Is there a way to learn a word with different meanings that doesn’t confuse the program? I was trying to add different definitions for a word but even if it shows that it’s the same word it marks it as a mistake each time. Is there any way to correct this??
Hi @BerryM,
If I have understood you correctly, I have thought about this and here are perhaps some suggestions that may help.
Have a look at the solutions offered in this thread:
But also:
MemRise chooses alternatives from the same column so if you have say AAA could be BBB, CCC, DDD, or EEE, why not create a course with 8 columns so only one alternatives ever appear in one column. (I am assuming you have lots like this that would go into the same column to avoid a column with only one answer.)
So the column entries would be
AAA BBB AAA CCC AAA DDD AAA EEE
You set up column 1 to test column 2 but NOT column 1 to test column 4 but 3 to test 4 etc
the next entry line in your database could be
FFF GGG FFF HHH GGG III FFF JJJ (where FFF has 4 other alternatives)
Hope I have explained that clearly and that it helps.
1 Like
Sorry I think I didn’t explain myself propery.
I am trying to learn attributes of a word, like yin and yang which can be light or dark, cold or hot, feminine or masculine and so on, so I worked my tables like this
But when I try to do the test it doesn’t make a difference between the words, so if for example the right answer is yin, you have to choose the yin that the program has registered for that line, if not even though you picked the right answer it marks it as wrong, is there anything that can be done to fix this?
You could try something like this:
Yin (gender) = femenino
Yang (gender = masculino
Yin (brightness) = obscuro
Yang (brightness) = light
Yin (openness) = oculto
Yang (openness) = abierto
Yes, it’s clumsy, but at least you won’t be selecting Yin (brightness) when the interface wants Yin (gender).
1 Like
Use a third column for synonyms, like I described here:
1 Like
Hi @BerryM,
As I said above, MemRise can not cope with that.
There are four ‘solutions’ as I see it:
1) As I suggested - move the other pairs into different columns,
2) Make each entry unique in some way as @kaspian has suggested (he got there first)
3) Combine the entries so ying = femenino, obscuro, oculto (where any answer would be accepted) or
4) Combine the entries so ying = femenino obscuro oculto (where every word has to be given to be accepted). (If you go for this option you will need to set up alternatives in different orders!)
PS it sounds as if @cos has suggested another.
Hi @BerryM ! You could add an attribute column and put it show in tests. In that column you’ll insert what type of attribute you are looking for that particular entry (i.e gender, brightness etc).
If you need any help just ask
(I don’t know why it went as a reply to Kaspian, sorry)
1 Like
@DW7
Good suggestion: unfortunately I can’t get it to work. As far as I can tell, Memrise only allows one set of two columns to be set as the “test on, prompt with” columns. So you can’t set up a course to:
Test on col1, prompt with col2
Test on col3, prompt with col4
Test on col5, prompt with col5
etc.
You can only test on one column, and prompt with one other column.
Yes @sebT you are right.
To clarify, in my suggestion you have 8 columns
Level 1 will test col 1 & 2
Level 2 will test col 3 & 4
Level 3 will test col 5 & 6
Level 4 will test col 7 & 8
And the word in columns 1, 3, 5 and 7 are the same but the ‘even’ columns have four possible alternatives.
Remember you can duplicate each level and reverse the testing.
But the suggestions above sound much simpler!
1 Like
Ahhhhh… that’s how it works!
The bit I was missing was that you have to create separate, duplicated levels, each of which tests/prompts on a different set of columns.
Thank you! I’ll make a note on the other thread to point people here.
For reference, to help others round this difficult problem, here’s what you have to do:
1. Work out the maximum number of duplicate meanings in your set of words. For my German definite articles, this number is 4, because der can be (nominative M, dative F, genitive F, genitive Plural) - die also has 4 meanings.
2. The number of columns you need will be 2x the maximum number of duplicates (for each, you need a “column to be tested on” - German in this case - and a “column to prompt with” - English in this case). For me, this is 8 columns.
3. Because Memrise already provides you with a single set of 2 columns, I only need to add (8-2) columns: 6 columns.
4. Columns have to have unique names within the course, so I couldn’t call them all “German” or “English”. However, the column names will be mentioned in tests, so call them something sensible. I used German(2), English(2), German(3) etc.
5. Get rid of the “plural and inflected forms” column, as it’ll just get in the way.
6. These new columns are not attribute columns. (Non-attribute is the default setting for a new column, so you don’t have to do anything)
7. Arrange your columns alternating between “column to test” (German for me) and “column to prompt with” (English for me): e.g. English, German, English(2), German(2), English(3), German(3)… etc.
8. Type in one row per unique word in your set. So, I typed in one row each for der, die, das, den, dem and des: even though these words have 4,4,2,2,2 and 2 meanings each, they each get one row.
8. For each word, use enough of the columns to cover all the duplicate meanings. So, for der and die, I used 8 columns - for the others, I used 4 columns. Make sure that the word in each of the columns you’re going to test on (German for me) is identical. You can see this in the first screenshot, though the English(4) columns is off-screen.
9. Now duplicate your level enough times to cover the maximum number of duplicates. That’s 4 duplicates for me, so I need 4 levels. I duplicate this level 3 times to make 4. (Leave renaming of the levels until after duplication, as the Memrise level-duplication is a little weird about putting the duplicated level before or after the one you’re duplicating).
10. Rename the levels with sensible names so you’ll know where you are from now on (I just used Level1-Level4 for the moment).
11. At this point, the levels are identical. Now you need to set each level to use a different pair of columns to test on/prompt on.
12. The first level is fine: it’ll test on the default German and English columns, and ignore the others. The second level needs to test on the second pair of columns: German(2) and English(2). Click on the weird thing next to the “test on…” setting at the top (which looks like the end of a giant fork pointing diagonally), and set which columns to test on:
13. Will all the rows work in this level? Yes, because all of them happen to have something in the second set of columns (German(2) and English(2)). Here’s what it looks like:
- For levels 3 and 4, I changed the testing/prompt columns to be German(3)/English(3) and German(4)/English(4). But if you look back at Level 2, you’ll see that most of the rows don’t have anything in those columns. They’re not needed in Levels 3 and 4, because the word in them only has 2 meanings. Delete the unwanted rows from these levels, to get something like this:
So, how does this then work when you learn the course? Pretty well:
- You don’t get the problem of multiple (correct) options appearing in multiple-choice tests, so that you can click on one and be marked incorrect. My guess is that Memrise somehow treats all the “der” rows (for example) in the various levels as the same word.
- You do get multiple incorrect options appearing in MC tests, especially in later levels with only a few words. This is because Memrise runs out of possibilities to fill all the MC boxes - but seems to be clever enough not to duplicate the correct answer. It’s a bit weird, because you can guarantee that the answer which appears more than once is the incorrect one! But this is only a minor problem.
- The only problem I can see (so far) is that this workaround makes the order of the words in the course illogical: ideally, I’d teach people the nominative (der, die, das, die), then the accusative (den, die, das, die) and so on. Again, this is a minor problem, and nothing compared to the mess you get trying to get Memrise to deal with duplicates in any other way.
3 Likes