Testing Listening Comprehension: The Listening Comprehension Construct

Testing Listening Comprehension (1 of 4 in a series)

The Listening Comprehension Construct

Testing listening comprehension is a complex undertaking. One good source of information is my book, Assessing Listening, by Gary Buck, published by Cambridge University Press. I have compiled some basic information on Testing Listening Comprehension into a series:

  1. The Listening Comprehension Construct
  2. Texts for Testing Listening Comprehension
  3. Writing Listening Comprehension Items
  4. Playing the Recording Once or Twice

The Nature of Spoken Language

When making language tests, the first place to start is the language. Processing spoken texts requires skills not required when processing written texts:

  • The information is conveyed through sound, which requires a knowledge of the phonetics of the language
  • There is considerable phonological modification (assimilation, elision, vowel reduction, etc.), which requires a knowledge of the phonology of the language
  • It must be processed in real time, not controlled by the listener, which means processing must be fast and efficient

Linguistically, spoken language differs considerably from written language:

  • Speakers use idea units, not sentences.
  • Idea units are shorter than sentences.
  • Spoken idea units are connected more by conjunction than embedding, and connections are often not explicit.
  • There are many hesitations: pauses, fillers, repairs, and repetitions.
  • The language is far more colloquial, less conservative, and it uses more slang.
  • Speakers use more personal involvement.
  • There is far less concern for truth and more poetic license.

Processing Spoken Language

Meaning is not found in the text but is constructed by the listener, using:

  • linguistic knowledge,
  • the co-text,
  • the context of situation,
  • general background or topic-specific knowledge.

There is no fixed bottom-up order of processing; expectations and inferences lead to top-down interpretations:

  • Speech is generally very fast, and hence processing must be automatic.
  • Faster speech rates are more difficult, and all listeners have a cut-off point beyond which processing breaks down completely.
  • Listeners don’t get all the meanings in a text, just a reasonable interpretation.
  • Listeners extract the meaning, or gist, which they store in memory, and then forget the form.

Remember, listening comprehension is not understanding what words mean but what speakers mean.

The Functions of Speech

There are two main functions of spoken language:

  • interactional—to maintain social relationships: the personal contact is the motivation for the message, for example, phatic, small talk
  • transactional—to transmit propositional information: the content is the motivation for the message

People virtually never intentionally listen to interactional language: to do so is boring and pointless.

Listening Situations

There are two main types of listening:

  • collaborative listening—the listener has the responsibility to cooperate to construct the discourse, provide back channeling, monitor their comprehension, and take turns where appropriate. Note: The listener can ask for repetition but always gets a re-statement.
  • non-collaborative listening—the listener has no chance to influence the discourse or seek clarification. Note: There is no chance to seek repetition.

However, technology has now made possible another type of listening:

  • non-collaborative listening to recordings—no chance to influence discourse, but can get a (perpetual) repetition of exactly the same words.

A listener can be

  • the addressee;
  • the participant, but non-addressee; or
  • the overhearer.

Listening is usually much easier for the addressee, who shares knowledge with the speaker, which is taken into account, whereas the overhearer often lacks the shared knowledge that makes the discourse comprehensible.

The speaker can be in three different situations:

  • monologic: listening to a monolog
  • dialogic: taking part in a dialog
  • group: taking part in a group interaction

Context and Cognitive Environment

Listeners use the context to help them interpret meanings. Traditionally, linguistics has recognized two main types of context:

  • context of situation: where the listening takes place
  • co-text: what has already been said

But the most important context is what is in the mind of the listener. This consists of

  • their purpose for listening;
  • their understanding up to this point—accurate or not;
  • what they are thinking about—which can differ considerably from one listener to another; and
  • the mental images they create—often very idiosyncratic, and often what is actually stored in memory.

It is the cognitive environment that provides the real context for interpretation: i.e. what’s in the listener’s mind.

Causes of Comprehension Breakdown

  • lack of knowledge of the linguistic code
  • mislexicalizations, i.e., slips of the ear
  • lack of cohesion, often due to too many “gaps” in comprehension so far
  • processing breakdowns; listeners have “knowledge” of the language but fail to apply it efficiently due to
    • lack of processing speed,
    • lack of sufficient effort or attention, or
    • inability to engage in other necessary cognitive activities while listening.

Note: Almost all processing breakdowns occur because processing is not sufficiently automatic.

Monitoring of Comprehension

  • If listeners aim for a reasonable interpretation, then they must have some mechanism to check, or monitor, whether their interpretation is reasonable.
  • Listeners often have a reasonably accurate idea of how well they have understood, but not always
  • Monitoring seems to work on the principle that the interpretation is accepted as alright as long as there are no obvious contradictions or discrepancies
  • Alertness and mindfulness of possible discrepancies can vary due to interest, motivation and, no doubt, “general intelligence”.

Oral-literate Continuum

Differences between listening and reading are really a question of degree. So texts can be ranged along an oral-literate continuum:

  • oral texts: more characteristics of spoken language, associated with casual conversation
  • literate texts: more characteristics of written language, especially expository prose

Features that determine the position on the continuum include:

  • linguistic features (the sound system, phonology, stress, intonation)
  • paralinguistic features (tone of voice, gestures, and facial expressions)
  • vocabulary choice
  • grammar
  • degree of planning
  • the type and degree of shared context between the participants
  • the function of the text

Spoken texts differ in their degree of orality depending on the listening situation:

  • Informal situations tend to have more oral features.
  • Formal situations tend to have more literate features.

Planned vs. Unplanned Discourse

One variable that relates to the degree of orality is whether the discourse is planned or not.

Unplanned Discourse

  • initial ideas and first reactions
  • loosely or poorly organized
  • fragments of language, with hesitations, false starts, restatements, vocabulary repair and grammatically “incorrect sentences”
  • context is used to connect propositions
  • referents are often missing

Planned Discourse

  • polished worked text
  • carefully structured
  • appropriate vocabulary and grammar
  • coherent and cohesive

Most informal texts are unplanned; formal texts tend to be more planned. In later blogs in this series, we will talk about the need to plan, unplanned discourse, and the dilemma this creates for test development.

Different Construct Definitions

If we examine listening tests, we can see a number of different listening constructs, from the most narrowly defined, usually in older tests, up to the broadest, usually in more modern tests:

Construct 1: Competence in knowing and using the sound system.

Construct 2: Competence in understanding short utterances on a literal linguistic level.

Construct 3: Competence in using the grammatical system as well as discourse competence.

Construct 4: Competence with all the linguistic, pragmatic or sociolinguistic aspects of language.

It is probably fair to say that these constructs derive from the work on communicative competence, or at least they can be understood within that framework. But in many cases, even the most “communicative” of these will often lack the oral characteristics of realistic spoken language and often will not require automatic processing.

I prefer a different approach to construct definition. Listening Comprehension is a very complex process. But normally, three things need to be stressed for construct validity:

The need for realistic texts: you have to test whether they can understand spoken language.

The need for realistic tasks: you have to test whether they can do what listeners normally do.

The need for language processing to be completely automatic: you have to test whether they can do this automatically, while paying no attention to their listening processes.

A good default listening construct might be:

The ability to process extended samples of realistic spoken language, automatically and in real time, to understand the linguistic information that is unequivocally included in the text and to make whatever inferences are unambiguously implicated by the content of the passage.

In my next blog in the Testing Listening Comprehension series, we’ll look at Texts for Testing Listening Comprehension (2 of 4 in the series).

 

Clip to Evernote

Comments are closed.

Corporate HQ

Lidget Green
700 Briggs Ave. #92
Pacific Grove CA 93950
Phone: (831) 222-3677
Fax: (831) 222-3672

Contact Us

Contact us for more information about our products and services, or to share comments.

Contact Us

Join Our Team

Lidget Green is always looking for great talent. Check to see if we have positions available.

Learn More

Viewpoints

Insights about education, test development, English language acquisition, and more!

Visit Viewpoints