Chi Squared (x^2) Flashcards
What type of data is the Chi squared used for?
Categorical data
How do you calculate the expected count?
Multiply the row proportion (row total/ total count) by the column proportion (column total/total count) by the total count.
This will produce two score (one for each column/row).
This should be the same value for all possible outcomes per category-assuming evidence for the null
What is the Chi-squared equation?
[the sum of the (observed value-expected value)^2]/[expected value]
repeat the equation [(observed value-expected value)^2]/[expected value] then add all the results together
What is the observed value?
The frequency of each outcome occurring for each category
What is meant by the degrees of freedom?
The number of cells that need to be defined so that you have enough data to work out the frequency of the other data values
What is the equation for degrees of freedom?
(number of rows-1)x(number of columns-1)
What are the assumptions of the chi-squared test?
1) the data points are independent (separate groups are used)
2) the data are described by the binomial distribution
3) For a good description of the test, at least 5 expected counts in each cell are needed
If the number of counts is too small, what test can be used instead?
The Fisher’s exact test