Kaplan Gb513 Unit 6 Assignment

Submitted by: Submitted by

Views: 282

Words: 727

Pages: 3

Category: Other Topics

Date Submitted: 02/24/2014 10:25 PM

Report This Essay

Descriptive Statistics

| BBS | ABN | CBC |

| | | |

Mean | 12.72 | 14.76 | 13.36 |

Standard Error | 0.6054 | 0.3688 | 0.4392 |

Median | 12.85 | 15.05 | 13.25 |

Mode | 10.8 | 14.7 | 13.6 |

Standard Deviation | 2.7075 | 2.2733 | 2.4055 |

Sample Variance | 7.3308 | 5.1679 | 5.7865 |

Kurtosis | -1.4827 | -0.2152 | 0.0080 |

Skewness | 0.1500 | -0.1652 | 0.2329 |

Range | 8.2 | 9.5 | 10.2 |

Minimum | 8.9 | 10 | 8.9 |

Maximum | 17.1 | 19.5 | 19.1 |

Sum | 254.3 | 560.9 | 400.9 |

Count | 20 | 38 | 30 |

Average rating for CBC movies is 13.36, for BBS it’s 12.72 while for ABN it is 14.76. So, CBC is faring better than BBS but is behind ABN. Standard deviation, a measure of variability, for CBC movies is 2.4055 against 2.7075 for BBS and 2.2733 for ABN. This shows ratings for CBC are more consistent than those for BBS but vary more than ratings of ABN movies. Skewness is very small for all three networks, meaning ratings are almost symmetrical around the mean value. Kurtosis shows the peaked-ness of the distribution. BBS ratings are least peaked while ABN and CBC rating distribution have peaks similar to a normal distribution.

Charting

Least square equation for the trendline is y = 0.0637x + 13.143. So, trendline predicts an increase of 0.0637 per month on average. But r-squared value indicates the actual prediction reliability using this equation. A value of 0.0463 means only 4.63% of the variability of ratings is explained by factor ‘month’, thus the equation can’t be reliably used for forecasting monthly ratings. May be including other factors for the least square equation could improve prediction accuracy.

Hypothesis Testing

Null hypothesis: Ratings are equal for movies with stars and those without stars.

Alternative hypothesis: There is a difference in ratings.

Before applying independent sample t-test for means, we checked using F-test whether variances for two populations could be assumed equal or not. Result indicates...