Teacher and para handoff kit
One page per goal that lives with the clipboard, not in a shared drive. It answers the questions a person actually has at 9:15 on a Tuesday when the direction has just been given twice.
What this fixes
Consistency between collectors is not a training problem, it is a memory problem. Two people agreed in September; it is now February; one of them has six other goals and the other started in November. Nobody is being careless. The definition just is not in front of them.
This is not observer calibration. Running a real interobserver agreement check means two people scoring the same live session independently and comparing, repeatedly. This sheet makes that check more likely to pass. It does not replace it, and we are not going to call it one.
The sheet
Fill it from your workshop output. Print one per goal, per collector.
Collection handoff — one goal
| What we are measuring | |
| One opportunity is | |
| It does not count when | |
| Met criterion means | |
| Prompted success counts as | |
| Unit we report | |
| When and where to collect | |
| Observation window | |
| How many opportunities per session | |
| Who collects | |
| Where the sheet goes afterwards | |
| Who reviews it, and how often | |
| Who to ask when unsure |
- I
- V
- G
- P
Codes describe the support given. They are not a scale and they do not add up.
The three rules on the back
Print these on the reverse. They cover most of what goes wrong.
| If this happens | Do this |
|---|---|
| No opportunities came up | Leave the row blank and mark it not observed. Do not write 0. |
| Opportunities came up and none were met | Write 0. It is a real observation and it belongs in the total. |
| You scored something wrong | Strike it through, write the correction beside it, initial it. Do not erase, and do not add a second row. |
| You are not sure it counted | Write what happened in the notes column and ask. A note is recoverable; a guess written as a score is not. |
| Somebody else covered the session | Put their initials on the row. Who collected is part of the data. |
A ten-minute check, once a month
Both collectors score the same session independently and compare before anybody says a number out loud. Disagreements are useful information about the definition, not about the person.
| Date | Your count | Their count | Where you differed | Change made to the definition |
|---|---|---|---|---|
If you change the definition as a result, that is a protocol change and it goes in the revision log. Data collected before and after are not automatically comparable.