Rough Notes

• adding a feature in amazon like adding a new button, click rate, CTR increase or decrease

• null and alternate,

• more buys? more clicks?

• null hypothesis

  • status quo
  • nothign is going on
  • you reject or fail to reject

• alternate hypothesis

  • logical opp of null hypothesis

• possible conclusions

  • reject $( H_0 )$
  • do not reject $(H_0)$ fail to prove $(H_1)$, but does not mean we have proven $( H_0 )$
  • rejecting the null hypothesis is a stronger statement, because we can claim that the alternative hypothesis testing is true. (at some significance level)
  • failing to reject the null is weaker as it just implies that we had insufficient evidence to provde that $(H_1)$ is true

• example:

  • Mean iq scores between urban and rural areas
  • $H_0$ no diff - this means any observed diff in IQ bw people in rural and urban countries is purely due to random chance or sampling error rather than an inherent diff
  • $H_1$ - there is a significant diff bw people who live in cities and rural areas

• assignment for me

  • find out some statements in trading and create null and alternative hypothesis

• random sampling (urban 90% college educated vs rural 10% college educated)

• normal distribution