# MechaCar_Statistical_Analysis ## Linear Regression to Predict MPG ![Del1](https://github.com/rindneremily/MechaCar_Statistical_Analysis/blob/main/images/Del1.png) Vehicle weight is unlikely to give a random variance as it is below 0.05. The vehicle weight is very small and should be considered to have a zero slope. Yes, the r squared value shows that it is a good model of prediction. ## Summary Statistics on Suspension Coils ![Total summary](https://github.com/rindneremily/MechaCar_Statistical_Analysis/blob/main/images/Total%20summary.png) ![Lot 1 test](https://github.com/rindneremily/MechaCar_Statistical_Analysis/blob/main/images/Lot%201%20test.png) ![Lot 2 test](https://github.com/rindneremily/MechaCar_Statistical_Analysis/blob/main/images/Lot%202%20test.png) ![Lot 3 test](https://github.com/rindneremily/MechaCar_Statistical_Analysis/blob/main/images/Lot%203%20test.png) Yes, the variance is less than the maximum of 100 pounds so it meets the design specifications. ## T-Tests on Suspension Coils ![One sample t test](https://github.com/rindneremily/MechaCar_Statistical_Analysis/blob/main/images/One%20sample%20t%20test.png) This tested the null mean equaling 1500 and the altnerative not equaling 1500. The 95% CI shows that it is between 1497.507 and 1500.053. ## Study Design: MechaCar vs Competition I would perform another multiple linear regression to compare fuel efficiency, average miles driven per month of the consumer, and safety rating. All of this data needs to be quantifiable so I would use miles per gallon, miles driven per month, and a rating out of ten. The null would be that the majority of the population want 30mpg+ in the city, drive 1,000+ miles per month, and want a safety rating 6+. The alternative would be less than 30mpg in the city, drive less than 1000 miles per month, and want a safety rating below 6.