Saturday 26 September 2015

Evaluation of the reading test

Oh well... Where should I start...

After yesterday's reading test, I went home with lots of material and I didn't really have a clue, how I was supposed to give them a grade. In my eyes they all did wonderful, but that won't do me any good for they report cards so I had to come up with some kind of key...

First, I scanned the stories as texts so I was able copy them into a word document and count the words each student read because I wanted to make it fair for those who had longer texts to read.

I thought I was so smart because now I started to highlight words they read wrong and then I just calculated the correct words out of all the words and had a percentage of how much they've read correctly. Sounds great, doesn't it? Well this became harder and harder to judge the more texts I listened to.
  • What do I do with kids who pronounced the words correctly but had to pause every other sentence? 
  • What do I do with kids who read words that were not in the text but skipped words that were there? 
  • And what do I do with those who pronounced the words correctly but due to their lack of practising couldn't read it fluently?

I started to realise that this whole thing was a lot more complex. And here is what I came up with:
I made three criteria: Pronunciation, Accuracy and Fluency

Pronunciation
  • if someone pronounced the "came" as "come" or the "new" as "now".
  • And I did NOT care if they said "wery" instead of "very" or if the "th" sound didn't come out correctly. Neither did I expect them to say "earth" 100% correctly (because that a real tough word), as long as I understood that they meant "earth", it was fine with me. Just one girl said "Youth" and that was a little too far away in my opinion.
  • So if one child's text was 271 words long and it made 5 errors, it got 266/271 or 98.15%
Accuracy
  • leaving out a word or mixing up words or just reading "he" instead of "she".
  • So if one child's text was 271 words long and it made 5 errors, it got 266/271 or 98.15%
Fluency
  • was if they read the whole sentence fluently. 
  • So if one child's text was 15 sentences long and it stumbled in two sentences it got 13/15 or 86.67%
So I started over listening to those readings... But after the third text I realised that if the kids or the parents wanted to know where those mistakes happened, I needed to be able to show them and so I started highlighting the words in the text in different colours.

Soon I stopped doing this on the computer but printed it out and just wrote the letters P, A or F in the text when ever there was a mistake. So here's the result:

Evaluation Reading Test

Since they did not always have the same amount of text in the first and in the second round, it looks a little confusing but this way, I can show everybody who wants to know how I came up with the grades.

And here are the calculations... I hope this one boy will be back on Monday so I can record him and his group too. Then I'll make a Audio CD out of all the students reading with their grades on it. So they have the book and the CD to take home to their parents.


2 comments:

  1. Haha! I am having a good laugh at this because this is the struggle I get my students to think about and they all look at me like I'm stupid!! My own example are the book reviews:
    How do I "score" reading: different kids had different book and during the oral presentations about the books, I wanted to see if they showed an indication of reading the book in-depth. So I just gave 3 levels. "seems like you did not read the book", "yes, you read the book but not a lot of details were apparent" and "yes, great, I see you read and really understood the book". Of course none of this actually proves anything because perhaps the lowest scorers really did read the book in depth, but their OUTPUT did not indicate that. And then the oral presentations.....Some read from their papers, but did it well; others spoke freely and goofed quite a bit - so here I also did the "fluency, accuracy, range" basics.

    All this leads me to believe that for reflective teachers such as yourself, the 4 skills are not really the way to go. Rather just have 2 - productive and receptive or comprehensibility and comprehension and maybe by simplifying the system, it will get other teachers to get away from simple tallies of vocab/translation tests into something more holistic!!!

    ReplyDelete
    Replies
    1. Hey Laura
      How right you are. I could have saved soooooo much time by thinking of how I will grade the readings before the project. AND I could have told the class to make it more transparent to them. Well, I'll tell them now as they get their grades how I did judged them and I have a "new aim's resolution"! From now on I'll decide on how I will grade them while I plan the project. And I could have made a reading comprehension grade on this too... Or a writing... Ahhh...! I thought of letting them write a couple of sentences about their story, or giving them a couple of questions about their texts to get grades for two more sections of the report card but I decided not to since the aim was clearly given: Reading a story to the class... I'll have to think of wider aims on the next project or I'm never gonna fill this report card. Well while going along the Young World, there'll definitely be listening, reading and writing but next time I do a project myself, I'll have to have those grades in mind :-)

      Delete