sem09_hypothesis/sem_09_hypo_184.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Наполеон задумался о результатах контрольной 1 по ноду :)\n",
    "\n",
    "\n",
    "<img src=\"https://upload.wikimedia.org/wikipedia/commons/5/53/Napoleon_Paul_Delaroche.jpg\" height=\"500\" width=\"400\"> \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "from scipy import stats # more style :)\n",
    "\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.style.use('ggplot')  # стиль для графиков\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Задаченька 1 или Французы клюют только на французских покойников\n",
    "\n",
    "Девятый семинар по наукам о данных мы начинаем на неделе с 18 мая. В этот день корсиканец Наполеон был провозглашён императором Франции. А нам настала пора загрузить файлик с оценками по ноду, `nod_2020-05-17.csv`.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>group</th>\n",
       "      <th>fio</th>\n",
       "      <th>k1</th>\n",
       "      <th>d1fraud</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>БЭК181</td>\n",
       "      <td>Аверьянов Николай Егорович</td>\n",
       "      <td>10.1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>БЭК181</td>\n",
       "      <td>Алиев Эльнур Гашимович</td>\n",
       "      <td>6.7</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>БЭК181</td>\n",
       "      <td>Алтунина Анастасия Игоревна</td>\n",
       "      <td>8.2</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>БЭК181</td>\n",
       "      <td>Байкалова Дарья Александровна</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>БЭК181</td>\n",
       "      <td>Васьков Алексей Васильевич</td>\n",
       "      <td>7.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    group                            fio    k1  d1fraud\n",
       "0  БЭК181     Аверьянов Николай Егорович  10.1      0.0\n",
       "1  БЭК181         Алиев Эльнур Гашимович   6.7      1.0\n",
       "2  БЭК181    Алтунина Анастасия Игоревна   8.2      0.0\n",
       "3  БЭК181  Байкалова Дарья Александровна   7.0      1.0\n",
       "4  БЭК181     Васьков Алексей Васильевич   7.0      2.0"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nod = pd.read_csv('nod_2020-05-17.csv')\n",
    "nod[[\"group\", \"fio\", \"k1\", \"d1fraud\"]].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "nod.dropna(subset=[\"k1\"], inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5.431282051282051"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nod.k1.values.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5.431282051282051"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nod.k1.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "class A:\n",
    "    def __init__(self,):\n",
    "        print(1)\n",
    "        \n",
    "class B(A):\n",
    "    def __init__(self,):\n",
    "        super(A, self).__init__()\n",
    "        print(2)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1\n",
      "2\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<__main__.B at 0x7efd7f673710>"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "B()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__а)__ Постройте гистограмму результатов за первую контрольную, переменная `k1` :)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "norm_distribution = stats.norm(loc=nod.k1.mean(), scale=nod.k1.std())\n",
    "\n",
    "x = np.linspace(0, 13, 300)\n",
    "y = norm_distribution.pdf(x)\n",
    "\n",
    "plt.figure()\n",
    "plt.plot(x, y)\n",
    "nod.k1.hist(bins=20, density=True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KstestResult(statistic=0.052264904482032776, pvalue=0.22936091387357344)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats.kstest(nod.k1, 'norm', args=(nod.k1.mean(), nod.k1.std()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Предположим невероятное! А именно, независимость и одинаковую распределённость наших наблюдений!\n",
    "\n",
    "__б)__ Постройте примерный $95\\%$-й доверительный интервал для ожидамоего результата за первую контрольную.\n",
    "\n",
    "$$\n",
    "\\left[\\bar x - z_{crit} \\sqrt{\\frac{\\hat\\sigma^2}{n}} ; \\bar x + z_{crit} \\sqrt{\\frac{\\hat\\sigma^2}{n}} \\right]\n",
    "$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.959963984540054"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats.norm().ppf(0.975)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-1.9599639845400545"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats.norm().ppf(0.025)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Деврительный интервал 5.115697343992563 5.746866758571539\n"
     ]
    }
   ],
   "source": [
    "alpha = 0.05\n",
    "\n",
    "n = nod.k1.size\n",
    "mu = nod.k1.mean()\n",
    "sd = np.sqrt(nod.k1.var() / n)\n",
    "\n",
    "z_crit = stats.t(n-1).ppf(1 - alpha / 2)\n",
    "\n",
    "print(\"Деврительный интервал\", mu - z_crit * sd,  mu + z_crit * sd)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Разобьём наши наблюдения на две группы: Ангелы и Демоны. У Ангелов `d1fraud == 0`, у Демонов `d1fraud > 0`.\n",
    "\n",
    "__в)__ Постройте на общем фоне гистограмму распределения оценок за первую контрольную у Ангелов и Демонов. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10, 5))\n",
    "nod[nod.d1fraud > 2.5].k1.hist(bins=20, alpha=0.5, density=True, label=\"demons\")\n",
    "nod[nod.d1fraud < 2.5].k1.hist(bins=20,alpha=0.5, density=True, label=\"angels\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__г)__ Постройте примерный $95\\%$-й доверительный интервал для разницы ожидаемого результата за первую контрольную у Ангелов и Демонов. Без предположения о равенстве дисперсий.\n",
    "\n",
    "$$\n",
    "\\left[\\bar x - \\bar y - z_{crit} \\sqrt{\\frac{\\hat\\sigma_x^2}{n_x} + \\frac{\\hat\\sigma_y^2}{n_y}} ; \\bar x - \\bar y + z_{crit} \\sqrt{\\frac{\\hat\\sigma_x^2}{n_x} + \\frac{\\hat\\sigma_y^2}{n_y}} \\right]\n",
    "$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = nod[nod.d1fraud > 0].k1\n",
    "y = nod[nod.d1fraud == 0].k1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Деврительный интервал -0.532833905384929 0.8087678503390602\n"
     ]
    }
   ],
   "source": [
    "alpha = 0.05\n",
    "mean_diff = x.mean() - y.mean()\n",
    "diff_sd = np.sqrt(x.var()/x.size + y.var()/y.size)\n",
    "\n",
    "z_crit = stats.norm.ppf(1 - alpha / 2)\n",
    "\n",
    "print(\"Деврительный интервал\", mean_diff - z_crit * diff_sd,  mean_diff + z_crit * diff_sd)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__д)__ Как изменятся результаты, если границу между Ангелами и Демонами провести по `d1fraud < 2.5` и `d1fraud > 2.5`?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Деврительный интервал -0.8893352912570267 0.53970010884823\n"
     ]
    }
   ],
   "source": [
    "x = nod[nod.d1fraud > 2.5].k1\n",
    "y = nod[nod.d1fraud < 2.5].k1\n",
    "\n",
    "lpha = 0.05\n",
    "mean_diff = x.mean() - y.mean()\n",
    "diff_sd = np.sqrt(x.var()/x.size + y.var()/y.size)\n",
    "\n",
    "z_crit = stats.norm.ppf(1 - alpha / 2)\n",
    "\n",
    "print(\"Деврительный интервал\", mean_diff - z_crit * diff_sd,  mean_diff + z_crit * diff_sd)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Предположим, что математическое ожидание результата за контрольную зависит от уровня подготовки, а дисперсия результата — от нервов.\n",
    "\n",
    "__е)__ Проверьте гипотезу об одинаковой нервозности Ангелов и Демонов на уровне значимости $5\\%$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F observ 0.8220777811901974\n",
      "F right crit 1.424729837579135\n",
      "F left crit 0.6715534278792631\n"
     ]
    }
   ],
   "source": [
    "x = nod[nod.d1fraud > 2.5].k1\n",
    "y = nod[nod.d1fraud < 2.5].k1\n",
    "\n",
    "print(\"F observ\", x.var() / y.var())\n",
    "\n",
    "print(\"F right crit\", stats.f.ppf(1 - alpha / 2, x.size - 1, y.size - 1,))\n",
    "print(\"F left crit\", stats.f.ppf(alpha / 2, x.size - 1, y.size - 1,))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__ё)__ Без буквы `ё` никуда, скажут нам [Пафнутий Чебышёв](https://ru.wikipedia.org/wiki/%D0%A7%D0%B5%D0%B1%D1%8B%D1%88%D1%91%D0%B2,_%D0%9F%D0%B0%D1%84%D0%BD%D1%83%D1%82%D0%B8%D0%B9_%D0%9B%D1%8C%D0%B2%D0%BE%D0%B2%D0%B8%D1%87) и [Лёв Толстой](https://arzamas.academy/special/ruslit/writers/tolstoy).  \n",
    "\n",
    "Нарисуйте симпатичную картинку для корреляционной матрицы по задачам первой контрольной для всех наблюдений, отдельно для Ангелов и Демонов. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# решение"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Симуляционная задачулька 2 про повторное тестирование\n",
    "\n",
    "Сказка про то, что последовательное применение тестов без корректировки — это не очень хорошо :)\n",
    "\n",
    "Жизнь исследователя Винни-Пуха прекрасна. Наблюдений много, гипотеза о равенстве ожиданий верна, а наблюдения — нормальны. \n",
    "\n",
    "А именно, величины $X_1$, ..., $X_{n}$ независимы и нормальны $N(0;1)$, величины $Y_1$, ..., $Y_n$ независимы и нормальны $N(0;2)$. И пусть $n_x=100$ и $n_y=200$.\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Винни-Пух правда этого ничего не знает, потому что у него в голове опилки, и тестирует сначала гипотезу о равенстве дисперсий двух выборок на уровне значимости $5\\%$.\n",
    "\n",
    "__а)__ Проведите $10^6$ экспериментов, то есть $10^6$ раз попробуйте за Винни-Пуха проверить гипотезу. Хорошо бы $10^8$, но оттестируйте сначала свой код на $10^6$. \n",
    "\n",
    "Постройте гистограмму распределения тестовой статистики о равенстве дисперсий. В скольки процентах случаев гипотеза о равенстве дисперсий отвергнута? "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "65473bbecfc64931b1e5f4ee439b72f6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=10000), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.6/dist-packages/scipy/stats/_continuous_distns.py:1680: RuntimeWarning: divide by zero encountered in log\n",
      "  lPx = m/2 * np.log(m) + n/2 * np.log(n) + (n/2 - 1) * np.log(x)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.0217"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import tqdm\n",
    "nx = 100\n",
    "ny = 200\n",
    "\n",
    "simulation_count = int(1e4)\n",
    "alpha = 0.05\n",
    "\n",
    "f_distrubution = stats.f(nx - 1, ny - 1)\n",
    "left_border = f_distrubution.ppf(alpha / 2)\n",
    "right_border = f_distrubution.ppf(1 - alpha / 2)\n",
    "\n",
    "statistics = []\n",
    "some_list = []\n",
    "\n",
    "for _ in tqdm.tqdm_notebook(range(simulation_count)):\n",
    "    x = np.random.normal(0, 1, nx)\n",
    "    y = np.random.normal(0, np.sqrt(2), ny)\n",
    "\n",
    "    statistic = x.var() / y.var()\n",
    "    statistics.append(statistic)\n",
    "    \n",
    "    if left_border < statistic < right_border:\n",
    "        some_list.append((x, y))\n",
    "\n",
    "x = np.linspace(0, 2, 1000)\n",
    "y = f_distrubution.pdf(x)\n",
    "    \n",
    "plt.figure()\n",
    "plt.hist(statistics, bins=20, density=True)\n",
    "plt.plot(x, y)\n",
    "plt.show()\n",
    "\n",
    "len(some_list) * 1.0 / simulation_count"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.6/dist-packages/scipy/stats/_continuous_distns.py:1680: RuntimeWarning: divide by zero encountered in log\n",
      "  lPx = m/2 * np.log(m) + n/2 * np.log(n) + (n/2 - 1) * np.log(x)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "x = np.linspace(0, 2, 1000)\n",
    "y = f_distrubution.pdf(x)\n",
    "    \n",
    "plt.figure()\n",
    "plt.plot(x, y)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__б)__ Если гипотеза о равенстве дисперсий не отвергается, то Винни использует $t$-статистику для проверки гипотезы о равенстве ожиданий при равенстве дисперсий. \n",
    "\n",
    "Для тех экспериментов, в которых гипотеза о равенстве дисперсий не отверглась, проверьте гипотезу о равенстве ожиданий на уровне значимости $5\\%$. \n",
    "\n",
    "Постройте гистограмму распределения $t$-статистики. В скольки процентах случаев гипотеза о равенстве ожиданий отвергнута?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# решение"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__в)__ Словами объясните, почему нехорошо сначала тестировать на равенство дисперсий, а потом на равенство средних при равных дисперсиях, если на первом шаге гипотеза $H_0$ не отверглась. "
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    " # решение"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Задачулька 3 о мимимишных мышках\n",
    "\n",
    "[https://nplus1.ru/news/2018/05/11/mice-trolley](https://nplus1.ru/news/2018/05/11/mice-trolley) \n",
    "\n",
    "Для изучения аспектов процесса принятия моральных решений психологи уже много лет используют этические дилеммы, с помощью которых оценивают действия людей в гипотетических ситуациях. Один из самых известных примеров — это [проблема вагонетки,](https://ru.wikipedia.org/wiki/Проблема_вагонетки) в которой необходимо принять решение о том, стоит ли пожертвовать одним человеком для спасения пятерых.\n",
    "\n",
    "Бельгийские психологи воплотили дилемму в реальную жизнь. Участники эксперимента должны были выбрать, ударить током пять мышей или одну мышь. Эксперимент проходил следующим образом. Участника сажали перед двумя клетками, в одной из которых сидели пять мышей, а в другой — одна. Перед клетками стоял ноутбук с $20$-секундным таймером: участникам сообщили, что по истечении этих $20$ секунд в клетку с пятью мышами пустят ток, и мыши получат не смертельный, но очень болезненный удар. Пока время идет, участник может нажать на кнопку: в этом случае ток пустят по клетке с одной мышью. В исследовании использовали живых мышей; следует, однако, уточнить, что удары тока были ненастоящими: сразу же после «удара» участников сопроводили в отдельную комнату, где разъяснили им, что мыши в полном порядке и током их не били (об этом заранее догадались только $12$ участников). В решении реальной проблемы вагонетки приняли участие $192$ человека, а еще $83$ решали такую же задачку, но гипотетическую (на бумаге). Все участники также прошли онлайн-опросы, в ходе которых учёные собрали о респондентах кучу дополнительной информации. \n",
    "\n",
    "В файле `mouse.csv` лежит информация о том, как прошёл эксперимент. Нас будут интересовать столбцы: \n",
    "\n",
    "* __STUDY:__ какую проблему вагонетки решал человек $1$ если реальную, $2$ если на бумаге\n",
    "* __AGE:__ возраст респондента \n",
    "* __GENDER:__ пол респондента\n",
    "* __DECISION:__ решение дилеммы ($1$ - жать на кнопу, $0$ - не жать) \n",
    "* __RT:__ время, которое респондент потратил, чтобы нажать на кнопку \n",
    "\n",
    "Подробное описание данных, сами данные и даже код на R, использованный при оценивании моделей, можно найти в [репозитории исследования.](https://osf.io/kvb99/) В статье авторы строили кучу логистических регрессий, чтобы очистить эффект от психологических особенностей респондентов. Этому добру вас будут учить на эконометрике, мы пойдём простым и не совсем правильным путём и будем проверять гипотезы в лоб. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"mouse.csv\", sep=\"\\t\")\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Отдавайте себе отчёт о предпосылках теста, который вы используете.__\n",
    "\n",
    "\n",
    "__а)__ Попытаемся ответить на тот же вопрос, что и исследователи. В какой из двух ситуаций респонденты чащи принимают решение нажать кнопку: в реальной или на бумаге?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# решение"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__б)__ Правда ли, что женщины и мужчины принимают решению о нажатии на кнопку одинаково часто? А на бумаге?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# решение"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__в)__ Правда ли, что женщины намного кровожаднее мужчин и жмут на кнопку гораздо быстрее?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# решение"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__в)__ Предположим, что дисперсия времени нажатия на кнопку определяется эмоциональностью и романтичностью индивида. Правда ли, что романтичность женщин равна романтичности мужчин? "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# решение"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "<img src=\"https://pp.userapi.com/c852220/v852220032/14dce1/r1zXBIig4B8.jpg\" height=\"300\" width=\"600\"> "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}