, the teacher would say each
word with pauses in the appropriate places, and the child would
try to figure out what word the teacher is saying. Blending is usually
much easier for children than segmentation.
To test phoneme awareness, segmentation and blending techniques
can also be used, but in a phoneme awareness task, the pauses would
be inserted after each phoneme (either when the teacher segments
the word or when the student segments the word). So in a phoneme
segmentation task, a pause is inserted after each phoneme (/sat/
ð /s/ /a/ /t/), and in a phoneme blending task, a segmented word
is blended together to make a whole word (/s/ /a/ /t/ ð /sat/).
In addition to phoneme segmentation or blending tasks, there are
several other phoneme awareness tasks can be used to show that the
child is aware of all of the phonemes in spoken words. For example,
a child can be asked to count the number of phonemes in a word (e.g.
how many phonemes are in the word PIN?), or a child may be asked
to delete a phoneme from a word (e.g. What would PIN be if you took
out the /p/ sound?), or add a phoneme (Add an /s/ sound to the beginning
of PIN), or substitute a phoneme (replace the /i/ in PIN with an
/a/ sound). Also, children can be asked to rearrange the sounds
in a word (move the first sound of SIT to the end - Note, children
who have been taught "Pig Latin" are particularly good at this task.).
Finally, children clearly have phoneme awareness if they are able
to identify a phoneme in different words. Children should know that
the words SAT and TOP both contain the /t/ sound, and that GAME
and PLAY both contain the /a/ sound.
It is worth mentioning that some of these phoneme and phonological
awareness tasks are harder than others. Blending is easiest, but
can be made more difficult if the word, when blended together, does
not form a word that the child is familiar with (e.g. SAZ or VIKE).
Segmentation is more difficult than blending, and becomes considerably
more difficult if the word to be segmented contains consonant clusters
(sometimes called digraphs -- e.g. MASK, SPIN or SLIP). Phoneme
addition, deletion and manipulation, the most difficult tasks, are
also made more difficult by creating words the child is unfamiliar
with, and by adding consonant clusters.
Find published reading
assessments that test Phonological/Phoneme Awareness.
 A
child’s understanding of the alphabetic principle can be assessed
very early, even before the child can read or write simple words.
The most direct approach is to ask the child to write words that
you dictate — even if the child can not write the words accurately,
her understanding of the alphabetic principle is revealed by whether
or not she writes one symbol for each sound in the word. Young children
often represent a whole word with a single symbol (Sometimes the
symbol the child chooses is the first letter of the word, so a child
might represent the word DOG with the letter D). This reflects their
view that a word only exists as a representation of an object. Children
who have an understanding of the alphabetic principle, however,
will attempt to encode all of the sounds they hear in the word,
although they may not use the right letters — in fact, they
may not use letters at all. The child who has internalized the alphabetic
principle may write the word BALL with three symbols, and ironically
may represent the word BOX with four symbols (e.g. BOKS).
Similarly, a child’s knowledge of the alphabetic principle
can be tested in other ways. Children can be presented with two
words (written) — one long word and one short word. The teacher
asks the child to pick the word they think she is saying (and she
would say either a very long word or a very short word; e.g. HAT
or HIPPOPOTOMOUS. The words can get closer in length as the child
learns the object of the assessment.). Even if the child can not
read yet, an understanding of the alphabetic principle will allow
her to pick the right word.
Find published reading
assessments that test Knowledge of Alphabetic Principle.
 Typically,
testing a child’s knowledge of the letters of the alphabet
involves presenting the child with a page full of letters, and asking
the child to name them. The page usually contains upper-case letters
and lower-case letters, and a few odd characters like the two versions
of the lower case "a" and the lower-case "g".
This is not, however, the only approach to assessing letter knowledge.
Young children who do not know the letter names yet can be given
a pile of manipulable letters and numbers and symbols and asked
to separate the letters from the numbers and symbols. Similarly,
children can be asked to "tell what they know" about each
letter — they may not know the name of the letter, but they
might know a sound that it represents or a word that starts with
that letter. Children that know all the letter names can be further
tested by their ability to separate the letters into upper- and
lower-case groups, or to separate them into vowels and consonants.
Find published reading
assessments that test Letter Knowledge.
 Again,
a child’s writing is a good way to reveal their understanding
of the mechanics of text. Even children that are not writing well-formed
letters can reveal what they know about print — very young
children who have some experience with text "write" starting
at the top, left corner of the page, writing in parallel, horizontal
lines from left to right, and from the top of the page to the bottom
of the page. The "words" the child forms are separated
by spaces, and may even contain letter-like symbols. Sometimes children
even insert some attempts at punctuation into their creations.
A teacher can also observe how the child handles a book, and can
assess the child’s knowledge about how information is presented
in the book. A teacher can determine the child’s general knowledge
of books (Does the child know where the cover is? Does the child
hold the book right-side-up? Does the child turn the pages appropriately?
Does the child know that the message of the book is contained in
the text?), and the teacher can gather knowledge about more specific
details (Does the child have one-to-one correspondence between printed
words and spoken words? Does the child know what a sentence is and
what punctuation is? Can the child identify capital letters and
lower-case letters?).
Find published reading
assessments that test Concepts About Print.
| 
