A T-test is used to compare the means of two different groups. It is used to see if the two difference between the groups are statistically significant. Which means that differences between the two groups did not happen due to sampling error or by chance.

\({(x_1-x_2) \over (\sqrt{s^2*({1 \over n_1}-{1 \over n_2}}))}\)

-Independent -Normally Distributed -Homogeneity of Variance

We have two Independent data sets. One group being Animation ratings and the other group being Romance Ratings.

ggplot(data=rom, aes(x=rating)) +
  geom_bar(stat="count", width=0.5, fill="steelblue") +
  theme_minimal()

## Warning: `position_stack()` requires non-overlapping x intervals

ggplot(data=rom, aes(x=rating)) +
  geom_bar(stat="count", width=0.7, fill="red") +
  theme_minimal()

## Warning: `position_stack()` requires non-overlapping x intervals

## Make a null hypothesis

H_0: anim=rom H_A:anim!=rom

\({(6.57-6.51) \over (\sqrt{1.17^2*({1 \over 3650}-{1 \over 4704}}))}=\)

You compare the t-test value you got and compare it values in a critical value chart. If t-test value> critical chart value then there is significant difference. If t-test value < critical cart valye then there is no significant difference