[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: list matching questions

If the intent is to associate a word with a re-phrasing of the word's
meaning, it will be difficult to keep the re-phrasing from exceeding
the test level, unless it is for a test at a high level.

Questions should not be failed due to an ignorance of geography or any
other fields of knowledge besides Lojban. For instance, I don't know
of any Japanese birds, offhand. Is it a mountainish-chicken? That
doesn't ring a bell.

Including so many options is time-consuming. It would be difficult to
time how long it takes for the brain to access Lojban, and tell it
apart from how long it takes the brain to sort through all the
options. Live, timed tests are incredibly important. There is little
point in testing mere persistence, since a conversation that gets by
on persistence is not a very fluent conversation. Facility of recall
is a sadly overlooked ability in this particular language community,
and we need to put rewards in place for that to change.

I'm not sure this style of question can avoid all three of these pitfalls.


On Sat, Oct 17, 2009 at 10:39 AM, Pierre Abbat <phma@phma.optus.nu> wrote:
> On Saturday 17 October 2009 10:15:38 Jorge LlambÃas wrote:
>> Sounds good to me. This list probably goes beyond level 1 vocabulary
>> though.
> That list goes up to level 4, as "vombatu" is a type-4 fu'ivla
> and "ma'arjipci" is a word found only in "The Restaurant with Many Orders".
> Someone who knows how to analyze a lujvo should be able to figure out which
> phrase matches "ma'arjipci", though.
> How should that sort of question be implemented in the database? And how
> should it be scored?
> Pierre
> --
> .i toljundi do .ibabo mi'afra tu'a do
> .ibabo damba do .ibabo do jinga
> .icu'u la ma'atman.