{
  "id": 195075,
  "title": "Meaning of prior_q_had_exp",
  "url": "/competitions/riiid-test-answer-prediction/discussion/195075",
  "author_name": "",
  "post_date": "2020-11-03T13:19:32.503987Z",
  "votes": 4,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Does this attribute mean that the user 'chose' to look at an explanation/correct response or does<br>\nit mean that this question 'has' an explanation/correct response and is automatically displayed?</p>\n<p>If latter, then do certain bundles/topics of questions have an explanation/correct response and other do not?</p>\n<p>I think this is an important question since it means that this attribute either reveals something about the user's recent behaviour or something about the future knowledge of the user.</p>",
  "messages": [
    {
      "id": "1068488",
      "postDate": "11/03/2020 13:19:32",
      "content": "<p>Does this attribute mean that the user 'chose' to look at an explanation/correct response or does<br>\nit mean that this question 'has' an explanation/correct response and is automatically displayed?</p>\n<p>If latter, then do certain bundles/topics of questions have an explanation/correct response and other do not?</p>\n<p>I think this is an important question since it means that this attribute either reveals something about the user's recent behaviour or something about the future knowledge of the user.</p>",
      "rawMarkdown": "Does this attribute mean that the user 'chose' to look at an explanation/correct response or does\nit mean that this question 'has' an explanation/correct response and is automatically displayed?\n\nIf latter, then do certain bundles/topics of questions have an explanation/correct response and other do not?\n\nI think this is an important question since it means that this attribute either reveals something about the user's recent behaviour or something about the future knowledge of the user.",
      "votes": null
    },
    {
      "id": "1069292",
      "postDate": "11/04/2020 09:37:55",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/srkopp\" target=\"_blank\">@srkopp</a>, the first 30 or so questions are part of an <em>onboarding diagnostic test</em> (check: data description) and as a result the users do not see any explanations for any of them. Beyond the first 30 questions, I believe that <strong>most of the questions have explanations to them and are shown by default, however the user may chose not to see their explanations sometimes</strong> (the vast majority of questions sometimes answered and sometimes not answered, confirms this).</p>\n<p>I say that explanations are shown <em>by default</em> because of this statement, I saw from the <a href=\"https://github.com/riiid/ednet\" target=\"_blank\">RIIID Ednet</a> Github Repo (check data description for KT3):</p>\n<pre><code>After each student solves given questions, corresponding explanations are given to him.\n</code></pre>\n<p>I haven't done analysis on the entire dataset, but for the first 1e6 rows, certain questions (35 to be precise) were never explained. I double checked these question's task_container_id to see if these were the questions that were asked for <em>diagnostics</em> but no, they do not overlap. I have no idea why these were never explained. Maybe they were too simple to be explained, maybe? Or Maybe.. I need to check with the entire dataset ;)</p>\n<p>I might be wrong about this, but this is just how I have inferred it. Hope this helps!</p>",
      "rawMarkdown": "Hi @srkopp, the first 30 or so questions are part of an *onboarding diagnostic test* (check: data description) and as a result the users do not see any explanations for any of them. Beyond the first 30 questions, I believe that **most of the questions have explanations to them and are shown by default, however the user may chose not to see their explanations sometimes** (the vast majority of questions sometimes answered and sometimes not answered, confirms this).\n\nI say that explanations are shown *by default* because of this statement, I saw from the [RIIID Ednet](https://github.com/riiid/ednet) Github Repo (check data description for KT3):\n```\nAfter each student solves given questions, corresponding explanations are given to him.\n```\n\nI haven't done analysis on the entire dataset, but for the first 1e6 rows, certain questions (35 to be precise) were never explained. I double checked these question's task_container_id to see if these were the questions that were asked for *diagnostics* but no, they do not overlap. I have no idea why these were never explained. Maybe they were too simple to be explained, maybe? Or Maybe.. I need to check with the entire dataset ;)\n\nI might be wrong about this, but this is just how I have inferred it. Hope this helps!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1069292,
      "author_name": "doctorkael",
      "author_url": "",
      "post_date": "11/04/2020 09:37:55",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/srkopp\" target=\"_blank\">@srkopp</a>, the first 30 or so questions are part of an <em>onboarding diagnostic test</em> (check: data description) and as a result the users do not see any explanations for any of them. Beyond the first 30 questions, I believe that <strong>most of the questions have explanations to them and are shown by default, however the user may chose not to see their explanations sometimes</strong> (the vast majority of questions sometimes answered and sometimes not answered, confirms this).</p>\n<p>I say that explanations are shown <em>by default</em> because of this statement, I saw from the <a href=\"https://github.com/riiid/ednet\" target=\"_blank\">RIIID Ednet</a> Github Repo (check data description for KT3):</p>\n<pre><code>After each student solves given questions, corresponding explanations are given to him.\n</code></pre>\n<p>I haven't done analysis on the entire dataset, but for the first 1e6 rows, certain questions (35 to be precise) were never explained. I double checked these question's task_container_id to see if these were the questions that were asked for <em>diagnostics</em> but no, they do not overlap. I have no idea why these were never explained. Maybe they were too simple to be explained, maybe? Or Maybe.. I need to check with the entire dataset ;)</p>\n<p>I might be wrong about this, but this is just how I have inferred it. Hope this helps!</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1068488": "Does this attribute mean that the user 'chose' to look at an explanation/correct response or does\nit mean that this question 'has' an explanation/correct response and is automatically displayed?\n\nIf latter, then do certain bundles/topics of questions have an explanation/correct response and other do not?\n\nI think this is an important question since it means that this attribute either reveals something about the user's recent behaviour or something about the future knowledge of the user.",
    "1069292": "Hi @srkopp, the first 30 or so questions are part of an *onboarding diagnostic test* (check: data description) and as a result the users do not see any explanations for any of them. Beyond the first 30 questions, I believe that **most of the questions have explanations to them and are shown by default, however the user may chose not to see their explanations sometimes** (the vast majority of questions sometimes answered and sometimes not answered, confirms this).\n\nI say that explanations are shown *by default* because of this statement, I saw from the [RIIID Ednet](https://github.com/riiid/ednet) Github Repo (check data description for KT3):\n```\nAfter each student solves given questions, corresponding explanations are given to him.\n```\n\nI haven't done analysis on the entire dataset, but for the first 1e6 rows, certain questions (35 to be precise) were never explained. I double checked these question's task_container_id to see if these were the questions that were asked for *diagnostics* but no, they do not overlap. I have no idea why these were never explained. Maybe they were too simple to be explained, maybe? Or Maybe.. I need to check with the entire dataset ;)\n\nI might be wrong about this, but this is just how I have inferred it. Hope this helps!"
  },
  "source": "meta"
}