Nested timeline with feedback

[y-j-huang]

I was trying to create a nested structure in the timeline so that the questions within the same item group won't be separated and the presentation of different items will be randomized (something like this: A1, A2, A3, will always appear together, but A, B, C, could appear in any order -- A, B, C, each has three questions A1, A2, A3, B1, B2, B3, C1, C2, C3). I also want to give feedback in my practice after each question. The plugin I'm using is survey-multi-choice. Because multi-choice has a nested structure, the code I used ended up to be quite complicated (omitting the feedback part which is irrelevant to the problem I ran into).

 timeline:[ 
      {   type: 'survey-multi-choice',

            questions: jsPsych.timelineVariable('questions1'),
            data:{
                trial_part: 'test',
                questions: jsPsych.timelineVariable('questions1'),
                correct_response: jsPsych.timelineVariable('correct_response'),
                },
            on_finish: function(data) {
                var acc = false;
                if (data.correct_response == JSON.stringify(data.response)) {
                    acc = true;
                }
            data.accuracy = acc;
           //var prompt = data.questions[0].prompt // sentence and question
           //data.prompt = prompt;
            }
        },

        {   type: 'survey-multi-choice',
            questions: jsPsych.timelineVariable('questions2'),
            data:{
                trial_part: 'test',
                questions: jsPsych.timelineVariable('questions2'),
                correct_response: jsPsych.timelineVariable('correct_response'),
                },
            on_finish: function(data) {
                var acc = false;
                if (data.correct_response == JSON.stringify(data.response)) {
                    acc = true;
            }
            data.accuracy = acc;
           //var prompt = data.questions[0].prompt // sentence and question
           //data.prompt = prompt;
                }
        },
        
        {   type: 'survey-multi-choice',
            questions: jsPsych.timelineVariable('questions3'),
            data:{
                trial_part: 'test',
                questions: jsPsych.timelineVariable('questions3'),
                correct_response: jsPsych.timelineVariable('correct_response'),
                },
            on_finish: function(data) {
                var acc = false;
                if (data.correct_response == JSON.stringify(data.response)) {
                    acc = true;
            }
            data.accuracy = acc;
           //var prompt = data.questions[0].prompt // sentence and question
           //data.prompt = prompt;
                }
        },
     ]

Where the stim info looks like this:
[{
questions1: [{prompt:"XXX", options:["A", "B", "V"], required: true, name: "X"}], correct_response: "A",
questions2: [{prompt:"XXX", options:["A", "B", "V"], required: true, name: "X"}], correct_response: "B",
questions3: [{prompt:"XXX", options:["A", "B", "V"], required: true, name: "X"}], correct_response: "V", },
{
questions1: [{prompt:"XXX", options:["A", "B", "V"], required: true, name: "X"}], correct_response: "A",
questions2: [{prompt:"XXX", options:["A", "B", "V"], required: true, name: "X"}], correct_response: "B",
questions3: [{prompt:"XXX", options:["A", "B", "V"], required: true, name: "X"}], correct_response: "V", },
{
questions1: [{prompt:"XXX", options:["A", "B", "V"], required: true, name: "X"}], correct_response: "A",
questions2: [{prompt:"XXX", options:["A", "B", "V"], required: true, name: "X"}], correct_response: "B",
questions3: [{prompt:"XXX", options:["A", "B", "V"], required: true, name: "X"}], correct_response: "V", },
]

With this structure, the correct_response variable only pulls out the last correct_response from each item, namely "V" in the above example. I was wondering how I should modify the structure so that the correct line could be identified. Thank you!