Lecture 4 - Chi-Squared Analysis Flashcards
What are three types of categorical data?
Binary, Nominal, Ordered
What are two types of continuous data?
Interval (zero is a value and values can go below zero) and Ratio (zero is absolute and no values can go below zero).
Chi-squared analysis is based on expected and observed frequencies, as opposed to central tendency and variance.
True or false?
True.
Who developed Chi-Squared Analysis?
Pearson in 1900.
What question is being asked in a Chi-Squared Analysis?
“How much of the variation of frequencies within a contingency table is due to the relationship between the categorical variables and how much is just due to chance?”
What is the equation for the chi-squared statistic?
X2=sum of ( (observed frequency - model frequncy)squared/model frequency )
How do you calculate expected frequencies
row total x column total / n (number of participants in the sample)
In Chi-squared analysis what is a RESIDUAL?
A RESIDUAL is the observed count minus the expected count.
For Chi-Squared Analysis how do you calculate standardised residuals?
Standardised residuals are calculated by taking the square root of the chi squared stat for each cell.
So, it works out as observed frequency - expected frequency / square root of the expected frequency
How are observed counts determined to be significant or not during a chi-squared analysis?
We look at standardised residuals.
If SR are below -1.96 or above 1.96 then the observed count is considered to be significant and the difference seen in those cells is significantly likely to be due to a relationship between the two categorical variables and not just due to chance.
What values are put in a contingency table that is presented in a lab report?
Observed counts
Expected Counts
Standardised Residuals
How do you calculate degrees of freedom when do a chi-squared analysis?
(Number of rows - 1) x (number of columns - 1)
What does the chi-squared statistic indicate?
The chi-squared statistic (X2) indicates the amount of variation within the contingency table that is due to the relationship between in the variables.
