Field Project 3

1
Field Project 3

• Ling 102

2
Reporting

• What is wrong with that?
• 85 of the uses of belike were used when
  speaking in a narrative setting

3
Reporting

• What is wrong with that?
• 85 of the uses of belike were used when
  speaking in a narrative setting
• Maybe 85 of the instances of the corpus were
  narrations!
• Good way of putting it
• Of the narrative tokens, x used be like and y
  used say
• Of the not narrative tokens, x used be like
  and y used say

4
Reporting

• Incorrect
• of the x variant of the dependent variable (ie.
  VQ) was used in the y variant of the independent
  variable (i.e. gender)
• Correct
• of the y variant of the independent variable
  (ie. gender) used the x variant of the dependent
  variable (i.e. VQ).
• The ppt VQ.pdf explains the same point (slides 36
  to 39)

5
Reporting statistics

• Incorrect
• Probability of keeping the null hypothesis

6
Reporting statistics

• Incorrect
• Probability of keeping the null hypothesis
• Correct
• Probability of wrongly rejecting the null
  hypothesis

7
Reporting statistics

• Incorrect
• The null hypothesis is that theres not enough
  information to confirm this correlation between
  males/females and the use of belike

8
Reporting statistics

• Incorrect
• The null hypothesis is that theres not enough
  information to confirm this correlation between
  males/females and the use of belike
• Correct
• The null hypothesis is that theres not a
  correlation between males/females and the use of
  belike

9
Reporting statistics

• Does anything look wrong?
• Chi test statistic 0.51679911
• This finding is not statistically significant,
  because there is only a 1 in 2 chance I have
  wrongly rejected the null hypothesis

10
Reporting statistics

• Does anything look wrong?
• Chi test statistic 0.51679911
• P-value 0.51679911
• This finding is not statistically significant,
  because there is only a 1 in 2 chance I have
  wrongly rejected the null hypothesis
• It is a very big probability!

11
Reporting statistics

• I came up with 5.01795E-9 as the probability that
  the null hypothesis is true, making the data
  significant

12
Reporting statistics

• I came up with 5.01795E-9 as the probability that
  the null hypothesis is true, making the data
  significant
• I came up with 5.01795E-9 as the probability that
  the null hypothesis is true,
• making the difference between the groups
  significant
• Meaning that there is a significant correlation
  between x and y

13
Reporting statistics

• With a p-value of 0.9, we must accept the null
  hypothesis

14
Reporting statistics

• With a p-value of 0.9, we must accept the null
  hypothesis
• With a p-value of 0.9, we cannot reject the null
  hypothesis

15
Reporting statistics

• Women use like VQs a bit more than men, but
  this difference is statistically insignificant

16
Reporting statistics

• Women use like VQs a bit more than men, but
  this difference is statistically insignificant
• Women use like VQs a bit more than men, but
  this difference is not statistically significant