If anyone has any questions about the data, feel free to ask here.
", "messages": [ { "id": "2229732", "postDate": "04/21/2023 16:30:36", "content": "If anyone has any questions about the data, feel free to ask here.
", "rawMarkdown": "If anyone has any questions about the data, feel free to ask here.", "votes": null }, { "id": "2235244", "postDate": "04/25/2023 21:19:20", "content": "Hi, I had a few questions on some parts. I could not find answers from the descriptions.
\nYes, there are dozens of units that a student could be given an end-of-unit test for.
It depends on the teacher, but to know how many unit tests a student was given in the training data, you could merge the assignment details table with the assignment relationships \"unit_test_assignment_log_id\" column and count how many unique assignment log IDs per student.
Nope.
Thanks for clarifying! I had another question:
\nYes. For some units there are two different tests, and even if there is only one, there's nothing stopping a teacher from assigning it multiple times.
", "rawMarkdown": "Yes. For some units there are two different tests, and even if there is only one, there's nothing stopping a teacher from assigning it multiple times.", "votes": null }, { "id": "2238806", "postDate": "04/28/2023 20:11:41", "content": "Hi Ethan, some problem_id in test data (evaluation_unit_test_scores.csv) I couldn't find in other files (problem_details.csv, sequence_details.csv), Example of row in test data:
\n| id | \nassignment_log_id | \nproblem_id | \n
|---|---|---|
| 17672 | \n19VJY40T5T | \n29NEJMLNEH | \n
problem_details.csv
\nThis file contains one row for every problem referenced in the dataset, except for some problems in the action logs, which have been deleted from the database. These problems likely had errors during their original transcription into ASSISTments that were corrected, but no record of the original problems was kept.
Hi, I had a question about the submission format. What should be the metric for predicted outcomes for submission: probability, or binary?
\nThanks!
Binary. \"Score\" of each problem has to be either 1 or 0.
", "rawMarkdown": "Binary. \"Score\" of each problem has to be either 1 or 0.", "votes": null }, { "id": "2251828", "postDate": "05/09/2023 16:39:27", "content": "Each problem does only have a binary 1 or 0 score, but the metric used to evaluate the predictions is AUC, so you're probably better off submitting probability(score = 1) as your submissions.
", "rawMarkdown": "Each problem does only have a binary 1 or 0 score, but the metric used to evaluate the predictions is AUC, so you're probably better off submitting probability(score = 1) as your submissions.", "votes": null }, { "id": "2273831", "postDate": "05/25/2023 12:45:00", "content": "Hi, can I ask for the paper submission, what is the word limitation?
", "rawMarkdown": "Hi, can I ask for the paper submission, what is the word limitation?", "votes": null }, { "id": "2275096", "postDate": "05/26/2023 13:52:32", "content": "There's no word limit. Make it as long or short as necessary. That being said, if the editors feel it is lacking detail, they could request you add more detail, and if they feel it's redundant in places, they could request you shorten it.
", "rawMarkdown": "There's no word limit. Make it as long or short as necessary. That being said, if the editors feel it is lacking detail, they could request you add more detail, and if they feel it's redundant in places, they could request you shorten it.", "votes": null } ], "comments": [ { "id": 2235244, "author_name": "stdevlimit", "author_url": "", "post_date": "04/25/2023 21:19:20", "content": "Hi, I had a few questions on some parts. I could not find answers from the descriptions.
\nYes, there are dozens of units that a student could be given an end-of-unit test for.
It depends on the teacher, but to know how many unit tests a student was given in the training data, you could merge the assignment details table with the assignment relationships \"unit_test_assignment_log_id\" column and count how many unique assignment log IDs per student.
Nope.
Thanks for clarifying! I had another question:
\nYes. For some units there are two different tests, and even if there is only one, there's nothing stopping a teacher from assigning it multiple times.
", "votes": null, "replies": [] } ] }, { "id": 2238806, "author_name": "ceruttivini", "author_url": "", "post_date": "04/28/2023 20:11:41", "content": "Hi Ethan, some problem_id in test data (evaluation_unit_test_scores.csv) I couldn't find in other files (problem_details.csv, sequence_details.csv), Example of row in test data:
\n| id | \nassignment_log_id | \nproblem_id | \n
|---|---|---|
| 17672 | \n19VJY40T5T | \n29NEJMLNEH | \n
problem_details.csv
\nThis file contains one row for every problem referenced in the dataset, except for some problems in the action logs, which have been deleted from the database. These problems likely had errors during their original transcription into ASSISTments that were corrected, but no record of the original problems was kept.
Hi, I had a question about the submission format. What should be the metric for predicted outcomes for submission: probability, or binary?
\nThanks!
Binary. \"Score\" of each problem has to be either 1 or 0.
", "votes": null, "replies": [ { "id": 2251828, "author_name": "ethanprihar", "author_url": "", "post_date": "05/09/2023 16:39:27", "content": "Each problem does only have a binary 1 or 0 score, but the metric used to evaluate the predictions is AUC, so you're probably better off submitting probability(score = 1) as your submissions.
", "votes": null, "replies": [] } ] } ] }, { "id": 2273831, "author_name": "newqbsfisher", "author_url": "", "post_date": "05/25/2023 12:45:00", "content": "Hi, can I ask for the paper submission, what is the word limitation?
", "votes": null, "replies": [ { "id": 2275096, "author_name": "ethanprihar", "author_url": "", "post_date": "05/26/2023 13:52:32", "content": "There's no word limit. Make it as long or short as necessary. That being said, if the editors feel it is lacking detail, they could request you add more detail, and if they feel it's redundant in places, they could request you shorten it.
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "2229732": "If anyone has any questions about the data, feel free to ask here.", "2235244": "Hi, I had a few questions on some parts. I could not find answers from the descriptions.\n\n1. Can a student receive more than two end-of-unit assignments?\n2. How many units does a student go through?\n3. Is there a language restriction? (Python or something else)", "2235855": "1. Yes, there are dozens of units that a student could be given an end-of-unit test for.\n\n2. It depends on the teacher, but to know how many unit tests a student was given in the training data, you could merge the assignment details table with the assignment relationships \"unit_test_assignment_log_id\" column and count how many unique assignment log IDs per student.\n\n3. Nope.", "2236381": "Thanks for clarifying! I had another question:\n1. Can a student receive more than two end-of-unit assignments in a single unit?", "2236390": "Yes. For some units there are two different tests, and even if there is only one, there's nothing stopping a teacher from assigning it multiple times.", "2238806": "Hi Ethan, some problem_id in test data (evaluation_unit_test_scores.csv) I couldn't find in other files (problem_details.csv, sequence_details.csv), Example of row in test data:\n| id | assignment_log_id |problem_id|\n| --- | --- | --- |\n| 17672 | 19VJY40T5T |29NEJMLNEH|", "2240302": "problem_details.csv\n\nThis file contains one row for every problem referenced in the dataset, except for some problems in the action logs, which have been deleted from the database. These problems likely had errors during their original transcription into ASSISTments that were corrected, but no record of the original problems was kept.", "2251105": "Hi, I had a question about the submission format. What should be the metric for predicted outcomes for submission: probability, or binary?\nThanks!", "2251814": "Binary. \"Score\" of each problem has to be either 1 or 0.", "2251828": "Each problem does only have a binary 1 or 0 score, but the metric used to evaluate the predictions is AUC, so you're probably better off submitting probability(score = 1) as your submissions.", "2273831": "Hi, can I ask for the paper submission, what is the word limitation?", "2275096": "There's no word limit. Make it as long or short as necessary. That being said, if the editors feel it is lacking detail, they could request you add more detail, and if they feel it's redundant in places, they could request you shorten it." }, "source": "meta" }