My Marking System for a Language Class Summarised.

“How do we mark kids in T.P.R.S. classes?” asks a C.I. newbie on the Yahoo listserv.  Assessment and evaluation are always hot topics.

First, definitions.  Assessment is observing how people are doing, and based on that observation

1. changing what you the teacher do and

2. giving students feedback in order to

3. improve student performance.

Assessment in a comprehensible input classroom means checking whether individual students (or the class) understand both statements and questions, and– if they don’t– clarifying. Weak choral responses?  No response?  Wrong answer?  Go back and clarify. The same goes for individual kids.  If Johnny can’t answer the question “What does ___ mean?”, go back and re-explain.  BOOM! that’s assessment.

Evaluation is assigning a number to student work based on a student’s performance in relation to standards (criteria).  This we do at the end of instruction.

Second, principles.  If you havn’t been in-serviced about “assessment for learning,” google it, I can’t explain it all here, but let’s make three things SUPER CLEAR.

A.   We only evaluate “final products,” not practice, homework, behaviour, attitude, etc.   The long and the short of this surprisingly controversial statement is that we expect our kids to do X, Y and Z, and the mark we ultimately assign them at the end of Year ____ of Blablabian Language Class should reflect how well they do X, Y and Z.  Whether Johnny is nice (or a total jerk) in class, and whether or not Suzy does her homework are irrelevant: what matters is how well they do in relation to criteria.

B.  We only evaluate what has been taught.  No random new vocab, no activities on tests the kids havn’t done, etc.  You test what you teach, period.

C. We always let ppl re-do work so they can improve their mark. In my class, this means that their mark is always their most recent writing, reading and speaking mark.

Assessment should be fast, simple and minimal, because people learn from interesting comprehensible input and not from testing, and because teachers need to see their families & have a life 👍.  

Now, here is what I do.


  • 10% all work done during year
  • 10% two culture projects (5% each)
  • 80% final listening, reading & writing exam (speaking for Level 2 & up)

First, this is how I mark writing.

Second, for reading and listening marks, the easiest thing is the exit quiz.  One a week for reading and one for listening.  Speak– or write on the board– five sentences.  For listening, the kids write in the target language and translate into English.  For reading, they translate into English.  You can have them trade papers and mark.  You can do an exit quiz in ten minutes, the marks recording is quick, and you’ll know immediately if they understand or not.

Another great idea– thanks, Ben Slavic and also ironically legacy methods– is dictation.  For dictation, read a very short (e.g. 10-sentence) story aloud.  The kids write down what they hear.  Then, they translate.  Finally, you project the story, and they fix their errors.  I have no idea how to mark dictation.  Suggestions?

I also mark my kids’ comics for a reading mark.  This is Adriana’s idea from her special-ed course and it’s simple & excellent:

  • read the extended version of the most recent story (if using embedded readings, read the most complex one)
  • make a 10-16 panel comic (Internet clip art etc fine) of the story
  • each panel MUST have at least one “narrating sentence,” (e.g. “John was hungry and went to the store”)
  • each panel MUST have EITHER dialogue or thought-bubbles (characters either describing themselves, e.g. “I like girls” or thinking, e.g. a girl thinking “I am hungry”)
  • the pictures MUST accurately support the meaning

This is what Adriana calls “deep reading.”  It makes the kids read, extract the main points, and illustrate them.  It’s also fairly easy, and the results– which I say should look decent, be in pen, have a bit of colour and not use lined paper– go on the wall, where the kids read each others’.  If you have Adminz or Headz who get excited about Technologiez, the kids can make them using Computerz!

The comic marks are cumulative: if they don’t do them, they get zero until they are done.  A comic typically takes a kid about 30-40 minutes to do, and is the only homework I assign (other than occasional reading).

Third, I do not assess “speaking skills” in Level 1.  Shocker, I know, but there’s a few reasons.  I know exactly how well each student can talk.  And–unless your modus operandi is hand out worksheets and put your feet up– so do you. Level 1 is where we lay foundations, where we plant seeds, and if the kids are getting a load of input, these seeds, as Ben Slavic puts it, will bloom in Level 2.  And practising speaking does not develop speaking skills.  My kids can talk– a lot— but I don’t pressure them with “prepare for your oral test” nonsense (how do you “prepare” for an oral test anyway?!?).

I assess speaking skills in one three-minute interview at the end of Level Two (and up).  These interviews are 100% random questions.  For Level Two, I am basically looking for their ability to understand and respond (they should be able to answer a question like “What do you like to do?” or “What did you do last weekend?” with 1-3 sentences).  I also get them to tell me a story.  For Level 3 and up, they should be able to provide longer answers and ask me questions as well.

Now, how does it all hang together?

Well, what I have is a “rolling” gradebook.  That is, their mark always reflects their latest performance (the exception:  the comics, which are all added together under the reading mark).  Their mark is always the most recent thing they have done (for writing, listening and reading).  We do assessment during the story-asking process for listening and reading skills, and at the end of a “story cycle” for writing.

This allows me to say “right now, you are getting ___%” and also “if you blow off coming to class, reading etc, your mark can drop.”  With T.P.R.S., most kids do quite well, so I have very few parent or kid complaints.  In the last three years, one parent has complained…and when I told her that her son missed an average of 60% of classes, she said “oh, I see” and that was that.  Also, the kids know that if they have a crappy day on writing assessment day, they can boost their mark next time around.  Generally, the marks are quite consistent.

When I get the dreaded question “how can Maninder improve her mark?” from a parent , I say all you can do is listen in class, read, and try to find easy listening or reading in Spanish outside of class.  This also happens to be the only thing that really helps.  There is no edubabble like “look in your portfolio and revise your ____” or “master your ____ verbs” or whatever (feedback is useless: if it adresses the conscious mind, it will be either forgotten or unavailable in real-time interactions).

Teachers say “well, a kid could do nothing all year and then try to ace the final.”  Yes, they could.  But…for one, as Bill VanPatten reminds us, the “work” in a language class is processing input.  No processing input = no (or very little) acquisition. And sure enough, the kids who blow off listening and reading (and/or skip classes) see their marks drop.  If they keep re-writing the same story, and don’t learn new vocab, the end-of-year writing criteria takes that into account.  They are also expected (in Level 1) to write about 75 more words in each 40-min story-write and 10 more words per 5-min speedwrite (describing themselves or a picture) than they did the previous time.  Kids who skip etc see their marks drop, because their output has less varied (and less well-used) vocab, and lower wordcount.

In 2015 Spring semester, I had two kids who missed about 50% of classes.  Sure enough, at the end of the year exam, their wordcounts for stories were in the 300 range (the rest of the class could do 600, and the top kids almost 900), their 5-min speedwrites were about 60 words (every other kid was over 100), and their reading and listening comprehension was significantly lower.  Most interesting:  these kids did not badly…but would have failed under the grammar grind/communicative system.

So, they have “ongoing” evaluation where their marks are always the last thing they’ve done.

For final evaluation, I do a 45-min story write and a 5-min describe-a-picture write.  I use this Timed Writing Rubric (from Kristin Duncan) at the end of the year.  In Level 1, the story word target is 800 words in 45 min; for Level 2 it is 1,000.  For the picture,t he target is 100 words in 5 min.

For listening, I do dictation:  they listen to a 20-sentence Spanish story, write it in Spanish, then translate into English. For reading, I give them 20 questions relating to the embedded (long) reading versions of the asked stories, and they have to hunt through the stories (re-reading! reps!) for answers.  The answers are 1. copy from the text and 2. translate.  The evaluation here again is of comprehension, not output.

Generally, I use 80% final evaluation (read, write, listen; oral for 2s and up), 10% culture projects and I make the entire year’s work worth 10% for their final Spanish mark.  Interestingly– I kept stats this year– there was very little difference between kids’ final exam marks and their during-the-year marks.

This has never happened, but I tell kids, if you blow it on the final writing or reading exam, you can re-do it.  As Vancouver T.P.R.S. teacher Steve Bruno remarks, one of the great things about T.P.R.S. is low test anxiety.  The kids know what they have to do, they can do it, and there is zero difference between a final exam and a regular test.

So anyway…to sum it up:

  • students are always marked on their most recent work
  • no oral evaluation for beginners
  • this is real “assessment for learning” where only final products are evaluated with reference to criteria
  • there is no marking of attitude, homework, participation, etc.  If kids screw around, or don’t work, I deal with it– but not by using marks as carrot or stick


  1. Mr. Stolz, I loved some of these ideas. I also use and love the comic idea, as in an illustration w/ captions to retell. It’s from my elementary days. I do think it’s cute that you call them “marks” -you adorable neighbor to the North.

    I am confused about the necessity of separate listening, writing, reading, and speaking assessments, as opposed to a more holistic assessment of language and/or literacy. Is this data you feel you can use to drive instruction, or just something your school requires?

    Thank you for sharing.

    1. I don’t have a way of assessing “holistically” or whatever. The marks program blends the 3 or 4 categories together to produce a mark.

      In a CI class, marks provide very little useful feedback to the teacher. The real feedback is comp checks esp. on barometer kids. If you figure out where meaning breaks down and then address that, and provide lots of CI, things tend to work out nicely.

      1. “In a CI class, marks provide very little useful feedback to the teacher.” Yes, that’s evident in the assessment instruments you are using. So why use those assessments? Could you not use the “real feedback” from comprehension checks you mentioned?

  2. You will obviously do formative “breakdown” assessments during the storytelling process. I suppose my question is why would the written/illustrated retell after the story is done–why would that not be enough to provide both grades and feedback on how well students understood a story that they listened to and read? This seems like a fair assessment that aligns to the instruction you used and allows students to respond to/interact with the language they heard and read in class.

    So why the additional speaking, writing, and listening test? Are those mandated in your school?

  3. A question about the TW rubric you linked to: the numbers at the top of the columns aren’t the points, are they? If they were, a student who is almost there would get a 75% (3/4 on each part). How do you translate those descriptors to numbers?

    1. They get a mark out of three for grammar and a % built on word count. A kid who gets 2/3 for grammar and 90/100 for words gets about 75%.

      Very few kids get less than 2/3 for grammar.

      1. Right, that’s your ongoing grading schema. I am interested in whether you have a rubric for that 3-point grammar score as well.

        But in the post’s section on final evaluation, you have a link to a Timed Writing Rubric from Kirstin Duncan that seems to have a scale of 1-4 (

        I was wondering whether those numbers were the actual points: if so, a student who:

        -met the word count
        -wrote in long compound sentences
        -made limited enough errors not to affect comprehension
        -used a variety of recent vocabulary

        would get 3/4 in each row, for a total of 12/16, or 75%, which seems a very low score for those descriptors.

        If they don’t indicate the points, then what are the numbers at the tops of the columns?

        Sorry to be unclear, and to keep pestering you about this. I’m just trying to move over to a consistent grading system to provide predictability and support for my students as I continue to move in a CI direction from a grammar focus.

        Thanks for your time and all your great posts here on what really makes for good language teaching!

      2. I’ll tet back to you in a few days. I have to go and re-read my own post!

        But one thing to keep in mind: eval. should be fast simple and as infrequent as possible. It does not help kids and it wastes teacher time.

  4. I waffle between your system and Lance’s. I did a year where only the last grade is in the grade book, no hw, and final exam overrides everything w retake if you want (nobody did). Grades were in line with work during the year, and I put estimated “proficiency level” in the report card notes.

    I loved it that the capable but manipulative scholarship athlete who was being moved through by admin who did not come to 1/3 of classes (w admin approval) passed the final (hence passed the class) because, apparently, he was getting input even though he minimally participated and had absences. But I was not satisfied when popular smart girl who distracted in class got an A, and hard-working slow-processor never moved beyond Novice Low+. I need to up my classroom management (call popular girl’s parents) and find more engaging material (but two kids just would not engage).

    HOWEVER, it also appeals to me that ALL students take ownership/accountability of their 50%. I can lead the horse to water – even give it soda – but I can’t make them drink it. The more input they receive, the more they will acquire – both popular smart girl and hardworking kid. If I make hw be input (rereading), and they “log” minutes and the pages read at home and during SSR (they can fib, but I can have a chat and check on that), if they are NOT distracting during class, (I do not worry about eye contact or body position), then they are doing their part to receive input, and that can boost slow-processor’s grade – not his or her fault what their processing speed is – and distracting smart kids should be getting more input and should not earn an A.

    This is a BIG dilemma for me and one I would love to resolve. I am open to all comments-arguments. Help solving this puzzle would be appreciated.

    1. Well everybody is diff. Slow processor is…a slow processor. Variability in ability is standard esp during the extremely limited time we have with kids. If they all do the same test, their results will show how much they acquired.

      1. I guess I unofficially do. I think that by Spanish 1 about 2/3 of my kids are at Novice High and 1/3 at Novice Mid. my 2s should all be at Int Low or Mid.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s