chi-square and anova Flashcards
it measures how a model compares to actual observed data.
chi square
- helps us see if what we predicted matches what actually happened in real life
The chi-square test is an example of a common approach to statistical analysis known as
statistical modeling
- seeks to develop a statistical expression (the model) that predicts the behavior of a dependent variable on the basis of knowledge of one or more independent variables.
what does the chi-square compares
the size of any discrepancies between the expected results and the actual results, given the size of the sample and the number of variables in the relationship
the values in cell should be a minimum of?
The values must exceed at least 5.
You can’t solve for chi-square if there are less than 20 people. As each cell must have at least 5.
what type of variables are used in chi-square
categorial (nominal or ordinal)
It is a data that compares two variables.
contingency table
p -
q -
p - row
q - column
Values of the cells where the rows and columns intersect can suggest whether or not the two sets are correlated.
IT represents two classifications of a set of counts or frequencies. The rows represent two classifications of one variable
two by two
or
fourfold contingency table
Number of times a particular event is occurring actually when an experiment is conducted in the real world. This data is observed and recorded in actuality from a particular sample of the specific group involved in the study.
observed frequency
- gathered through data gathering
- conducting the actual research
Number of times a particular event is expected to occur or claimed to occur. The figure is computed using a simple formula to predict the outcomes of a specific event using known data.
expected frequency
- existing
the formula if the expected frequency is not given (independent)
f = row sum - column sum / N
what are the steps to find the chi-square
- find expected frequency if not given
- compute X^2
- df
- look at p value by using the df and the level of significance
- conclude
what is the degree of freedom used
df = (p-1)(q-1) =1
what happens if the level of significance is not given
assume that confidence level is 0.05
NO observable value:
HAVE observable value:
NO observable value:
X^2 is smaller than the p value
hence, fail to reject the null hypothesis
HAVE observable value:
X^2 is larger than the p value
reject the null hypothesis, accept alternative hypothesis
