**Numerical Simulation** and **Symbolic Math** only

# Introduction

In the Response Outcome, you can provide response-specific feedback. Whether a candidate got the answer correct, incorrect, or maybe made a common error you factored into your Program Model, you can tailor the automatic feedback accordingly.

This feedback can also include the variable values themselves. So rather than the candidate receiving variable names in the feedback, they'll see the exact values they were presented when they attempted the question.

- Example Question
- Set Feedback For Correct Answer
- Set Feedback For Partially Correct Answer
- Set Feedback for No Response Outcomes Met
- Set Feedback for Set Incorrect Answers
- Enabling Feedback On A Test

# Example Question

To help demonstrate Response-specific feedback, we'll use the following question.

## Solution

To get the correct the response for Part 1 we must first take distance and divide by the combined speed of the trains.

To get the correct the response for Part 2, we need to calculate how far Train B has travelled at the point of crossing (Train B speed x Part 1 answer) and then subtract that distance from the overall distance between the two towns.

## Program Model

`trainA: rand_with_step(30,100,1);`

trainB: trainA + rand_with_step(5,35,1);

distance: rand_with_step(120,420,1);

combined_speed: trainA + trainB;

time_answer: decimalplaces(distance/combined_speed,2);

trainb_distance: time_answer * trainB;

distance_answer: distance - trainb_distance;

# Set Feedback For Correct Answer

In the **Response Outcome** under where you set the correct answer is the Feedback area.

When you start writing in the Feedback area, an editor will appear allowing for basic text decoration, inserting LaTeX, as well as inserting the variables declare in the Program Model. The value of these variables will then appear in the response feedback.

# Set Feedback For Partially Correct Answer

If using Error Carry Forward, you can include ECF-specific feedback. That way, you can highlight what part of the question they handled correctly while also highlighting where they went wrong.

ECF-specific feedback is entered under the **Error carry forward outcome **response outcome. By inserting the relevant `RESPONSE-n`

variable, the feedback will include the candidate's given answer.

# Set Feedback for No Response Outcomes Met

If a candidate's response does not match any of the response outcomes, you can give feedback explaining how the correct outcome is achieved. This feedback is entered under **Define feedback if no response outcomes are met**. Inserting the variables in the feedback will help contextualise it too.

# Set Feedback for Set Incorrect Answers

Response Feedback isn't just for mark-bearing or 'none of the above' responses. If there are common mistakes you expect to come up, you can pre-empt them and include feedback accordingly.

For example, with Part 2 of the train question, the answer is achieved in two steps

- Take Part 1's answer (time to cross) and multiplying by Train B's speed. (result =
`trainb_distance`

) - Subtract this value from the overall distance (
`distance`

).

This is the part of the Program Model that relates to these calculations:

`trainb_distance: time_answer * trainB;`

distance_answer: distance - trainb_distance;

If a common error is candidates forget to do the second step and answer with `trainb_distance`

, you could set-up feedback in advance.

Under **Response Outcome** click **+ Add Outcome** and the expected response. In this case it's `trainb_distance`

and the result is **Incorrect**. Now if a candidate makes this common error, they will receive specific feedback explaining where they went wrong.

This can be done with any variable you set in the Progam Model. It won't be possible to preempt every candidate error, but for common errors it can be a helpful way to show the candidate where they went off track.

# Enabling Feedback On A Test

As will all question-level feedback in Inspera Assessment, it needs to be enabled in **Deliver** in **After test**. It is its own setting **Include feedback on response outcomes.**