Types of error

• Type 1

  • when the null hypothesis is rejected and the alternative is accepted when it should be the other way around
  • more likely to do this if the significance level is too lenient
  • leads to a false positive
  • making a claim that a correlation exists, when it doesnt
• A 5% significance level is better to balance between the two
• Type 2

  • when the null hypothesis is accepted and the alternative is rejected when it should be the other way around
  • likely to do this if the sig level is too stringent
  • false negative
  • making a claim a correlation doesnt exist when it does