Python/New/Learn

[dn]

To the OP: there are many MOOCs available on the likes of the Coursera
and edX platform. (rationale, below)

Disclaimer: I work on courses on the edX platform (but not Python).


On 06/05/2022 23.37, o1bigtenor wrote:
> As we're now discussing tutorial methods - - - -
> 
> On Thu, May 5, 2022 at 8:57 PM Chris Angelico <rosuav at gmail.com> wrote:
>>
>> On Fri, 6 May 2022 at 09:53, Grant Edwards <grant.b.edwards at gmail.com> wrote:
>>>
>>> On 2022-05-05, Mats Wichmann <mats at wichmann.us> wrote:
>>>
>>>> Without having any data at all on it, just my impressions, more
>>>> people these days learn from in-person or video experiences.
>>>
>>> I've always been utterly baffled by video tutorials for
>>> programming. There must be people who prefer that format, but it seems
>>> like absolutely the worst possible option for me. You can't cut/paste
>>> snippets from the examples. You have to constantly pause them so you
>>> can try out examples. Sometimes it's not even easy to read the
>>> examples. Perhaps if there was an accompanying web page or PDF...
>>>
>>
>> Video tutorials make GREAT sense for learning complicated programs
>> like Adobe PhotoShop or some 3D game design engines, because (a) most
>> of what you need is in the menus somewhere, but it's hard to find; (b)
>> you can aim the tutorial at a specific version, and it'll be the same
>> for all users; and (c) you can talk about it at the same speed that
>> people can follow along.
> 
> Respectfully - - - I would disagree. FreeCAD is a mind bogglingly
> complex architecture (a lumping together of a lot of things without a real
> central vision imo in fact) and a video tutorial would seem to make sense
> - - - except - - - trying to see what is being done is at best tricky - - - at
> worse - - - impossible. If the instructions were text - - - well you could do
> things one step at a time and you wouldn't have to scroll back and forth 10
> times trying to see exactly which part of what was the mouse applied
> to and then which toolbar . . .   .  One would think that something that
> you manipulate visually would be best served by video instruction.
> 
> Personally I find video instruction the most difficult to follow and the most
> awkward. And then if the speaker is not easily understandable or is using
> translated terms (not necessarily the same as those in the program
> itself) well - - - the frustration level is most definitely NOT small and
> the amount of learning - - - not that large - - - especially given the
> effort needed to create video tutorials.  IMO video is too often used
> because its there - - - not because this enhances the experience.
> 
> (Or the instructor is an academic who is reading their video screens
> - - - you know - - - like the 85 or 90% of the profs at the uni - - - - )
>>
>> Video tutorials do NOT make sense for anything where you'll be using
>> your own editor, typing in code, and having it behave the same way.
>> There's nothing to point-and-click, and everything to type.
>>
>> But some people start making tutorials of the first kind, and then go
>> on to make some of the second kind, thinking they'll also be useful.
>>
> I think you, that is Chris, are a very generous person. My experience
> has been that many consider video tutorials to be cool or sexy or of the
> highest art - - - and never even consider the uncool, mundane, boring,
> old, text option.
> 
> I'm wondering if the difference is that in general education itself less
> and less emphasis is placed on reading (and comprehension) skills.
> This fits alongside a return to pictographic language supposedly to
> assist in multi-lingual barrier reduction.


To all:

The problem with some of the advice given in this thread, eg using
StackOverflow or YouTube videos, is that a beginner (particularly) has
no measure of the material's quality. Both platforms are riddled with
utter-junk - even 'dangerous' advice.

Some of such posted-content has been encouraged by 'teachers' who think
the old adage "to understand something properly one must be able to
explain it to someone else" can be translated into an (effective)
learning practice. Which is fine, until the results are posted on a
public forum without any qualitative assessment - I've even come-across
'teachers' (the quote-marks could be taken to indicate serious question
or in some cases, disdain) who think that allowing/encouraging students
to post such is "encouraging the student". Sadly, (IMHO) its effect is
'intellectual pollution' for those who come-after, and the generation of
an unrealistic ego/self-assessment on the part of the post-er.

Then there are others which have been posted by well-meaning
individuals. These may be motivated, similarly, by ego or altruism. Few
know what they're doing (from a training PoV) and the fact that some
such contributors stand-out so far about 'the crowd' speaks to this. How
many have you seen which fail to account for the differences between
your system and the one the author uses? How many seem determined to
show that by clicking 'here' and selecting 'there', the most complex of
environments can be 'mastered' by anyone, clickety-click - without a
single 'if this (step) goes wrong, here's what you might see/need to do...?

The advantage of a text-book, which has been properly and competently
edited and reviewed, is that the content has passed the 'many eyes'
test. Each explanation has been examined critically, and is no longer
(only) the author's own view-point, way of explaining, or even opinion;
of how things could/should be done.

There is little learning-value in asking people to learn-by-typing -
although I suppose one might learn something of the importance of
accurate syntax. (how could an author hope to cover all the possible
errors of this type that a learner might make whilst transcribing 'my
first Python program[me]'?) This is why many texts offer a download, and
why my in-person classes start with trainees being given code-sets to
edit/customise.

When we set up our MOOC (remember, not Python, but the message is much
the same) we swallowed two lots of 'Kool Aid'. Firstly, that 'the modern
student' wants everything on video. Secondly, that 'long lectures' are
'out', and we should look at +/- seven minute talks.

The edX platform provides excellent usage statistics, both on a cohort
basis, and over time. These reveal (as a group) that most trainees
either don't start, or certainly don't finish, an entire video - without
regard for 'length'. Surprise!

Unfortunately, the stats don't differentiate between a person watching a
video for the first (and only) time, and someone who goes back to watch
it again, eg after attempting a quiz/progress challenge, not doing well,
and wanting to review. However, the more 'repeat views' that have taken
place, imply that the average must shift to revealing more-and-more
trainees who don't even open the video.

The 'lecture' advice does fit with the way people organise self-paced
study in-and-amongst all the other facets of life. Many working
full-time, find that a 'short burst' can be managed whenever an
opportunity presents itself - even during a commuter bus-ride. Whereas
one would have to make a more deliberate plan to sit down for one whole
hour at a time.

Yes, it can be more complicated to design a sensible (alternate)
presentation-structure. However, as we have discovered, this works well
for 'us' as well. Replacing a long video-based lecture is much more
expensive than replacing a short one, eg if a new feature has been added
by a recent software-release - particularly if we are only talking about
a small 'tweak'.

The claimed-advantage of MOOCs is that trainees from anywhere in the
world can access 'the best' in the world - usually this is a claim
against the institution, but sometimes deservedly, the lecturer. I have
to say though, some of these claims are not justified. A staff-member at
a local-institution might have a much better explanation! I'm distressed
by many (usually American) institutions claiming world-status but who
use teaching-techniques that remind me of my own days as an under-grad
years, um-er, many (many) decades ago. The world has turned, and
moved-on! Similarly, there is a considerable difference between the
teaching styles used in much of Asia, eg India and China; and the more
student-centered approaches used in 'the west'. Would you prefer to
write some code-examples using simple data-structures, or to sit down
and be tested on your ability to recall the range of list-methods and
their parameters?

The 'professionals' will often have a veritable studio and likely
video-staff involved in ensuring the output-standard (similar to above
reference to books and 'reviewers'). This ensures that as many 'takes'
as necessary are performed to ensure quality. Whereas our 'YouTubers'
works are frequently punctuated by 'ers', 'ums', and re-typing errors
during demos, and the like - a 'stream of consciousness approach'. Our
course-structure considerations extend down to the detailed-level of
'story boards' detailing which topic will be covered 'when', how it will
be covered, what the sequence will be, and so-on. This means that 'my
first Python program[me]' will not be confused with 'how to install the
Python interpreter' - another lesson the earnest-YouTubers could emulate
to-advantage (also likely the +/- seven minutes advice).

Sadly MOOCs are not the 'be-all and end-all'! For a start, there are
many students who cannot gain (easy) access to the Internet - a rather
basic requirement to 'attend'. Also (and as appears elsewhere in this
thread) the 'lecture' approach to training is not 'the best way' - it is
a 'factory approach' to teaching (one 'sage on the stage' speaking to a
large number in an audience). However, the real question is: is
attending a lecture the best way to learn?

That said, a MOOC will usually be structured so that there are
video-recorded lectures, but that much?all of the content also appears
in the directive or accompanying text. Thus, both 'YouTube' and 'text
book' media; combined into one presentation. In addition there are
built-in 'Discussion Lists' where all the trainees working on a
particular topic exchange views, ask questions, help each-other through,
etc (unlike StackOverflow which is very wide and diverse collection of
conversations, with no indication of 'level', competence, need...).

Because a MOOC's text is both copy-paste-able and/or code-examples can
be provided for download (within context, rather than as a book-wide
.ZIP archive), those advantages are also available (as previously
discussed 'here').

I don't see any sort of 'cut-over point' where text-books give way to,
or are over-taken by videos. Complexity of the subject is not the issue
or inflection-point. Some things are best presented in schematic form,
for example - but although you could portray the try...except construct
that way, sooner-or-later the presentation must revert to text - what is
code if not 'text'?

What can be a very important point, (as previously mentioned in this
thread) MOOCs are produced by folk who are used to presenting
information, and to the discipline of writing 'papers'. Accordingly,
dates and version numbers are included as a matter of course. Thus,
there is a very low possibility of finding that 'this version' is
different from the one you have (assuming you installed the one the
course recommends/requires). How many 'amateurs' think to do this - even
bloggers who don't see themselves as primarily presenting 'learning
material'.

Speaking personally, I'm like the person who said that learning from
videos is more difficult that text-based materials. The most obvious
explanation for this relates to when/how I grew-up and enjoyed?suffered
my education. We didn't have videos, or even (domestic) VCRs (until the
very end of my high school years. So, everything was 'in a book'.

These days it is quite a different story. Whilst I'm prepared to accept
that young 'digerati' may prefer video, even that assertion is clouded
by the nature of watching a video not only being a change in medium but
it is also "asynchronous". Accordingly, the advantages of 'self-paced'
and being able to 'watch' at a time of one's own choosing, cf attending
a (synchronous) lecture at a specific time in a particular location.

Interestingly, ans somewhat-related, Librarians report that there is
little difference between the age-groups when library-users are offered
the choice of the same material either as an eBook or in the form of
'dead trees', people select the latter/old-style more than 90% of the time!

However, I may disagree (for my own case) with the person talking about
'notes'. It is easily and widely proven that people who make notes will
improve their learning and "retention" rates. It is worth knowing that
when we read something we are overly confident of remembering points
later. Indeed, even reading and re-reading text (as many of us were
taught to do for exam-prep) is a remarkably-ineffectual way to learn
(for that reason).

Making notes has a first-pass value of helping us to consolidate
new-learning and link it to existing knowledge (a key to retaining
knowledge of this 'new stuff' effectively). There is something about
writing things down, even after reading them, that helps 'engage the
brain'. Indeed, to the ridiculous point that when I write something down
I may not ever need to go back to my notes to be able to recall the
fact(s), eg my shopping list! So, why write them down then? As
mentioned, it is the very act of writing which effects this 'learning'.

Bad news though: typing into a computer does not create this
psychomotor-benefit. There is something in the 'picture' of writing
which cannot be re-created in the text-only alternative of a word
processor!

Secondly, if one takes notes today, and re-reads same tomorrow, and
again next week; the 'learning' is not only effected, but consolidated.
So, "repetition" at that level is more likely to lead to "retention".

Now applying a little psychological 'warning'. We (humans) tend to
approach devices in a particular mode, eg most people regard their car,
bike (etc) as a means of getting from 'A' to 'B'. Yes, there are others
who have a different perspective, and want to go racing, for example.
The same idea of 'this is what the machine does' or 'this is what I use
if for', applies to computers and smart-phones - and again, those of us
who regard them as a machine to 'master' or which which to build 'apps'
and develop software are akin to the 'racers'! Thus, if someone mainly
uses their device to watch movies, tweet, put their face in a 'book',
send 'telegrams', or record humorous events/Dad-dancing; is the 'mode'
in which they habitually use their device and thus the 'state of their
brain', congruent to or encouraging of "learning"?

The technology one poster 'here' requested, that video and text might
appear together and synchronised is (surprisingly?) available today and
has been for five~ten years. It is a base-feature of HTML5's video
facility. IIRC something similar is also reproduced and available
through the YouTube 'engine' (I don't use that). It is not common
because it is MORE WORK! In this part of the game, most people seek a
technology which will take the video's audio-track and transliterate
(speech-to-text). I've had to take up a practice that I never previously
used, and write every single word of a presentation (like a formal
speech). This forms the text component of a future video, but I (or the
presenter) then have to read the words whilst making our voices sound
more conversational - again, a skill most have not acquired, even having
been required to read-aloud for the class in junior/grade school!

This form of 'multi-media' forms one aspect of my research into
Cognitive Psychology (how one learns) and tool-building to ease
presentation-production and/or to promote learning.

Finally, I was amused by @BigTenor's comment about "pictographic
language". Some years ago, a theory was proposed that different people
learn in different ways, eg some prefer to hear whereas others to read.
Sadly, whilst this idea has immediate appeal because a casual
introspection will have many readily saying "I'm a ... learner"), the
theory has been debunked - both by primary-research and by
secondary-researchers critiquing the 'evidence' which didn't actually
make a proper attempt to prove the theory in-reality. However, it can be
a useful 'guide' - as long as that is as far as it is ever applied. A
lot of 'damage' is done by people telling themselves (or being told by
their 'betters') that they can only learn in one particular style (to
the exclusion of all others). Don't you believe it!

As above, some things are best explained one way, and others another, eg
an annotated map cf a list of directions. @Stefan's French spelling and
phonetic pronunciation guide in-combination illustrates this nicely.
(see also 'try...except', above) The interesting observation is that
whilst I may grasp some new concept from a text explanation in the
morning, I may need a diagrammatic approach to convey something similar,
this afternoon, ie we don't always learn using the same approach, every
time! Which leads to the idea of having multiple means of conveying
information, eg the MOOC's text and video content, because we hope that
if one doesn't 'do it for you', the other will!

Apologies for the lack of 'Python', but thanks for the discussion (and
am hoping you won't mind me passing it around amongst my colleagues -
perhaps there's not much that is 'new' (for us) but the fact that it is
being discussed by folk who are primarily practitioners is most helpful!)
-- 
Regards,
=dn