Content-type: text/html ~ Stephen's Web ~ Interactivity: Another Tack On It (4)

Stephen Downes

Knowledge, Learning, Community

Oct 04, 2002

In this series:

- Interactivity and Best Practices in Web Based Training
- Interactivity: Another Tack On It (Part 1)
- Interactivity: Another Tack On It (Part 2)
- Interactivity: Another Tack On It (Part 3)
- Interactivity: Another Tack On It (Part 4)

Posted to the E-Learning Course List October 1, 2002

George Siemens wrote,

Thanks for the links to your articles, Stephen.
I've distilled the points in your articles as follows (let me know if
I'm way off base):

  • Type/sort of interaction (human-human, human-computer,
  • Conditions that influences the types of interaction (Time (real
    or asynch.), number of people, location (proximate, distance)
  • The uses of interaction types (i.e. what is each best suited
  • Degree of interactivity (information exchange between
    participants as one measure)
  • Amount of info conveyed is relative to the needs of the receiver
    (aka learner)
  • "Thus, a measure of the quality of learning materials would
    include, first, the quality of information transmitted, and second, the
    degree of interactivity afforded by means of transmission."
  • "Greater interactivity tends to increase time spent"

This is a good summary. But Siemens continues,

Here is a definition of interaction I'd like to draw to your attention:
"Interactions occur when these objects and events mutually influence one
another. An instructional interaction is an event that takes place
between a learner and the learner's environment. Its purpose is to
respond to the learner in a way intended to change his or her behavior
toward and educational goal. Instructional interactions have two
purposes: to change learners and to move them toward achieving their
goals." (Wagner 1994)

More below, but it is important not to confuse interaction itself with:

  • the goal of interaction (eg, leading students to reflection)
  • the effect of interaction (eg., changing behaviour)

On my account interaction is nothing over and above the exchange of information.

Siemens continues,

The goal of interaction is to lead students to a point of reflection
that causes them to evaluate existing assumptions and then choose to
integrate or discard the new information. By itself, interaction has
very little value. I've often interacted at length with concepts, only
to find that I've not been focused - and everything I read/heard was a
blur. Why? Because I was not engaged in reflecting/validating the
content I was exposed to. Effective (effective being defined as
achieving the intended purpose of interaction stated earlier)
interaction, then, is a process of awakening a students' internal
reflective processes.

OK, this discussion looks at the role of interaction as it relates to learning. The suggestion
here is:

  • effective learning requires reflection and validation
  • reflection and validation require interaction
  • therefore effective learning requires interaction

This sounds like a good argument. However, though the premises are sometimes true, they are
not always true.

  • suppose, for example, I wanted to learn how to make microwave Kraft
    Dinner. I read the instructions, follow them, and assuming I've measured
    my water correctly, successfully demonstrate that learning. Reflection
    and validation are not required to learn how to make microwave Kraft Dinner.

  • suppose, for example, I want to work out a concept for myself. I write
    a draft, review it, make changes, think about it, trace implications,
    and finally convince myself that I have it. I have engaged in reflection
    and validation, but have not engaged in interaction.

This is not to invalidate the original argument. But it is important to keep in mind
that the need for reflection, validation and interaction varies according to context.
Rather than simply say that we need them, we need to say why we need them. This in
turn leads us to a greater understanding of the goals of interaction in a given
context, and therefore, to the best mode of interaction to employ (as defined by my
variables above.)

To expand on this a bit:

  • It is one thing to learn how to make Kraft Dinner and quite another to
    learn the essentials of integral calculus, and quite another to comprehend
    the importance of Descartes's Meditations, and quite another again to learn
    how to play darts.

    Different types of learning require different methodologies (which is why
    I lose patience with people who say things like 'The key to learning is...'
    of that 'Learning is essentially about...'

    Various accounts of different types of learning exist, the most quoted
    of which is Bloom's taxonomy. But we can approach the subject with a
    rougher taxonomy:

    • Learning of skills, methods, processes, etc ('knowing how')
    • Learning of facts, data, statistics, dates, etc. ('knowing that')
    • Learning of concepts, representations, abstract systems, etc ('knowing why')

    To express the relation very roughly, imagine a gradient between learning
    by rote and learning by understanding. In the former, the emphasis is on
    observation and experience, recitation and practice, while in the latter,
    the emphasis is on the formulation of abstracts and the placement
    of new information within an overall mental model or conceptual scheme (and
    involves the construction and refinements of these mental models and schemes).

    Then from the list above, the first relies most of all on learning by rote,
    and the best means of learning is by doing - by repeating multiplication
    tables, by memorizing key dates, etc. Reflection and validation are of
    minimal importance; repetition is the order of the day. The second relies
    on learning by association - one thing is like another in just this way. In
    such cases, reflection and validation are of maximal importance as it is
    necessary to constantly refine and adapt one's conceptual model.

  • While interaction has a role to play in rote learning ("watch and then
    repeat after me") most discussion of interaction deals with its role in
    learning by understanding (hence the argument that began this section).

    In such cases, interaction plays a role in the four major forms of reasoning:

    • Definition - acquiring, through conversation, a common understanding of
      language and syntax

    • Description - acquiring, through conversation, a common understanding
      of reference and representation

    • Argument - acquiring, through conversation, a common understanding of
      what beliefs may be inferred from definition and description

    • Explanation - acquiring, through conversation, a common understanding
      of the theorietical constructs, world views, or models within which
      definition, description and argument operate

    Again, as in the previous case, we obtain a gradient. Definition, for example,
    may proceed by stipulation (in a learning environment, it almost always
    operates by stipulation). Description involves a mutual sharing of stories
    and other experiences, but the contents of each person's message are (typically)
    unrelated to each other. In argument and explanation, the messages begin to
    inform each other, affecting the truth or meaning of previous messages.

To approach the question of interaction, then, is to take the following steps:

  1. To identify the type of learning involved, and therefore, the degree of
    reflection and validation required

  2. To identify the type of reflection and validation required, and therefore,
    the degree of interaction required.

The typical learning scenario will not be unidimensional. That is, it may involve
a bit of 'learning that' and a bit of 'learning why'. It may involve some definition
of key terms and a bit of argumentative reasoning in order to derive a new statement
of fact (a typical math class, for example, is very much like this. Then you get
a list of problems to solve, learning by rote the concepts you explored in class).

Siemens continues,

Learning is essentially communication with an objective or goal. As
such, it is helpful to modify existing communication models as an
"interaction cycle".

I disagree with the definition of 'learning' proposed. If pressed, I would define learning as
the acquisition of new knowledge. Specifically, there can be different types of learning,
corresponding with the different types of knowledge:

  • learning that
  • learning how
  • learning why

The phrase 'communication with an objective or goal' is a description one way to
facilitate learning. But even the requirement of an 'objective or goal' is not necessary
in all cases of learning. Consider:

  • You are teaching someone how to speak French by uttering words and having the
    learner repeat them. This type of learning by rote is most clearly goal-defined: there
    is a specific outcome (correct pronounciation) intended.

  • You are leading a class discussion in which each participant advances his or her
    views on animal rights. This is an instance of reflection and validation, with the use
    of argument as the reasoning form. There is no clearly defined goal or outcome
    definable for this exercise, and yet the process of interaction is likely to help some
    people more clearly articulate the conceptual model governing morality in their
    world view (and it may help others revise their conceptual model, and it may
    help others develop a conceptual model, and it may give others practice in oration,
    and it may allow others some necessary space in which to vent).

In general: the clearer the objective or goal of the learning activity, the more concrete
the learning (that is, it is more likely to be by rote than by understanding, and the more
likely it is to be definition or description than it is to be argument or explanation).

In a certain sense, interactivity and learning objectives are in opposition to each
other: the more interactive, the less certain the outcome. Indeed, if an instructor
has clear learning objectives, then interaction (especially unfettered interaction among
the students) can be an impediment to that goal.

Siemens now begins to describe the interaction cycle. (Note here - and interestingly - that Siemens, like Anderson, wants content to be an agent:

The interaction cycle is as follows:

  • Sender - this can be: content (content has two components - the
    content itself, and the expression of content - i.e. html document - in
    this case it is the latter. Content itself is the message),
    student-student discussions, student-instructor, or interface (i.e.

  • The message (generally content expressed via one of the four
    senders listed above). The nature of the message is influenced by the
    conditions mentioned in Stephen's articles

  • Receiver - this is the learner who receives the message

    ... (list continued below, after my comment)

Now let me turn to the interaction cycle as described above. We need to draw out
and explicate some of the components:

  • The agents - these are the entities (the people or the computer programs) that
    are actually doing the communicating

  • The medium - this is the means by which the communication takes place (eg.,
    by internet, by telephone, by tapping on a water pipe)

  • The data - this is the actual message being sent - binary code, analog audio
    transmission, tapping noises)

  • The expression - this is the manner in which the data is encoded. It has the
    following components (in many communications, these are implied):

    • structural information - message headers, chapter headings, spaces
    • syntactic information - an account of the language being used, eg text/html, pdf, gif
    • presentation information - eg., font size (may be empty in some communications)

  • The content - this is the part of the data that is not part of the expression and
    is identifical by having a semantical rather than a syntactical or structural import
    (that is, we recognize content by the fact that it has meaning)
  • The information - this is the part of the content that is new to the receiver, that is,
    the part that (to allude to my previous posts) reduces in the receiver the number of
    possible states of affairs in the world

We now continue with Siemens's cycle...

  • Interferance/distractors - these are concerns that can obstruct
    the intended learning. Some are controllable by the sender (e.g. through
    good design (content), instructor training) others are controllable by
    the receiver (the state of mind of the learner), and others are not
    controllable (i.e. internet traffic).

This part of the taxonomy is intended to distinguish between
those parts of the communication that are not a part of the message
and those parts that are. Of course, we can see by this phrasing
that this phrasing is vague. We can identify the following types
of interference or distractors:

  • agent interference - either in the sender (speaking without
    thinking) or in the listener (not listening, busy elsewhere)

  • medium interference - non-data noise that is a property of
    the medium itself and not the data being sent (though it
    often becomes a part of the data received): the hiss of a
    tape recorder, packet loss, pipe noises caused by the
    movement of water

  • expression interference - this may include the effects
    of encryption, but also includes those meaningless
    strings of HTML text that MS word imports into HTML
    documents; it also includes coding errors, grammatical
    and spelling mistakes, mispronounciations, and the like

  • content interference - any of a variety of logical fallacies
    including vagueness, ambiguity, faulty reasoning, personal
    attacks, and the like

  • informational interference - that part of the content that
    is not information - restating the obvious, for example

In other words, an account of interference is not something which is described as a part or a step of
the interaction cycle, but rather, is an additional layer of description applied to all parts of the
interaction cycle

Continuing with Siemens's interaction cycle...

  • A process for feedback/validation/correction to ensure that the
    intended message (learning) has been received and
    integrated/validated/rejected by the receiver/learner.

As in the case of interference, feedback / verification mechanisms
operate at all points in the cycle. Specifically:

  • agent verification - introductions or handshakes (in both the
    technical and non-technical sense), presentation of identification
    or authentication (PGP, coded 'shave-and-a-haircut' taps)

  • medium verification - test patters, pings, and other sorts of

  • expression verification - syntax checkers, people who complain
    about grammar and spelling, requests for resend

  • content verification - parity bits, acknowledgement (by restating)
    that the content received was the content sent

  • informational verification - application against known information,
    response in the form of questions or requests for clarification,
    assessments of verisimiltude (eg., a response asserting that the
    information sent is fase)

Siemens's cycle, continued...

  • The manifestation of learning through some type of
    change/expression by the learner (this validates not only that
    reflection occured via interaction, but that the reflection was
    significant enough to evoke a change in behaviour in the learner). This
    is often neglected in many learning models...because learning is
    evaluated via a test/essay, rather than through a more authentic
    assessment approach - like performance, portfolio, presentation

This lies outside the interaction cycle. Arguably, evaluation of learning by changes
in behavious are useful in learning, but are not essential in interaction. An
interaction may occur with no manifest change in the receiver (this is why we need
a separate testing or evaluative process).

That said, there are two very different forms of evaluation (as is
well known), and they play very different roles:

  • formative, or correective - this is the sending of information
    in direct response to the display provided. It is best thought
    of as 'back propogation' - a reflection of the original content
    with an evaluative addendum to be integrated into the learner's

  • summative, or evaluative - this is an assessment for external
    purposes of the degree of learning attained. It is best thought
    of as 'forward propogation', and may consist of the evaluative
    addendum (a grade, for example) alone (that is, not containing
    any part of the learner's display

The mode of evaluation - test, essay, protfolio - most usefully
employed depends on the type of evaluation being performed and the
type of learning being assessed. In general (and with exceptions):

  • 'learning that' is usually measured with some sort of test
  • 'learning how' is usually measured with some sort of demonstration
  • 'learning why' is usually measured with some sort of creation

Note: parts of this note are original to me (eg., the four basic types of reasoning), parts borrowed (eg., references

to Bloom) and parts probably previously stated in other works about which I have no knowledge. Supply your

own references.

Stephen Downes Stephen Downes, Casselman, Canada

Copyright 2024
Last Updated: Jul 16, 2024 07:46 a.m.

Canadian Flag Creative Commons License.